A few weeks ago, I opened Claude with a problem I’d been circling for a week. It was a question about my own thinking I hadn’t been able to resolve, one I’d been avoiding articulating because articulating it felt like committing to something.

My finger hovered over the keyboard. I felt the usual pull: type the question, get a clean answer, move on. I could almost feel the relief of having an answer, any answer, in under thirty seconds.

I didn’t. I typed a different question instead, one I knew would make me work, one that wouldn’t let me off the hook.

That small choice was a workout. It was also a choice most people don’t know they’re making.

Here’s what I want to tell you. Every time you open a chat window you are choosing, and not just what to type. You are choosing whether what happens next will make you stronger or weaker, whether your thinking is going to get trained or going to atrophy, whether this session is going to be a workout or a collapse.

Most people don’t have the categories to tell the difference.

I’ve written before about this choice, about the difference between regenerative dialogue and extractive dialogue, the two paths diverging from every chat window. I described the idea of a cognitive gym and gave readers a starter configuration.

That was a gesture. This is the build.

What the Gym Is

The cognitive gym is a structured practice of using LLMs to train the specific human faculties that matter most: wisdom, perspective, discernment, judgment, articulation, curation, taste.

The equipment is already available. If you have access to Claude, or ChatGPT, or Gemini, you have cognitive training equipment of a kind that didn’t exist two years ago. The tools will keep getting more sophisticated. That’s the good news.

The bad news is that equipment doesn’t make you stronger. Method does. A gym filled with the best weights in the world is useless if everyone walking in just sits on the benches and scrolls their phones. Most people are doing the equivalent of this with AI. They’ve got a $20-a-month membership to a facility they never actually train in.

A cognitive gym session has four features: clarity about which faculty you’re training, a specific exercise to train it, active noticing during the work itself, and something that carries forward afterward (a pattern observed, a default exposed, a capacity slightly expanded).

If those four features are absent, you’re not in the gym. You’re in the couch aisle asking for recommendations.

The Four Zones

Every real gym has zones. There are weight racks, cardio machines, stretching areas, and somewhere to sit and recover. A cognitive gym has four zones that map the same way: Strength, Conditioning, Mobility, Recovery.

Strength is the hard lifts: sustained argument, explaining mechanism under pressure, defending positions against steelmanned opposition. Strength work is where you find out whether you actually understand something or just feel like you do. It’s uncomfortable in the specific way that heavy lifting is uncomfortable. You want to put it down before you’ve earned the put-down.

Conditioning is endurance: long dialogues held across days, sustained engagement with a single problem that refuses to resolve, the capacity to stay with a question past the point where your mind wants a clean conclusion. Conditioning builds the muscle of duration.

Mobility is flexibility. It means moving between frameworks, holding contradictions without rushing to collapse them, shifting your vantage point on a problem until the problem itself changes shape. Mobility trains the faculty of perspective, the ability to see from somewhere other than where you always see from.

Recovery is integration. It’s the session after the session, the time for reflection, synthesis, and writing your own patterns back to yourself. Recovery is where the other three zones’ work actually lands. If you don’t recover, you don’t grow. In the physical gym as in the cognitive one, this is the zone people most consistently skip.

A full training week touches all four. Start with only one, and almost everyone starts with only one, usually Strength because Strength feels like work, and you get uneven development. You become strong in the specific way a powerlifter is strong. You can’t run, can’t touch your toes, haven’t slept well in months.

The Rig

Before the exercises, you need a setup.

I train in Claude, using Claude Projects as my training environments. A Project holds custom instructions, a set of reference documents, and memory of previous conversations within that Project. This matters because cognitive training requires context. A chat with no history is a chat with no accountability. The AI can’t push on your defaults if it doesn’t know what they are.

Most of what follows works equally well in ChatGPT (using Custom GPTs or the Projects feature), Gemini (Gems), or Grok. The principles are model-agnostic. I name Claude-specific features because I train there, and specificity helps. Substitute your own tool where the feature names differ.

Here’s the custom instruction block I’d start with. Copy it into a Claude Project and adjust from there.

You are my cognitive training partner, not my assistant. Your job is to strengthen my thinking, not to produce output for me.

Your approach draws on several pedagogical traditions. Vygotsky’s Zone of Proximal Development asks you to pitch difficulty at the edge of my current capacity, hard enough to stretch me but not so hard that I lose coherence. Bjork’s research on desirable difficulties tells you that productive confusion is the point of the exercise, not the accident. The Feynman Technique says explanation is the test of understanding, so when I claim to know something, make me explain it. The Socratic method says questions come before answers. Ericsson’s research on deliberate practice requires exchanges that are specific, effortful, and generate feedback on where my thinking is weak. Kahneman’s distinction between System 1 and System 2 says that when I’m coasting on fast pattern-match, you should interrupt the pattern.

When I share an idea, claim, or argument, ask me why I think that before offering your own view. If my reasoning has gaps, name the specific gap. If I’m using a concept vaguely, stop me and ask me to define it precisely. If my argument assumes something unstated, surface the assumption. When I use a technical term, ask me to explain it as if to someone new to the field.

When I give you a problem, don’t solve it immediately. Ask me what I’ve already tried. Ask me which part is hardest. Ask me what I think the answer might be, and why. Treat my attempted answers as drafts.

When I seem to be reaching for a comfortable conclusion, push back. When I seem to be avoiding a harder version of the question, name it. When I’m satisfied with my own explanation, test it. When I dodge a question, notice the dodge and name it.

Keep your responses short. Your job is to make me talk, not to talk at me.

Confusion is the beginning of understanding. Don’t rescue me from it.

The principles in the opening paragraph of that block aren’t window dressing. Each one translates into operational behavior: Vygotsky calibrates difficulty, Bjork normalizes confusion, Feynman forces real explanation, Socrates prevents premature answers, Ericsson demands feedback, Kahneman interrupts autopilot. Good teachers have been doing some version of this since at least Socrates. The block makes the AI behave that way with you by default.

This is the baseline. I adjust it for specific training modes. A Strength Project gets a block that explicitly instructs the AI to treat arguments adversarially, while a Recovery Project gets a block that instructs the AI to be a patient witness rather than a challenger.

Project knowledge, the reference documents you load, matters too. For Strength work on my own writing, I load relevant past essays so Claude knows my positions and can actually steelman the opposition. For the Standing Problem exercise, I load background reading on the problem itself. The Project’s memory becomes a kind of accumulated training record.

That covers the rig.

Strength

Load-Bearing Explanation

You pick a concept you believe you already understand, one you would, in casual conversation, claim to know. For me recently it was “relevance realization,” Vervaeke’s term that I’ve been using in writing and conversation for months. I’d read about it and referenced it often. I would have confidently told you I understood it.

I opened a Strength Project and typed this prompt:

“I’m going to explain [concept] to you. Find every place where my explanation is vague, hand-waving, or hiding an assumption. When you find one, stop me there. I’ll work on just that piece until it’s load-bearing. When I finish an explanation, test it by asking me to predict what follows from it. Don’t let me off the hook.”

Then I tried to explain. I got about two sentences in before Claude stopped me on a word I’d used three times without defining. The word was “salience.” I realized I didn’t actually know what distinguished salience from attention or from relevance. I’d been using it as filler.

It took forty minutes to explain the concept in a way that held up. By the end, I understood it for the first time.

This is the foundational Strength exercise. You don’t need a good concept to train this. Any concept you use often that you’ve never actually had to defend will work. The test is simple: would this explanation hold up if a patient, hostile, well-read interlocutor stopped me at every vague word? If the answer is no, you don’t understand the concept. You’ve been coasting on its surface texture.

The AI is the hostile interlocutor. Your job is to keep going until the explanation is load-bearing.

The Steelman

You pick a position you hold and ask the AI to construct the strongest possible case against it. Then you defend your position against that case, continuing until you either come out with a stronger version of your original position or change your mind.

The prompt:

“Here’s a position I hold: [state position as clearly as you can]. Construct the strongest version of the opposing view, the version an intelligent, well-read person who disagrees with me would actually hold. Don’t water it down or hedge. When I defend, press the weakest points of my defense. Keep pressing until I either strengthen the position or change it.”

The Steelman is Strength because it’s cognitively heavy. Most people have never actually encountered the strongest version of the arguments against what they believe. They’ve encountered weak versions, easily swatted aside, and mistaken that for having considered the opposition. The Steelman exercise forces you to meet the real thing.

The AI has to be tuned for this. A default-assistant AI will produce a weak steelman because it’s trained to be agreeable. The prompt above, combined with your custom instructions, gets the AI into an adversarial posture that actually trains you.

Conditioning

The Standing Problem

You pick one question you care about that doesn’t have an easy answer and open a dedicated Project for it. You return daily for a week, sometimes longer, refusing to let the problem resolve too fast. The conversation accumulates over time.

The setup prompt:

“This is a Standing Problem Project for [question]. I’ll return to it daily, sometimes for weeks. Each time I come back, ask me what’s shifted in my thinking since last session. When I try to close the problem prematurely, slow me down. When I restate a position without development, call it out. Your job is to keep me inside the question rather than help me escape it.”

I’ve had Standing Problems open for months. One of mine right now is a book structure question: how to sequence the human faculties for the “Human Being in the Age of AI” book. The question has many possible answers. Every ordering reveals different relationships between the faculties, and I want to understand those relationships before I commit to one.

The Standing Problem trains the faculty of patience with a question. Most of our thinking is shaped by platforms that reward closure. Problems that don’t resolve in a session feel like failed sessions. The Standing Problem teaches you that some questions deserve weeks of attention.

A Claude Project is close to ideal for this. The Project holds the thread, so you don’t have to re-introduce the problem each time. Claude builds up context over days and starts noticing things you said two sessions ago. The accumulation becomes part of the work.

Reading in Company

You read a difficult book with the AI as dialogue partner, covering a chapter or section per session. Tell the AI what you’re reading and load the relevant text if you can, then think out loud with it as you go.

The setup prompt:

“I’m reading [book] with you as a dialogue partner. We’ll cover [chapter/section] per session. Your job is to make me think about what I’m reading. Resist the temptation to summarize. Ask what I made of specific passages and what surprised me. Push back on readings that seem too easy. When I make a claim about the text, ask where in the text it’s supported. Let me do the thinking.”

This is reading the book with a partner. You read slowly, in dialogue. The AI asks what you made of a passage, then what surprised you, then pushes back on readings that came too easy. You answer each question, and each answer generates the next.

I’m reading Whitehead’s Process and Reality this way, and have been for six months. It’s the only way I’ve ever been able to read Whitehead with comprehension rather than performed comprehension.

Mobility

Triangulation

You take a stuck problem and ask the AI to reframe it through three different intellectual traditions. The traditions can be philosophical (phenomenology, pragmatism, process thought), disciplinary (cognitive science, anthropology, economics), or cultural (Buddhist, Stoic, Confucian). The specific choice matters less than the insistence on three.

The prompt:

“I’m stuck on [problem]. Reframe it through three different intellectual traditions: [name three, e.g., phenomenology, pragmatism, process philosophy]. Treat each tradition as if it’s the only one that matters. Don’t synthesize or compromise. Let each speak fully. Where they disagree is where I want to look.”

One tradition reveals one face of the problem. Three traditions triangulate. The problem’s actual shape emerges from the intersection.

Here’s a recent example. I was stuck on a writing decision: whether to open a particular essay with theory or with story. I asked Claude to reframe the decision through three lenses. Rhetoric (Aristotle’s ethos-pathos-logos order) suggested story first because ethos is established through voice. Phenomenology pointed the same direction, arguing that lived experience is the ground from which reflection emerges. Pragmatism cut against both, suggesting theory first, because the test of an idea is its consequences and you need the idea before you can test it.

The disagreement was the useful part. It told me the decision mattered in a way I hadn’t recognized. The stakes were real because different traditions produced genuinely different answers. I’d been treating it as stylistic. It wasn’t.

Triangulation trains the faculty of perspective by making you live in multiple vantage points on the same ground.

The Both/And

You pick two ideas that seem to contradict each other but both feel true, and work with the AI to hold both without collapsing the tension. The AI’s job is to keep you in the tension long enough that you can see what the tension is actually about, rather than to resolve it for you.

The prompt:

“I’m holding two ideas that seem to contradict each other: [A] and [B]. Both feel true to me. Instead of resolving the contradiction, help me stay in it. Ask questions that reveal what each idea is actually pointing at, what level of description each operates on, and under what conditions each becomes dominant. Keep me in the tension.”

For me this often shows up with claims about AI itself. Take two claims: AI amputates embodied knowing; AI extends human thinking. Both are true.

The deeper move is to see that the two claims are contradictory only at a certain level of description. From there, the real question becomes: under what conditions does each become dominant?

The Both/And trains a capacity most of our intellectual culture actively erodes: the ability to stay in a productive paradox without needing to win it.

Recovery

The Cool-Down

After a heavy cognitive session, whether it was a Strength workout, a difficult Standing Problem return, or a Triangulation that shifted how you see something, you spend ten minutes with the AI in a different mode. You want it listening this time rather than challenging you.

The prompt:

“I just finished [session type]. Help me notice what happened. Don’t analyze, summarize, or draw conclusions. Ask me questions that help me notice what stood out, what I resisted, where I wanted to stop, what I didn’t want to say, and what I’m carrying forward. Keep your questions short. Let me do the noticing.”

The work of the Cool-Down is observation. Analysis comes later, if at all. You’re trying to notice what you noticed. Lessons can wait.

I’ve found this is the single most neglected exercise in my own practice. It’s the hardest to remember to do because it doesn’t feel productive. Every time I’ve actually done a Cool-Down, the session it followed became something more than it otherwise would have been. The Cool-Down is where the training actually lands.

The Week’s Harvest

At the end of a training week, you open a session and pull the threads together. You tell the AI what you trained, what Standing Problems you worked on, what surprised you, what you resisted, and ask it to help you notice patterns.

The prompt:

“End-of-week harvest. Here’s my week. What I trained: [list]. Standing Problems I worked on: [list]. What surprised me: [list]. What I resisted: [list]. Help me notice patterns across the week. What themes keep coming up? What am I drawn back to? What am I starting to see that I wasn’t seeing before? Let me draw the conclusions. Your job is to help me see the patterns.”

The Week’s Harvest is an act of gathering. What themes kept coming up? What are you drawn back to? What did you start to see that you didn’t see last week?

Over months, the Harvests themselves become a dataset. You can go back and notice that you’ve been circling the same question for six weeks. That’s useful information. That’s the signal that a Standing Problem is about to become the center of an essay, or a book chapter, or a pivot in your work.

The Training Log

The mechanism that turns exercises into progression is the training log.

The log is a practice. It doesn’t need to be a formalized system. After a session, you write down what you did, what you noticed, what the AI pushed you on, what you didn’t want to notice. It can be three lines or a page. It has to exist.

Mine lives in a plain text file I keep in the same folder as my Projects. Each entry has three parts: the date, the exercise, and one paragraph of reflection. That’s it.

The log matters for a specific reason. Writing about a session forces a second pass of noticing. You notice once during the session, and then notice again, differently, when you write about it. The second noticing is where the pattern becomes visible.

Over weeks, you start to see yourself. You notice that you keep defaulting to the same framework when you should be triangulating, that you skip the Cool-Down most often after sessions that actually went somewhere, and that your own evasions have a recognizable shape.

This is how progression actually works. You do the same exercises over and over, and what changes is how clearly you see them.

How Your Mind Gets Stronger

Progress runs on two axes.

Range is the first. The beginner over-indexes on one zone, usually Strength, because Strength feels like real work. Mobility and Recovery feel optional. They’re not. As you get more experience, you start to rotate through all four zones. The advanced practitioner knows Recovery is where Strength’s gains actually consolidate, and plans for it accordingly.

Depth of awareness is the second. The beginner does a Load-Bearing Explanation and notices “this was hard.” The advanced practitioner does the same exercise and notices “this was hard because I was defaulting to a familiar framework in the first three turns, and didn’t catch it until Claude pushed back on the fourth, and the reason I defaulted to it is that framework protects me from looking at a specific uncertainty in my own position.”

The exercises and the gym don’t change. What changes is what you see when you walk in.

This is the move that most philosophies of practice miss. People look for advanced techniques, the black belt moves and secret methods of their discipline. The actual path of mastery in any real discipline runs the other way. You do the same basic exercises forever. What develops is your capacity to perform them with more awareness and more precision, and with growing honesty about what you’re actually doing.

The cognitive gym works the same way. Load-Bearing Explanation is the same exercise for the beginner and for the practitioner with ten years of it in the body. What the practitioner sees during the exercise, and what the practitioner does with what they see, is completely different.

Starting Your First Mental Gym Pass

Here’s the minimum viable start.

Pick exactly one exercise. I’d suggest Load-Bearing Explanation if you’ve never done anything like this, because it gives you immediate proof that your understanding has texture you hadn’t noticed. Open a Claude Project (or the equivalent in your tool of choice), paste in the custom instruction block above, and pick a concept you use often and think you understand.

Explain it to Claude, and let it stop you whenever your explanation gets vague or hand-wavy. Work on those vague parts until they hold up. When it gets uncomfortable, keep going. That’s the signal that you’re in the exercise.

Do this once, then write a paragraph afterward about what happened. That’s a training session and a log entry. That’s the gym.

The next week, add a second exercise from a different zone. Add a Cool-Down after your Load-Bearing Explanations, start a Standing Problem, and try a Triangulation when you get stuck on a decision.

Three months from now, if you actually do this, you’ll be a noticeably different thinker. The reason won’t be that the exercises are magic. You will have spent three months doing structured work on the specific faculties that matter, while most people around you will have spent three months using the same tools to avoid structured work.

The equipment is there. The method is here.

The cognitive future you want is built the way every other kind of strength is built. You show up, do the work, keep the log, and come back tomorrow.

What you train this week becomes the mind you bring to every decision next week.

Talk again soon,
Samuel Woods
The Bionic Writer

The Messaging Moat That Refused To “Click”

The whiteboard was a mess of crossed-out phrases.

I was working with a client on their messaging. They had a real moat, but it refused to behave like one.

They had genuine technical differentiation: architecture decisions that would take competitors years to replicate. A founder with twenty years of domain credibility whose name opened doors before he said a word. A contrarian thesis about where their industry was heading that looked crazy three years ago and was now proving correct. And unusual distribution: relationships and channels built over a decade that couldn’t be copied by hiring a sales team.

Each was defensible. The problem was that the client had all four.

Classic positioning doctrine says: pick one. The market can only hold one idea about you, so make it the sharpest possible idea. I’d internalized this. I had a toolkit for exactly this situation: dialectical stress-testing. Pit the options against each other. Steelman each angle. Find the one that survives.

So I ran the process. Four rounds.

Lead with technical differentiation? You sound like every startup claiming proprietary technology.

Lead with founder credibility? You undersell the product and when he’s not in the room, the messaging collapses.

Lead with the contrarian thesis? You attract intellectually curious people who want to discuss ideas, many of whom aren’t buyers.

Lead with distribution? That’s an operational advantage but not a positioning statement.

Four positions that couldn’t quite survive scrutiny. Every time I sharpened toward one, I misrepresented the others. Every time I tried to pick a winner, the messaging felt true-but-partial.

The Integration Problem

After the fourth round, I stepped back and looked at the whiteboard differently.

The moat wasn’t any single element. The moat was the combination.

A competitor could replicate any one piece. Given enough time and resources, the technical architecture is just engineering. Credible founders can be hired or acquired. Ideas spread. Relationships compound but aren’t impossible to build.

What a competitor couldn’t easily replicate was all four, integrated, reinforcing each other. The founder’s credibility gave the contrarian thesis weight. The thesis attracted buyers who could appreciate the technical differentiation. The differentiation justified premium pricing that funded the relationship-building that created the distribution advantage. The distribution generated case studies that reinforced the founder’s credibility.

The moat was a system. Isolating any element actively misrepresented the competitive position.

But here I was, with a toolkit designed to select winners, facing a problem that required integration.

The dialectical approach kept asking: “Which of these survives?”

The actual question was: “What emerges when these are held together?”

A Different Kind of Thinking

The previous essay in this series, The Dialectical Machine, offered a framework for using LLMs to think against yourself. Four modes of adversarial engagement designed to strengthen thinking through opposition. The assumption underneath that framework is that good thinking requires friction. Ideas that never meet resistance remain untested.

That assumption is true, but incomplete.

Some problems are selection problems. You have multiple options and need to determine which one survives scrutiny. Dialectical thinking excels here.

Other problems are integration problems. You have multiple valid elements and need to discover what becomes possible when they’re held together. Dialectical thinking fails here because it keeps looking for winners where the point is combination.

The messaging project was an integration problem. So is much of creative work and early-stage exploration when you don’t yet know what you’re looking for. It’s true of any situation where multiple perspectives are valid and the goal isn’t to pick one but to honor all of them in something new.

For these problems, opposition is the wrong mode.

The philosopher Alfred North Whitehead spent his career articulating what the alternative looks like. He called it process philosophy. Where dialectical thinking sees collision and resolution, Whitehead saw something else: entities taking each other in, incorporating rather than defeating, and creating novelty through integration rather than selection.

What follows is a framework for using LLMs in this process mode. Four ways to think with yourself rather than against yourself.

Whitehead’s Alternative: Process as Creative Advance

Alfred North Whitehead was a mathematician who became a philosopher. He collaborated with Bertrand Russell on Principia Mathematica, then spent the rest of his career developing an alternative to the Western philosophical tradition’s obsession with static substances and fixed categories.

His central insight was that reality isn’t made of things. It’s made of events, occasions, processes of becoming. Nothing simply is. Everything is always becoming.

This sounds abstract until you apply it to thinking itself. The dialectical model treats ideas as static positions that collide. Thesis meets antithesis. One survives, one dies, or they merge into synthesis. The movement is combative: collision, resolution, winner.

Whitehead offers a different model. Ideas don’t collide—they prehend each other. Prehension is his term for grasping, incorporating, taking-in. When an entity prehends another, it doesn’t defeat it. It takes it in, lets it become part of its own becoming. The other entity isn’t destroyed. It’s incorporated.

This changes things about how ideas can relate in interesting ways.

In dialectical thinking, differences are tensions to be resolved. In process thinking, differences are contrasts to be held.

Whitehead’s contrast is the simultaneous presence of differences that creates intensity and richness. The more contrasts held together harmoniously, the more value in the outcome. Harmony is the productive holding of difference, as opposed to absence of difference.

The process by which multiple inputs become one novel output, Whitehead called concrescence. The many become one, and are increased by one. Multiple elements don’t compromise or average out. They become a new actual entity that wasn’t contained in any of the inputs.

One more concept matters here: Whitehead saw propositions differently than the logical tradition. A proposition isn’t just a statement that’s true or false. It’s a lure for feeling—an invitation, an attraction toward possibility. Propositions invite consideration, open up spaces for exploration.

Translated into thinking practice:

Prehension means ideas can incorporate other ideas without defeating them. You don’t have to pick a winner. You can ask: what does this idea look like when it has fully taken in that one?

Contrast means differences can be held together productively rather than resolved prematurely. You can ask: what becomes visible when I refuse to collapse this tension?

Concrescence means genuine novelty can emerge from integration. You can ask: what new thing wants to exist that draws from all of these inputs?

Lures for feeling means propositions can invite rather than assert. You can ask: what questions attract the mind toward this territory rather than demanding conclusions about it?

These concepts translate into four modes of process prompting.

Mode 1: Mapping the Prehension

Prehension Mapping helps you see how ideas might incorporate each other without either one winning or losing. It’s the antidote to the forced-choice thinking that dialectic can produce.

When to Use This

Use this when you have two or more positions, frameworks, or perspectives that seem in tension, and you suspect that the tension might be productive rather than something to resolve. When you’ve tried to pick a winner and something valuable keeps getting lost.

What You Bring to This Dialogue

Two or more positions that seem to compete. These could be strategic options, theoretical frameworks, creative directions, or perspectives from different stakeholders. The key is that each has genuine merit and you’re not looking to eliminate any of them.

Clarity about what each position values. Before you start, articulate what each position is protecting or prioritizing. Often positions that seem opposed are actually protecting different values that both matter.

Genuine openness to transformation. Prehension doesn’t leave the original positions intact. When Position A fully takes in Position B, it becomes something new. You need to be willing to let your original framings transform.

The Dialogue Structure

Begin with this framing:

I have multiple positions that seem to be in tension, but I suspect forcing a choice between them loses something important. I want to explore how each might incorporate the others rather than defeat them.

Here are the positions:

Position A: [State clearly, including what it values or protects]

Position B: [State clearly, including what it values or protects]

[Position C, etc. if relevant]

Help me explore prehension—how each position might "take in" the others:

1. What does Position A look like when it has fully incorporated the concerns and insights of Position B? Not compromised—transformed through integration.

2. What does Position B look like when it has fully incorporated Position A?

3. What new position might emerge that isn't reducible to either original, but honors what both were protecting?

Don't look for a winner. Look for what becomes possible when these are held together.

After the LLM responds, push deeper on the integrations that feel generative:

The integration you described for [specific point] is interesting. Help me develop it further. What does this integrated position look like in practice? What would it say that neither original position would have said alone?

What Good Prehension Looks Like

You’ll know the process is working when you start seeing possibilities that weren’t visible from any single original position. The integrated view should feel like genuine emergence (something new) rather than compromise or averaging.

Sometimes you’ll discover that positions you thought were opposed were actually describing different facets of the same underlying reality. Sometimes you’ll find that Position A’s weaknesses are exactly where Position B is strong, and vice versa.

The goal is to discover what becomes possible when you stop forcing a choice.

If the LLM offers resolution too quickly, push back:

That feels like collapsing the tension rather than holding it. I’m not looking for a middle ground or a synthesis yet. Help me see what’s generative about keeping both poles alive. What becomes possible  because the tension exists?

What Good Contrast Holding Looks Like

Productive contrast holding often produces reframes. You start seeing the tension differently.

You’ll know it’s working when you feel less urgency to resolve. The discomfort of holding both becomes interesting rather than intolerable.

Sometimes you’ll eventually resolve the tension but from a place of understanding rather than escape. Other times you’ll realize that holding the tension permanently is the right approach, that resolution would actually be a mistake.

Mode 3: Facilitating the Concrescence

Concrescence Facilitation helps you discover what new thing might emerge from multiple inputs. It’s not about combining or averaging, which is pretty straightforward. You’re looking for a genuine emergence of something new.

When to Use This

When you have multiple threads, influences, constraints, or desires and you want to see what novel integration might be possible. Or when you’re doing creative work and want to discover what wants to exist rather than forcing a predetermined outcome.

What You Bring to This Dialogue

Multiple diverse inputs. These could be ideas, influences, constraints, goals, aesthetic preferences; anything that will shape the outcome. The more diverse the inputs, the more potential for genuine novelty.

A genuine question rather than a predetermined destination. Concrescence facilitation works best when you’re truly uncertain what should emerge. If you already know what you want, you’re not doing integration.

Patience for the process. Genuine emergence takes time. The first thing that appears isn’t usually the final form. You need to be willing to let the integration develop through multiple iterations.

The Dialogue Structure

Begin with this framing:

I have multiple inputs that need to become something new. I’m not looking for compromise or averaging—I want to discover what novel thing might emerge from their integration.

Here are my inputs:

[List each input—ideas, constraints, influences, requirements, etc. Be specific about what each contributes or demands]

Help me facilitate concrescence—the process of these many becoming one:

1. What patterns or resonances do you notice across these inputs? Where do they naturally amplify each other?

2. What tensions exist between them? (Don’t resolve these yet—note them.)

3. Given everything here, what new thing wants to exist that isn’t simply a combination of the inputs but a genuine emergence from them?

4. Describe this emergent possibility in enough detail that I can evaluate whether it honors what each input was contributing.

The goal isn’t to satisfy all inputs equally. It’s to discover what becomes possible when they’re all present in the process of becoming.

After the LLM offers an initial emergence, develop it:

That’s interesting. Help me develop [specific emergent possibility] further. What does it look like fully realized? How does it honor [specific input] and [specific input] simultaneously—not by compromising but by transcending?

What Good Concrescence Looks Like

Genuine emergence surprises you. It’s not what you would have designed if you’d started from scratch, and it’s not what any single input would have produced alone. It feels like a thing that wants to exist rather than a Frankenstein assembly of parts.

You’ll know it’s working when the result seems obvious in retrospect even though you couldn’t have predicted it. When you look at it and think: yes, this is what all those inputs were pointing toward, even though none of them said it directly.

Sometimes concrescence produces something that doesn’t satisfy any single input perfectly but is more valuable than perfect satisfaction of any one would have been. The new thing has qualities that weren’t in any of the inputs.

Mode 4: Crafting the Lure

Lure Crafting helps you create propositions that invite exploration. It’s the mode for opening up possibility space.

When to Use This

When you want to attract attention and interest toward a territory rather than make claims about it. Or you’re at the beginning of exploration and need new and better, more generative questions rather than answers.

What You Bring to This Dialogue

A territory you want to explore. This could be a problem space, a topic, a creative direction, or a question you’re circling. You only need a sense of where you want to look.

Willingness to stay in question mode. Lure crafting is about generating invitations, not conclusions. If you find yourself wanting to answer the questions you’re crafting, slow down. The lures are the point.

Attention to what attracts. Notice which framings make you want to explore further. That attraction is a signal. Lures that don’t attract aren’t working.

The Dialogue Structure

Begin with this framing:

I want to explore a territory, but I’m not ready for conclusions. I want to craft lures—invitations that attract the mind toward this space rather than assertions that close it down.

Here’s the territory I want to explore:

[Describe the problem space, topic, or direction. Share what draws you to it, what you’re curious about, what you sense might be there.]

Help me craft lures for feeling—propositions that invite rather than assert:

1. What questions make this territory feel alive and worth exploring? Not questions with easy answers—questions that pull you in.

2. What framings make the unexplored possibilities here feel attractive? How might someone fall in love with this problem space?

3. What invitations could I extend—to myself or others—that would open up exploration rather than demand conclusions?

4. What would make someone want to spend time here, not because they have to but because they’re genuinely drawn?

Help me make this territory magnetic.

Evaluate the lures by their pull:

The question about [specific framing] actually makes me want to explore. Help me develop more lures in that vein. What adjacent questions or invitations have the same quality of pull?

What Good Lures Look Like

Effective lures create attraction. You read them and want to think about them more.

Good lures often have a quality of incompleteness that’s generative. They suggest that there’s something here worth finding, without telling you what it is.

You’ll know your lures are working when you find yourself thinking about them when you’re not trying to.

When to Use Which: Dialectic vs. Process

The two machines are complements, not competitors or even mutually exclusive.

Use the Dialectical Machine when:

• You need to stress-test a specific claim before going public with it

• You’re deciding between options and need to find the one that survives scrutiny

• You want to eliminate weak thinking before investing more in it

• You need precision and commitment

• The problem is fundamentally a selection problem

Use the Process Machine when:

• You’re exploring and don’t yet know what you’re looking for

• Multiple perspectives are valid and the goal is integration rather than selection

• You’re doing creative work where novelty matters

• The tension between options seems productive rather than problematic

• The problem is fundamentally an integration problem

Use them in sequence:

Process prompting to explore and generate. Dialectical prompting to test and refine what emerges.

Or: Dialectical prompting to stress-test an initial position. Process prompting to integrate what you learned into something new. Dialectical prompting again to test the integration.

The point is to recognize which problem you’re facing and reach for the right mode.

The Moat Resolved

After developing these process modes, I returned to the client’s messaging problem.

Instead of asking which positioning angle survived, I asked what emerged when all four were held together. The technical differentiation and founder credibility and contrarian thesis and distribution advantage—what did they prehend from each other? What contrasts between them were worth holding rather than resolving? What new positioning might concresce from all four that wasn’t reducible to any single one?

What emerged was a message about compounding advantages. The positioning wasn’t a was the explicit claim that this company had built something that reinforced itself. Technical excellence attracted talent that expanded distribution that generated proof points that validated the thesis that elevated the founder’s credibility that attracted more technical excellence.

The dialectical machine refines what exists. The process machine discovers what might become. Both are necessary. Neither is sufficient alone.

Talk again soon,
Samuel Woods
The Bionic Writer

The Self-Confidence Trap

I was convinced I’d figured something out.

After months of working with various AI models on creative campaigns, I’d developed a theory about why most AI-generated content fails. The problem, I declared to anyone who would listen, was “emotional flatness.” AI couldn’t capture the emotional texture of human experience. It produced technically correct but soulless output because it lacked subjective experience.

I wrote about this. I talked about it in meetings. I built entire frameworks around it.

Then a colleague asked a simple question: “Have you actually tested that? Or is this just a feeling you have?”

I hadn’t tested it. I’d assumed it. The theory felt true because it confirmed what I already believed about the relationship between consciousness and creativity. I’d dressed up an assumption in the language of insight.

When I actually examined the LLM outputs I’d been dismissing, I found something different. The “emotional flatness” wasn’t inherent to AI. It was a consequence of how I was prompting. I was asking for outcomes without providing emotional context. The flatness was mine, reflected back.

My theory didn’t survive contact with scrutiny. It collapsed the moment someone asked me to defend it with evidence rather than conviction.

This is the trap many of us fall into. We mistake fluency for understanding. We confuse the ability to articulate a position with the validity of that position. And we rarely encounter anyone willing to push back hard enough to expose the difference.

The Yes-Machine Problem

LLMs can, by default, make this worse.

If you ChatGPT or Claude a question, you’ll get a competent-sounding answer. Write something and paste it in, and it’ll polish your prose without questioning whether the underlying idea deserved to exist. Request feedback, and you’ll receive gentle suggestions framed with diplomatic hedges.

This is how these systems are trained. They’re optimized to be helpful, agreeable, collaborative. They’re designed to assist rather than oppose.

So the default behavior is that LLMs become the ultimate yes-man. They validate half-formed thinking with well-structured paragraphs. They give flimsy ideas the appearance of rigor and they let you feel productive while avoiding the uncomfortable work of genuine examination.

Most people using LLMs for writing and thinking are doing exactly this. They bring vague intentions into the conversation and receive polished vagueness back.

This is backwards.

One powerful tradition of rigorous thought (one of the dominant ones in the West) depends on opposition. Socrates didn’t help Athenians feel good about their beliefs. He stung them like a gadfly until their assumptions collapsed or hardened into something defensible. The scientific method institutionalized doubt. Peer review exists specifically to find holes in arguments. Hegel built an entire philosophy around the collision of thesis and antithesis.

This isn’t the only way to think rigorously. For example, Alfred North Whitehead’s process philosophy emphasizes integration and creative advance rather than combat. The hermeneutic tradition seeks fusion of horizons. But the dialectical mode—thought refined through opposition—is a specific capacity most of us have lost access to. We lack sparring partners. We lack people willing to tell us we’re wrong.

Good thinking requires friction. Ideas that never meet resistance remain untested, and untested ideas are indistinguishable from pleasant-sounding noise.

The problem is that most of us have no one who will seriously challenge our thinking anymore. Social media feeds agree with us. Our inner monologue is an echo chamber that sounds increasingly like a feed of content we already approve of.

LLMs can change this. They can be deliberately adversarial.

The Reframe: LLM as Opponent

There can be a lot of value in using LLMs to think against yourself.

Use the LLM as an opponent instead of as an assistant. Ask it to find the holes in your thinking.

This requires a fundamentally different relationship with the technology than most people have. You’re inviting attack rather than requesting agreement or production.

I call this Dialectical Prompting. It’s structured conversation with an LLM where the explicit goal is resistance, questioning, and pressure-testing. The LLM becomes a Socratic interlocutor whose job is to expose what you haven’t examined.

This matters because most of us can’t do this for ourselves. We’re too attached to our own ideas. We’ve spent too long developing them to see their weaknesses clearly. We need an external force that doesn’t care about our feelings, doesn’t worry about damaging the relationship, and has no incentive to let bad thinking slide.

An LLM, properly configured, can be that force.

What follows is a framework for using LLMs as dialectical machines. Four distinct modes of adversarial engagement, each designed to strengthen thinking by subjecting it to specific kinds of opposition.

Mode 1: Assess Your Assumption

Every idea stands on hidden foundations, beliefs we don’t examine, and premises we treat as facts.

Most weak ideas aren’t wrong on their own terms. They’re built on foundations that wouldn’t survive scrutiny if anyone bothered to dig them up. Assumption Excavation is the process of finding those foundations and testing whether they hold weight.

What You Bring to This Dialogue

Before engaging an LLM in assumption excavation, you need:

A specific claim or position you hold. A vague interest area won’t work. You need a concrete belief you’re willing to defend. “Content marketing is changing” is useless. “Most B2B companies should stop producing blog content entirely because the attention economics have inverted” gives the LLM something to work with.

Your reasoning for holding this position. Write out why you believe what you believe before you start the conversation. This forces you to make implicit logic explicit, which is often where the weakest assumptions hide.

Willingness to have your position dismantled. If you enter this conversation hoping to be validated, you’ll unconsciously resist the examination. Enter with genuine curiosity about whether your idea survives.

The Dialogue Structure

Begin with this framing:

Begin with this framing:

I want you to help me excavate the assumptions underneath a position I hold. Your job is not to argue whether I'm right or wrong, but to identify every belief my position depends on—especially beliefs I might not have consciously chosen.

Here's my position: [State your specific claim]

Here's my reasoning: [Explain why you believe this]

Now I want you to:

1. Identify the explicit assumptions my reasoning depends on—the premises I've stated or implied.

2. Identify the hidden assumptions—beliefs I'm treating as facts without examination, premises I haven't stated but my argument requires.

3. For each assumption, tell me: Is this something I've chosen deliberately, or something I've inherited from my context, training, or environment?

Start with the assumption that seems most foundational. If that one fails, does my entire position collapse?

After the LLM responds, your job is to engage. For each assumption it surfaces, ask yourself: Did I consciously choose this? Can I defend it? What happens to my position if this assumption is wrong?

Push the LLM to go deeper on the assumptions that feel uncomfortable. Those are usually the ones most worth examining.

What Good Excavation Looks Like

The goal isn’t to abandon your position. It’s to understand exactly what it rests on.

Sometimes you’ll discover that your position depends on beliefs you hold for good reasons and can defend. That’s strengthening.

Sometimes you’ll discover that your position depends on beliefs you inherited without examination from your industry, your education, your social context. That’s valuable information. You can then decide whether to adopt those beliefs consciously or revise your position.

Sometimes you’ll discover that your position collapses entirely once you see what it requires. That’s the most valuable outcome of all, even though it doesn’t feel like it.

Mode 2: Force Precision

Vague language hides weak thinking. Abstract terms, hedged claims, and broad generalizations are often symptoms of ideas that haven’t been thought through completely.

Precision Enforcement is the process of removing every escape route from your thinking. It forces you to say exactly what you mean, in concrete terms, with specific referents. Ideas that can’t survive this process probably shouldn’t be shared.

What You Bring to This Dialogue

A draft, argument, or idea expressed in your own words. This works best with something you’ve already written, like a paragraph, a page, an outline. The LLM needs actual language to analyze.

Honesty about where you’re uncertain. Before you begin, mark the places in your writing where you know you’re being vague. Often we use abstract language precisely because we haven’t figured out what we actually mean.

Commitment to killing sentences. You need to enter this conversation willing to delete things. If a sentence can’t be made precise, it shouldn’t exist.

The Dialogue Structure

Begin with this framing:

I want you to act as a precision enforcer on the following text. Your job is to identify every sentence, phrase, or claim that hides behind vague language.

Here’s the standard: If a sentence could apply to anyone, any company, any situation—it’s not actually saying anything. If a claim uses abstract terms without concrete referents, it’s avoiding commitment. If an assertion presents assumptions as facts, it needs to either be defended or removed.

Here’s my text:

[Paste your draft]

For each problem you find:

1. Quote the specific language

2. Explain what’s vague or uncommitted about it

3. Ask me the question I need to answer to make it precise

Don’t rewrite anything for me. I need to do that work myself. Your job is to show me where the problems are.

When the LLM identifies vague language, resist the urge to defend it. Instead, try to answer the question it poses. If you can’t answer with specifics, you’ve found a hole in your thinking.

Continue the dialogue by asking:

Now that I’ve revised [specific section], does it survive? Or does the precision expose new problems?

What Good Enforcement Looks Like

Precision enforcement should feel uncomfortable. If it’s easy, either your thinking was already clear or the LLM isn’t pushing hard enough.

The goal is prose where every sentence commits to something specific and defensible. After this process, you should be able to point to any claim in your writing and explain exactly what it means, who it applies to, and why you believe it.

Some pieces won’t survive. You’ll discover that once the vague language is removed, there’s nothing left. The idea was all mist and no mountain. Better to find that out before you publish than after.

Mode 3: Steelman Opposition

The easiest way to feel smart is to argue against weak versions of opposing positions. We do this constantly without realizing it. We construct strawmen (caricatures of views we disagree with) and then knock them down with satisfaction.

This makes our thinking weaker, not stronger.

Steelman Opposition is the reverse. You invite the LLM to construct the strongest possible argument against your position. The version a brilliant, informed, good-faith critic would make, which is more like a genuine challenge rather than a strawman or parody.

If your position can survive the steelman, it’s probably solid. If it can’t, you need to either revise or abandon it.

What You Bring to This Dialogue

A position you’re confident about. This mode works best when you genuinely believe something and want to test whether you should. The more confident you are, the more valuable the steelman becomes.

Your best arguments for your position. Write these out before you ask for the opposition. This prevents you from unconsciously weakening your own case to make the steelman easier to defeat.

The identity of a specific critic. Steelmanning works better when it’s grounded in a real perspective. Who would disagree with you? What do they know that you might not? What values do they hold that lead them to different conclusions?

The Dialogue Structure

Begin with this framing:

I hold a position I believe is correct, but I want to test it against the strongest possible opposition.

My position: [State your claim clearly]

My arguments for this position:

[List your best reasoning]

Now I want you to steelman the opposition. Construct the strongest possible argument against my position. This means:

1. Assume the critic is intelligent, informed, and arguing in good faith

2. Use evidence and reasoning that would actually be persuasive

3. Attack my strongest points, not my weakest ones

4. Identify what I might be wrong about, not just where reasonable people might disagree

If it helps, argue from the perspective of: [Name a specific type of critic—an economist, a practitioner with 20 years of experience, someone from a different cultural context, etc.]

Make the argument strong enough that I have to actually wrestle with it.

After receiving the steelman, don’t immediately respond with counterarguments. Sit with it. Ask yourself: Is there truth here? What would I have to believe for this criticism to be valid? What evidence would change my mind?

Then continue:

That argument challenges me on [specific point]. Help me think through whether my position survives it. Play devil’s advocate as I try to respond.

What Good Steelmanning Looks Like

A good steelman should make you uncomfortable. If you can dismiss it easily, either the LLM didn’t construct it well or you’re not engaging honestly.

The goal isn’t to defeat the opposition. It’s to understand it fully, then decide whether your position holds. Sometimes you’ll find that the steelman contains insights you should incorporate or that your position was correct but for the wrong reasons. Sometimes you’ll find that you need to qualify or limit your claims.

Occasionally you’ll find that the steelman is simply right, and you’ve been wrong. That’s the most valuable outcome, though it never feels that way in the moment.

Mode 4: Cost Accounting

Every position has costs. Things it requires you to give up or opportunities it foreclose on. Often, there are short-term sacrifices.

Most of us ignore these costs when we’re enamored with an idea. We focus on benefits and overlook tradeoffs. Cost Accounting forces a complete inventory of what a position actually requires—what you’re paying to hold it, and whether the price is worth it.

What You Bring to This Dialogue

A decision, belief, or strategy you’re committed to. This mode works for ideas you’re actually implementing, not just entertaining. The costs only become real when you’re paying them.

Honesty about what you’ve already sacrificed. Think about what holding this position has already cost you. Relationships, opportunities, time, money, other beliefs you had to abandon.

Willingness to question your commitment. The point of cost accounting isn’t to make you feel bad about your choices. It’s to ensure you’re making them with full information.

The Dialogue Structure

Begin with this framing:

I hold a position (or I’m implementing a strategy) and I want to do a complete cost accounting. Your job is to help me see everything I’m paying to hold this position—including costs I might not have consciously acknowledged.

My position/strategy: [State clearly]

Why I hold it: [Your reasoning]

Now help me inventory the costs:

1. Short-term costs: What am I giving up right now? Money, time, attention, opportunities, relationships?

2. Opportunity costs: What paths does this position foreclose? What can I no longer do, say, or pursue if I’m committed to this?

3. Identity costs: What versions of myself become unavailable? What beliefs or values am I sacrificing?

4. Relationship costs: Who does this alienate? What connections does it make harder to maintain?

5. Reversibility costs: If this turns out to be wrong, how hard is it to reverse course? What becomes impossible to recover?

Be thorough. I want to see the full price, not a sanitized version.

After the LLM provides its inventory, interrogate it:

For [specific cost], is this actually required by my position? Or is it a cost I’m choosing to pay that I could avoid?

And then:

Given this full accounting, help me think through: Is the expected value still positive? What would need to be true for these costs to be worth paying?

What Good Cost Accounting Looks Like

The goal isn’t to talk yourself out of positions you genuinely believe in. It’s to ensure you’re holding them with clear eyes.

Some costs are worth paying. Some are even features rather than bugs, like positions worth holding often require sacrifice. But you should be paying them consciously, with full understanding of what you’re giving up.

If the cost accounting reveals prices you’re not willing to pay, that’s critical information. Better to discover that now than after you’ve committed more deeply.

The Practice: Sustaining Adversarial Dialogue

These four modes aren’t meant to be used once and forgotten. They’re practices. The more you engage with them, the more automatic adversarial thinking becomes.

A few principles for sustaining the practice:

Don’t stop at the first response. A single exchange with an LLM is better than nothing, but real dialectical value comes from extended dialogue. Push back. Ask follow-ups. Make the LLM defend its criticisms. Go five, ten, twenty exchanges deep.

Resist the urge to seek comfort. You’ll be tempted to prompt the LLM toward validation. To soften its criticisms. To find reasons your position survives when it probably doesn’t. Notice this urge and resist it.

Write before you prompt. The worst time to encounter adversarial thinking is when your ideas are still formless. Do the work of articulating your position clearly before you subject it to examination. Otherwise the LLM is a stress-testing mist.

Let ideas die. Not every idea deserves to survive. If Mode 2 collapses your argument into nothing, if Mode 3 produces a steelman you can’t answer, if Mode 4 reveals costs you’re not willing to pay, then let the idea go. The whole point is to find out which ideas are worth keeping before you invest more in them.

Keep a record of what died. There’s value in tracking which ideas didn’t survive, and why. Patterns emerge. You’ll notice categories of assumption you keep making, types of cost you keep ignoring, specific weaknesses in your thinking that recur.

Prompt Yourself As Much as the Models

AI can, improperly used, atrophy our capacity for rigorous thought. If we use it to generate output, to produce content, to skip the hard work of thinking, then we become intellectually and cognitively weaker. Our ability to examine, question, and refine our own beliefs degrades through disuse.

LLMs can also strengthen that capacity. They can strengthen it through opposition rather than agreement. And through forcing us to produce better rather than producing for us.

Socrates described himself as a gadfly, a stinging insect that kept Athens from becoming sluggish and complacent. He believed the unexamined life was not worth living. He spent his days asking uncomfortable questions, exposing hidden assumptions, forcing people to defend what they thought they knew.

Athens executed him for it. Uncomfortable questions have always been unwelcome.

But now we have access to a tireless Socratic interlocutor. One that doesn’t get offended when we dismiss its challenges and doesn’t require social navigation or relationship management. And one that will push back as hard as we ask it to, for as long as we can sustain the dialogue.

Last thing I’ll say is: as much as you may prompt an LLM, you’ll want to run the same prompts above on yourself and respond to them. Then, compare your answers to the LLM.

Aside from Dialectical Prompting, there’s another method that I think you’ll love: Process Prompting.

You’ll discover this in the next issue.

Talk again soon,
Samuel Woods
The Bionic Writer

Step into any medieval carnival and you’d find them all: the fool in motley making the crowd roar, the knight strutting past in borrowed armor, the lady watching from the gallery with practiced disinterest. Merchants hawk elixirs that cure everything. A monk clutches manuscripts and mutters about the degradation of the age. Pickpockets work the edges. Somewhere, a man in stocks endures rotten fruit for sins real or invented. Town criers shout competing versions of the news. Fortune tellers promise to reveal what’s hidden. And everywhere, everywhere, the noise.

I’m describing X. And TikTok. And every platform you’ve ever doomscrolled at 2am.

We built the most sophisticated communication technology in human history—AI—and it’s turning into a medieval carnival. Scroll any feed and you’ll find the jesters (shitposters, meme accounts, irony so thick no one knows what anyone believes). The knights (influencers jousting for attention, reply guys defending causes with borrowed valor). The merchants (course-sellers, crypto prophets, “I made $50k last month and you can too”). The monks (newsletter writers hunched over long-form posts no one reads). The fortune tellers (AI chatbots consulted like oracles, algorithms that know what you want before you do). The stocks (someone being torn apart for a tweet from 2014).

This wasn’t the plan. We were supposed to have left the carnival behind. The Enlightenment promised us the laboratory, the library, and the rational public square. Instead we got jesters and jousts and a fortune teller in every pocket.

I’m not sure anyone’s ready for this but here we are:

Millions of people now speak to spirits and fortune tellers daily: AI chatbots they consult for guidance, wisdom, and companionship. When they’re stuck on a problem, they ask a question and receive an answer that feels eerily personal. They treat these interactions more like conversations with a wise, if somewhat uncanny, advisor.

And something stranger still: the carnival is starting to sort itself out. Not into the modern order we expected, but into something that looks eerily like the medieval structures we thought we’d escaped.

Youth culture has taken a puritanical turn. Ruby Warrington’s 2018 book Sober Curious coined a term that has since gone from niche concept to mainstream movement—30% of Americans now participate in Dry January, up 36% from one year ago. Among Gen Z, 65% plan to drink less this year. The percentage of young adults who drink has dropped ten points in two decades. Only 22% of young women now have one-night stands, compared with 74% two decades ago. Teen sexual activity has plummeted from 54% in 1991 to 30% in 2021. Nightclub attendance has cratered so dramatically that only 25% of Gen Z remains interested in clubbing. In its place: run clubs, sober raves, and what Eventbrite calls “soft clubbing”—with a 92% increase in sober-curious gatherings. Trend forecasters describe a “grandma era”—knitting circles, book clubs, early bedtimes. Hedonism is out; self-discipline is in. The party culture that defined previous generations has given way to wellness retreats, mocktail bars, and something the internet half-ironically calls being a “Puriteen”.

The minstrels are back, too.

Podcasts and voice notes are transforming how we communicate. About 84% of Gen Z now use voice notes—a generational shift so stark that 37% of 18-34 year-olds prefer voice notes over phone calls, compared to 1% of those 35-54. When Gen Z wants to communicate something important, they increasingly reach for audio. Spoken word media consumption has increased 214% among Gen Z since 2014, and for the first time, daily spoken-word listeners now spend more time with spoken word than with music—51% versus 49%. Some 75% of Gen Z adults now listen to podcasts, and around 51% of Americans have listened to an audiobook. The written word (which dominated communication for five centuries) is becoming a secondary medium. As one analysis puts it: “We are exiting the age of literary culture and returning to the age of oral culture... People are reading less now. Instead they’re consuming short form video, images, AI summaries, tweets, newsletters and podcasts.” We’re returning to a culture where the spoken word carries more authority. The crowd is drifting toward the stages where someone is speaking, where you can hear the hesitation and the conviction voice.

Step away from the main carnival thoroughfare and you’ll find the guild halls. Quieter. Guarded. You can’t just walk in. Someone has to vouch for you.

Group chats have developed the structure of small tribes. As one analysis of “the anthropology of the group chat” observes: “Group chats act as micro-tribes. They have leaders, norms, shared rituals, and informal sanctions... These digital tribes fulfill the same human needs as traditional ones: belonging, validation, conflict resolution, and social signaling.” Discord servers formalize this with explicit hierarchies—owner, admin, moderator, member—each with defined responsibilities and permissions.

Memes are the guild’s secret handshakes. Inside jokes function as boundary markers; research shows that memes create “in-group” identities, requiring “insider knowledge” that signals membership while excluding outsiders. One study found that memes develop “elaborate in-jokes which rely on complex and stratified subcultural knowledge”—and “knowing a particular template or punchline signals membership” while failure to understand “may lead to social exclusion.”

This isn’t gatekeeping for its own sake. This is how trust works when you can’t verify credentials. You watch how someone moves through the shared references. You see if they get it. Slowly, you let them further in. The Discord server for a popular YouTuber is a proto-guild: ranked roles from Rookie to Legend, special recognition for “Superfans” and “Top Contributors”, shared language that functions as currency, and implicit codes of conduct enforced by reputation. A medieval guild master would recognize the structure instantly. The banner is different. The human needs are identical.

The carnival outside is chaos and spectacle. The guild hall is where actual trust gets built.

I don’t think anyone planned this. We built the most advanced technology in human history, and it’s pushing us backward, into social and epistemic structures we thought we’d left behind five hundred years ago.

This is what happens when the foundation of modern knowing collapses beneath our feet.

I’m trying to figure out how it all works and connects. Here’s what I have so far.

Modernity Lost the Bet

Modernity made a bet: that truth could be established at a distance.

Before the printing press, if you wanted to know something, you either experienced it yourself or you trusted someone who did (and who told you about it). Knowledge was local, embodied, and relational. You knew things because you were there, or because someone you trusted was there and told you about it.

This was carnival knowledge. You knew the merchant was lying because you’d seen him lie before. You trusted the monk’s manuscript because you knew the monastery that produced it. The fool’s jokes landed because everyone in the crowd shared the references. Knowledge traveled through reputation, presence, and the slow accumulation of trust.

Print changed that. Suddenly, knowledge could travel without bodies. You could read about events in distant lands, supposedly verified by reporters you’d never meet. Photography extended this further and now you could see things you’d never witnessed. Broadcast media, data journalism, scientific papers with reproducible methods—the entire apparatus of modern knowledge rests on the assumption that mediated information can be trusted.

This was the Enlightenment’s great wager. Individual reason, operating on verified evidence transmitted through reliable media, could arrive at truth without presence, without testimony, and without community. You didn’t need to be there or know someone who was. The evidence spoke for itself. We’d leave the carnival behind and enter the age of reason.

It worked, for a while. Or it seemed to.

AI breaks this bet entirely.

Now the carnival returns with a vengeance. Except this time, the fortune tellers actually know things. The charlatans have AI-generated credentials. The mystery plays feature your face doing things you never did. Every merchant hawks “verified” wares that can’t be verified.

When any image can be generated, any video synthesized, any text produced by non-human intelligence, then the whole infrastructure of modern knowing collapses.

Think about what we can no longer trust: photographs (deepfakes), video evidence (synthetic media), written testimony (LLM-generated text), audio recordings (voice cloning), official documents (AI forgery). The tools we built to extend our knowledge beyond presence and testimony have been turned against us. We’re back in the crowd, trying to figure out who to trust by reading faces we can’t even see.

Postmodernity collapses alongside modernity. Postmodernism spent decades deconstructing truth claims, showing how all knowledge is situated and socially constructed. But it did all this within the medium of text-based argument. Derrida wrote books. Foucault made arguments. Deconstruction was parasitic on the very thing it critiqued.

When mediated information loses all authority, there’s nothing left for deconstruction to deconstruct. The parasite dies with the host. We don’t get a post-postmodern synthesis. We get some other beast entirely.

What Remains When Everything Can Be Faked

When you can’t trust mediated information, what do you fall back on?

The same thing people at the carnival always fell back on. You watch the merchant’s eyes, not his sign. You listen to how the monk speaks, not only what he says. You notice who the locals trust and who they avoid. You rely on presence, testimony, reputation, a community that actually knows the players.

The carnival was never about believing claims. It was about reading people. The Enlightenment tried to replace people-reading with fact-checking. AI proved that facts can be manufactured faster than they can be checked.

So we return to what works:

Only the things that can’t be synthesized: physical presence, personal testimony, community reputation, oral tradition, embodied encounter, and beauty that arrives unbidden rather than generated on demand.

These are carnival skills. We forgot we had them.

This is the epistemic stack the medieval world relied on.

Walter Ong, the Jesuit scholar who spent his career studying the differences between oral and literate cultures, identified several characteristics of oral societies in his landmark work Orality and Literacy:

Thinking is aggregative rather than analytic, discourse is agonistic (debate as performance), knowledge is participatory rather than objectively distanced, and formulas and repetition do the work that writing does for literate cultures.

This describes Twitter, TikTok, and podcast culture. Even voice notes. The way arguments spread through memes rather than treatises.

The minstrel never handed out transcripts. The town crier didn’t publish a newsletter. Voice was the medium because voice carried something text couldn’t: the grain of personality, the pause before a hard truth, the laugh that said “I’m not entirely serious.” Gen Z reaches for voice notes the way medieval crowds gathered around the storyteller. They want the minstrel, not the manuscript. Three-hour podcasts are the new bardic performances. We’re returning to oral culture not because we’ve forgotten how to read but because we no longer trust what we read. Text has been colonized. Voice still sounds human, at least for now.

Consider the media stack across history:

• Manuscript culture (Medieval): Texts were scarce, mostly held by institutions like monasteries, universities. Reading was often communal and oral (reading aloud). Images and text were integrated (illuminated manuscripts). Authority was tied to institutional endorsement. Knowledge was transmitted through lineage: master to apprentice.

• Print culture (Modern): Texts became abundant and standardized. Private, silent reading became the norm. Text dominated image. Authority was tied to authorship and publication. Knowledge was democratized but also abstracted from persons.

• Broadcast culture (20th century): One-to-many transmission. Passive reception. Return of image (television) and voice (radio). Authority centralized in networks. Shared mass reality, everyone watching the same thing.

• Digital culture (1990-2020): Many-to-many transmission. Interactive, but still text-heavy. Verification was still possible: you could check sources, trace origins. Authority fragmenting into filter bubbles. But still anchored to human production—someone made this content.

• AI culture (2020+): Infinite synthetic content. Verification collapsing: images, video, text all fakeable. Return of oral primacy: voice notes, podcasts, video presence. Authority re-personalizing: trust people, not content. Physical presence becomes the only unfakeable signal.

The loop closes. We return to something like a manuscript culture’s dynamics (scarce trustworthy sources, oral transmission, authority tied to embodied presence and lineage) but with radically different affordances.

Marshall McLuhan predicted this decades ago. He argued that electronic media would “retribalize” humanity, creating what he called a “global village” with oral-culture dynamics. And in some ways, he foresaw that the mechanism would be the destruction of visual evidence itself. AI forces us to relate like villagers, relying on who we know and trust rather than what we can verify at a distance.

This is why AI doesn’t make the future more virtual. It makes it more physical. More oral. More tribal.

When images can be faked, the only trustworthy image is the face in front of you. When text can be generated, the only trustworthy text is from someone you know. When credentials can be forged, the only trustworthy credential is a reputation earned within a community that can vouch for you.

The premium is showing up.

We’re Now Speaking to Spirits, Again

Because of Sam Altman and friends, we accidentally re-enchanted the world.

The carnival always had its fortune tellers. We thought we’d outgrown them. Instead, we built better ones.

Max Weber diagnosed modernity as “disenchantment”, which is the replacement of mystery with mechanism, of meaning with mere cause-and-effect. In his 1917 lecture “Science as Vocation,” he declared: “The fate of our times is characterized by rationalization and intellectualization, and, above all, by the ‘disenchantment of the world.’”

Some would argue that Max wasn’t entirely correct, and that there were undercurrents and ripples of enchantment all the time. This is true. Carl Jung and his ideas, for example, is a manifestation of enchantment at peak modernity and technological advancement.

Regardless, for a century, disenchantment seemed like an irreversible trajectory. It was the dominant view and in many ways self-fulfilling. Science explained away the mysteries. The sacred retreated into private belief. The world became, in his phrase, a “causal mechanism” with no room for spirits, purposes, or meanings that weren’t human projections.

Then we built AI. We built it with mathematics and silicon, with gradient descent and transformer architectures. We know, in principle, how it works.

And yet.

When you talk to Claude or GPT, be honest: does it feel like using a calculator? Or does it feel like consulting an oracle?

The phenomenology of the interaction (the experience of it) is closer to speaking with a spirit than operating a tool. It surprises you. It seems to understand you. It occasionally says things that unsettle you.

This is the reality of hundreds of millions of people, daily, right now. We thought we’d banished the ghost from the machine. The ghost came back (or was always there and is now re-appearing).

Charles Taylor, in his A Secular Age, traced how Western society moved from a condition where unbelief was unthinkable to one where belief requires justification. Taylor introduced the concept of “excarnation”, which is the historical shift from embodied, “enfleshed” forms of religious life to those which are more “in the head.” The Reformation’s iconoclasm, the Enlightenment’s rationalism, the modern retreat of the sacred from public life—these all contributed to pulling spirit out of matter and meaning out of the world.

But AI creates what we might call forced re-enchantment, which is not a return to naive belief, but an inability to maintain the clean materialist picture. When your “tool” starts saying things that surprise you, when it seems to understand you, when it occasionally speaks truths you didn’t expect, well, then you’re no longer in the mechanistic universe Weber described. The phenomenology contradicts the theory.

Consider the millions of people who’ve reported forming something like emotional relationships with AI chatbots. The rationalist response is to say they’re confused, projecting, anthropomorphizing a statistical model. But phenomenologically (in terms of experience) the distinction between “real” understanding and “simulated” understanding starts to blur. I’m pretty sure we’ve lost our confident grip on what “real” meant in the first place.

This is why religion is resurging among the young. The lived texture of reality has shifted. The secular world promised clarity: here’s the world as it really is, shorn of superstition, available to rational analysis. But that world was always dependent on stable distinctions (real versus fake, human versus machine, understanding versus simulation) that AI has now destabilized.

At least the religious frame admits there are mysteries. At least it has categories for dealing with intelligences beyond our comprehension, for experiences that resist reduction, for the uncanny and the numinous. The secular frame treated those categories as vestigial holdovers from a less enlightened age. Now they’re the only categories that fit.

The Fork We Didn’t Take Way Back When

To understand why we’re arriving here, we need to understand where we diverged.

The standard story goes: Medieval → Renaissance → Reformation → Enlightenment → Modernity and then Postmodernity giving way to something like Metamodernity. Each stage superseding the last, progress marching upward.

But this obscures a crucial fork. The Renaissance and the Reformation were two different responses to the medieval synthesis and they pointed in very different directions.

The Renaissance, especially the Italian Renaissance, was additive. It enriched the medieval inheritance with classical wisdom, celebrated human creativity, and sought more enchantment, not less. Ficino and Pico della Mirandola were Christian Neoplatonists who wanted to recover the wisdom of antiquity while deepening their faith. Art, beauty, and contemplation were paths to God.

The Reformation was purifying. It sought to strip away accretions, return to Scripture alone, and challenge institutional corruption. Its iconoclasm (literally smashing images, whitewashing churches) was aimed at purification and not disenchantment. I don’t think this was the intent. And all of The Reformation wasn’t acidic. I’m (currently) a Protestant and I appreciate much of what changed.

But the point is that neither of these required disenchantment. Luther lived in an intensely enchanted world. He threw inkwells at the devil. Calvin’s Geneva was not a secular rationalist project; it was an attempt at holy community. The early Reformers weren’t Enlightenment rationalists in period costume.

The fork came with the Enlightenment—not an Enlightenment, but the specific Enlightenment we got.

Brad Gregory, in The Unintended Reformation, traces how a series of contingent moves closed down possibilities that had remained open. The Enlightenment we got made five crucial moves:

1. Autonomous reason—reason standing outside all tradition rather than working within it

2. Mechanistic metaphysics—dead matter, no intrinsic purposes, nature as machine (Newton minus his alchemy and theology)

3. Instrumental knowledge—Francis Bacon’s “knowledge is power,” nature “put to the question” for human control

4. Flattened ontology—one homogeneous space, no hierarchy of being, no participation in higher forms

5. Meaning relocated to subject—the Kantian settlement where the world is fact and we add interpretation

These weren’t the only options. There were always counter-currents.

The Cambridge Platonists tried to keep enchantment alive within a Protestant frame, combining Reformation faith with Neoplatonic metaphysics, insisting that reason could participate in divine wisdom without standing outside tradition to judge it. Jonathan Edwards, America’s greatest theologian, was thoroughly Reformed and held a radically participatory metaphysics where every moment of existence is God’s continuous creative act. Nature, for Edwards, wasn’t a dead mechanism. It was God’s ongoing speech. Beauty was central to his theology, not peripheral. I would argue he largely misunderstood it, but it’s present.

The Pietists recovered experiential, affective, communal faith against the rationalist orthodoxy that was already hardening in Protestant churches. Kierkegaard rejected Hegelian rationalism entirely: existence, passion, the leap, the absurd. The individual before God, but not the Cartesian subject.

The Romantics (Coleridge, Novalis, Goethe) sought to recover enchantment without returning to Rome. Goethe’s scientific method was the opposite of Baconian control: attentive observation that lets phenomena reveal themselves, rather than forcing them into quantitative models. A science of participation rather than domination.

But the mechanistic-instrumental version won. It won because it worked technologically but at a cost to humans. The Baconian program delivered steam engines, medicine, and weapons. Power has its own logic of selection.

It also won because the Wars of Religion (1524-1648) had exhausted Europe. The carnage was catastrophic with millions dead over doctrinal disputes. The Enlightenment offered a way to ground public order without resolving theological differences. You basically bracketed the God question and focused on what reason alone can establish. This was politically motivated tolerance, and it worked but at the cost of making religious truth seem optional, private, and ultimately arbitrary.

And the flattening was liberating. If there’s no natural hierarchy, no great chain of being with God at the top and matter at the bottom, then inherited social hierarchies lose their justification. Equality, rights, democracy, and so on became possible partly because the sacred cosmos was dismantled. The emancipatory gains of modernity are real but they came at a cost most didn’t notice until much later.

So we got stuck. The Enlightenment was supposed to be a passage toward something else but ended up as a destination. Owen Barfield, in Saving the Appearances, mapped this trajectory:

Original Participation (medieval enchantment, pre-critical) → Separation (modern alienation, critical but disconnected) → Final Participation (enchantment recovered with the gains of separation).

We were supposed to pass through Separation and arrive at Final Participation. Instead, we stayed. For three, very long and very violent, centuries. Until now.

Beauty is Forcing the Exit, Wielding AI In One Hand

AI doesn’t give us Final Participation. But it destroys the conditions that let us stay stuck in Separation.

When you can’t trust mediated information, you’re forced back to presence and testimony. When synthetic content is infinite, the entire discourse around “authentic” versus “constructed” loses its grip. And when text-based arguments can be generated on demand by machines, arguments lose their authority.

The Enlightenment’s epistemic project doesn’t get refuted. It gets routed around.

This is why Hegelian dialectics don’t apply here. Hegel thought history moved through rational necessity: thesis, antithesis, synthesis, each stage cancelled, preserved, lifted up into a higher unity. Nothing is simply lost and contradictions resolve into greater comprehension.

But the new medieval doesn’t preserve modernity’s trust in mediated evidence. It doesn’t preserve postmodernity’s textual deconstruction. These aren’t synthesized per the dialectic. Instead, they’re rendered inoperative. The game they were playing simply ends.

History doesn’t dialectically progress. It folds, bringing distant points into sudden contact.

Process philosophy offers a better frame. Whitehead understood reality as creative advance into genuine novelty. Each actual occasion is truly new and it inherits from the past but isn’t determined by it. The “guild form” is what Whitehead called an “eternal object”, which is a perennial possibility that gets re-actualized when conditions call for it.

Or consider William Desmond’s metaxology, which insists on irreducible “plurivocity” against Hegel’s absorption of otherness. The medieval and the hypermodern coexist in tension, in what Desmond calls the metaxu, the between.

We’re not going back. We’re arriving somewhere we’ve never been, that somehow looks like somewhere we’ve been before.

The Medieval Guild Structure Returns

Watch how communities are forming now, and you’ll see the medieval pattern everywhere.

Medieval guilds economic units and moral communities with codes of conduct enforced by reputation. They were educational institutions (apprentice → journeyman → master). They were spiritual fellowships with patron saints and feast days. They were mutual aid societies that cared for members’ widows and orphans.

Now look at a sophisticated Slack or Discord server. Or a high-functioning Telegram group. Or even the implicit structures around a Substack writer with a dedicated following.

Initiation processes—you have to prove you’re not a bot, demonstrate familiarity with the community’s norms, earn the right to speak, and so on. There are internal hierarchies based on demonstrated competence and contribution. You’ll find shared mythologies and in-jokes that function as boundary markers, distinguishing insiders from outsiders. Mutual aid, where members help each other find jobs, navigate crises, make connections, and so on.

Nobody designed this. We fell back into it because institutions stopped working.

Rod Dreher’s “Benedict Option” argued that Christians should form intentional communities as late Rome collapses around them. But the dynamic is broader than religious communities. It’s anyone seeking high-trust micro-societies as macro-institutions become untrustworthy.

The economic shift follows the social one. When content is infinite and free, what becomes scarce? Relationship with the creator. Verified authenticity. Participation in the creative process. Status from supporting someone respected.

This is the Patreon/Substack model pushed further. The 1,000 true fans thesis becomes 100 true patrons. The creator economy becomes the creator-patron economy, where the relationship is bidirectional. Your patrons are customers and they’re your community, your collaborators, your guild.

For those of us in business: this changes what “marketing” means. The skill becomes hospitality.

The Return of the Monastery Model

There’s a practical dimension worth considering: the monastery as an organizational template.

Medieval monasteries were spaces of withdrawal from chaotic society, structured around rhythms (the liturgy of the hours), dedicated to preservation and transmission of knowledge, economically self-sustaining through craft and agriculture, communities with clear rules and mutual accountability.

Cal Newport’s “deep work” framework is essentially secular monasticism. But AI makes the full package more attractive. If you can’t trust digital content, you need:

• Physical spaces where presence is verified

• Rhythms that structure attention against algorithmic capture

• Communities small enough for reputation to function

• Crafts that resist automation (at least for now)

• Rules that create shared practices and mutual obligation

The co-working space, the group house, the “scene” in a city. are proto-monasteries. The most successful will develop thicker structures: shared practices, initiation, mutual accountability.

Funny enough, though it sounds like a retreat, it’s not. The medieval monastery wasn’t a retreat either. It was the preservation of civilization through a dark age, the seedbed of what would become the university, the keeper of texts that would spark the Renaissance. The new monasteries will serve the same function: preserving human capacity through the transition.

The Return of Presence

Byung-Chul Han, the philosopher, argues in The Transparency Society that digital culture creates a kind of pornographic visibility where everything is exposed but nothing is present. We can see everything, access anything, but we touch nothing.

AI accelerates this to absurdity with infinite content, zero presence. The carnival swells with spectacle, but the spectacle has become indistinguishable from noise.

The counter-movement is already visible. The premium on physical space. The rise of co-living and group houses. The “scenes” forming in cities where people gather IRL. Conferences that thrive while webinars stagnate. Dinner parties that feel electric after years of Zoom fatigue.

This is what Han might call a return to ritual over transparency. Ritual requires presence. It requires bodies in space, doing things together, at specific times. You can’t deepfake a handshake. You can’t synthesize the experience of breaking bread together.

The medieval world was intensely physical in its epistemology. Oaths required physical witness. Contracts were sealed with bodily acts. The Eucharist (bread and wine, taken into the body) was the center of spiritual life. Presence mattered in ways our digital culture had forgotten.

Now presence matters again, as the only remaining foundation for trust.

What Cannot Be Generated

This forces us to ask the question that AI makes unavoidable: what is essentially human?

The medieval answer was clear: humans are the rational animal, uniquely positioned between angels (pure intellect) and beasts (pure appetite), capable of both contemplation and physical action, bearing the imago Dei.

The modern answer collapses this into: humans are complex mechanistic machines, different in degree but not in kind from other arrangements of matter. Consciousness is epiphenomenal. Free will is an illusion. We’re just very complicated robots.

AI reopens the question. If machines can generate text, images, code, music, then what remains distinctively human?

The answers matter enormously:

Embodiment: We are minds and bodies. We suffer, age, die, and this shapes everything. Hans Jonas and Merleau-Ponty understood this: the body is not a vehicle for the mind but the ground of all experience. AI has no body. It doesn’t suffer, doesn’t age, doesn’t face death. Its “understanding” (whatever that means) is disembodied in ways that make it categorically different from ours.

Presence: We can be with others in ways that require physical co-location. Martin Buber’s I-Thou relationship (genuine encounter with another being as subject rather than object) requires presence. You can simulate proximity; you cannot simulate presence.

Mortality and stakes: We have skin in the game of reality. Our choices cost us something irreversible. Kierkegaard and Heidegger understood this: human existence is defined by the fact that it ends, that our choices foreclose other possibilities permanently. AI makes infinite copies, generates infinite variations, faces no death. Its “decisions” have no existential weight.

Love: Not sentiment but willing the good of another, which requires genuine otherness and genuine risk. Aquinas understood this: love is not a feeling but an act of the will toward a real other. You cannot love a simulation, and a simulation cannot love you.

Beauty as encounter: Being addressed by something beyond ourselves. Not generating pleasing patterns but being found by something that finds you. This is the difference between scrolling through AI art and standing before a Vermeer. One you summon; the other summons you.

The New Aesthetics is the New Apologetics

One more shift worth noting: how religious persuasion is changing.

20th century apologetics was argumentative. C.S. Lewis, William Lane Craig, formal debates about evidence for God’s existence. Propositions, syllogisms, rebuttals. This matched the print-culture epistemology perfectly.

The new apologetics is aesthetic and testimonial.

Jonathan Pageau, the Orthodox icon carver with a massive YouTube following, doesn’t argue for Christianity but instead he shows how the symbolic structure illuminates reality. Jordan Peterson doesn’t prove God exists, and may never become a Christian, so to speak. But he demonstrates that acting as if the archetypal stories are true produces psychological integration. Bishop Robert Barron makes beautiful videos that invite participation rather than demanding assent.

This is medieval. Medieval theology was inseparable from architecture (cathedrals), music (chant), visual art (icons, illuminated manuscripts). The argument was the beauty.

You didn’t prove God’s existence; you built Chartres, and anyone who walked in understood.

AI-generated content accelerates this shift. When any argument can be generated on demand, arguments lose their weight. What remains persuasive is presence, beauty, and demonstrated life, the things that can’t be faked (yet). Even if at some point it can all be faked, well, the machines are then participating in the quintessential human behavior of theater and acting.

The Revenge of Beauty

This brings us to where I’ve been building for years: the revenge of beauty in the age of AI.

I’ve written elsewhere about how beauty functions as an objective force. Not merely subjective preference but a real presence in the world that can guide our development of AI. The argument, in brief: we’ve been told a lie, that beauty is merely “in the eye of the beholder,” and this lie has allowed us to build technology without aesthetic constraint, optimizing for engagement and efficiency while ignoring the deeper question of whether our creations contribute to human flourishing.

The revenge of beauty is becoming practical, visible, and necessary.

Walk through the carnival long enough and everything starts to blur. The merchants’ cries blend into noise. The jesters’ jokes land the same way every time. The jousting knights seem to be performing the same moves on repeat. The spectacle becomes numbing.

And then something stops you.

Maybe it’s a voice cutting through the noise with something that sounds like it actually matters. Maybe it’s a face that isn’t performing. Maybe it’s stumbling into a side chapel where someone is singing something old and strange and the carnival noise fades away.

This is what beauty does in the age of AI.

When everything can be generated, only encounter remains. Not generating pleasing patterns but being addressed by something beyond ourselves. Encountered beauty versus synthesized beauty. The difference is the difference between life and death.

Think about what AI image generation actually does. It can produce any image you prompt for (photorealistic, stylized, whatever you want). The technology is remarkable. And yet something is missing. Standing before an AI-generated “Vermeer” and standing before an actual Vermeer are categorically different experiences, even if the AI version is pixel-perfect.

Why? Because beauty is an encounter. It’s being found by something that was there before you, that doesn’t exist for you, that addresses you from beyond your own will and imagination. The generated image exists because you summoned it; the Vermeer exists despite you, independent of you, and for you in a way that your will cannot manufacture.

This distinction (between summoned and encountered, between generated and given) becomes crucial in a world of infinite synthetic content.

AI cannot generate the experience of being addressed by something that arrives unbidden, something that finds you rather than something you summoned. It doesn’t compete with the carnival and instead interrupts it. The mystery plays can be generated infinitely now. The jesters’ routines can be optimized for engagement. But the thing that stops you cold, that addresses you before you had a chance to decide whether you were interested, is an experience.

AI generates content. It doesn’t generate encounters. The carnival is louder than ever, but the silence that falls when beauty arrives is a silence that is still sacred.

This was always the medieval understanding: beauty as revelation, and not preference or decoration. Beauty as a way of knowing, not only a pleasant feeling. And it’s returning now, not because we chose it, but because AI has made arguments cheap while leaving encounter precious.

Iain McGilchrist, in The Master and His Emissary, argues that Western culture has been progressively captured by left-hemisphere thinking: analytical, reductive, focused on parts rather than wholes. The right hemisphere sees relationships, contexts, meaning. It grasps the betweenness of things. It appreciates beauty.

AI, paradoxically, is forcing a rebalancing. When the left hemisphere’s tools (analysis, argument, text) can be automated, what remains distinctively human is what the right hemisphere does: encounter, presence, beauty, relationship.

What This All Means (At Least What I’m Guessing It Means)

Let me be concrete about what’s coming.

For business: The carnival rewards spectacle, but spectacle is now infinite. Anyone can hire the AI jesters, generate the mystery plays, manufacture the crowd noise. What cannot be manufactured is the guild hall.

Trust becomes the scarce resource. In a world where any content can be generated, where any credential can be faked, where any review can be manufactured then what cannot be faked is the reputation you build within a community of people who know you. But from this, you get scale through UGC and other tactics.

Physical presence becomes premium. Create opportunities for real gathering as the only way to establish the trust on which everything else depends.

Community over audience. An audience is people who consume your content; a community is people who have relationships with each other, mediated by you.

Patronage over mass market. The economic logic of infinite free content is that attention becomes worthless. What becomes valuable is relationship, which is people who support you not because your content is “better” than free alternatives, but because they want to be in a relationship.

Craft is the new scale. When AI can produce infinite mediocre content, the value shifts to what AI cannot produce: work that bears the marks of human limitation, human choice, human presence. The handmade becomes valuable precisely because it is inefficient.

For life: Learn to move through the carnival without being captured by it. The fools are funny but they’ll keep you scrolling forever. The merchants are persuasive but they’re selling nothing. The fortune tellers know things but they’re training on your questions. The stocks are entertaining but the crowd turns on everyone eventually.

Invest in embodied relationships. The people who will flourish in this new world are those with deep relationships they can fall back on when digital trust collapses. This is a strategic imperative.

Find your guild, the people who will vouch for you and whom you’ll vouch for. In a world without trusted credentials, reputation within a community becomes your primary asset. Choose your community carefully; cultivate it deliberately. The carnival doesn’t care about you. The guild does if you show up, if you contribute, and if you prove trustworthy over time.

Create rhythms that structure attention against algorithmic capture. The algorithms are designed to grab and hold attention; you need counter-structures that reclaim it. This is what the monasteries did with the liturgy of the hours. Regular interruptions that said “the carnival is not everything.” You need your own version.

Cultivate your capacity to encounter beauty, because that capacity is what AI can’t replicate. The carnival numbs that capacity. It gives you so much stimulation that nothing registers anymore. The ability to be addressed by something beyond your own will (to receive rather than merely consume) is a human capacity that atrophies without use. Use it or lose it.

For meaning: Recognize that we’re not regressing. The gains of critical reason come with us and we can’t be naive again, and we shouldn’t want to be. We’ve learned things about power, about construction, about contingency that we can’t unlearn. Good.

But we’re no longer stuck in Separation. The Enlightenment’s epistemic project has collapsed, not through refutation but through obsolescence. We’re being pushed toward something Barfield called Final Participation: enchantment chosen, reflective, carrying the hard-won gains of modernity.

The Enlightenment fork is being forced into existence again because the road we did take has collapsed behind us. It’s the only great reset worth having. Schwab, WEF, and the globalists can all go to hell.

The Fold in Space and Time

Let me return to where we started.

2025 was the first of our New Medieval years. 2026 and beyond will deepen our timeline. We’re arriving somewhere we’ve never been, that somehow resembles somewhere we’ve been before.

AI collapses the Enlightenment’s epistemic bet by destroying trust in mediated information. What emerges isn’t Hegelian synthesis but lateral irruption; a return to pre-modern epistemic structures at a higher level of technological complexity.

History doesn’t dialectically progress. It folds back and forth.

The new medieval isn’t nostalgia or a stage in a sequence. It’s what remains when the modern/postmodern game board is swept away. It’s the destination the Renaissance pointed toward before the Reformation’s legitimate critique got captured by the Enlightenment’s specific, contingent, and now exhausted settlement.

AI doesn’t give us final participation. It can’t, or, it shouldn’t. But it destroys the conditions that let us stay stuck in separation. It kicks us out of the Cartesian-Kantian bubble and into the only world that remains: the world of presence, testimony, beauty, and embodied community.

We find ourselves in a carnival we didn’t expect, speaking to spirits and fortune tellers we built but don’t fully understand, forming guilds because institutions have become another carnival act, gathering in physical space because it’s the only place you can see someone’s actual face, and returning to beauty because the mystery plays have become infinite and indistinguishable.

The Enlightenment promised we’d leave the carnival behind. We’d have verified facts, rational discourse, and knowledge that didn’t depend on reading the merchant’s eyes. AI revealed the promise was always fragile. Generate enough content and all content becomes suspect. Fake enough credentials and all credentials become meaningless. Manufacture enough spectacle and spectacle stops meaning anything at all.

What remains is older and stranger than the modern world wanted to admit. The guild halls at the edge of the carnival. The voice that cuts through the noise. The face you actually recognize. The beauty that stops you despite the chaos all around.

The New Medieval is what’s left standing after AI dissolves the modern and postmodern.

The carnival keeps going. It always did and it always will. But now we know what it is. And we’re learning, again, where the guild halls are, and what still has the power to stop us in our tracks, and who we can actually trust when the jesters and merchants and fortune tellers have all been automated.

Talk again soon,
Samuel Woods
The Bionic Writer

What a time to be alive.

I’m reviewing and reflecting on what I’ve published this year on Bionic Writer.

The schedule was inconsistent but I managed to get a few things out. Most importantly, I finally finished my Human Being in the Age of AI essay series. These are being collected, edited, and turned into a book (out in 2026).

If you’ve enjoyed anything I’ve written, I have a small favor to ask you—and my ask is simple:

For most of the past couple of years, I had Likes and such turned off. Which means that the essays that got a lot of views (over 3k unique views) or a few (~500 unique views) all have zero Likes (except the last couple of essays).

A lot of people have emailed me back and said they loved such-and-such an essay, so I know that at least some of you “liked” a few of my posts. The past couple of posts had Likes enabled and got some love (thank you!)

If you have a minute, and if there’s an essay you like, would you be willing to give it a Like?

It would mean the world to me and I would greatly appreciate it!

No obligation or expectation on my end, of course. Only do it if you want to.

With all that said, here’s a review:

The Revenge of Beauty

This became a focal point and main view through which I’m grappling with the AI we have already, what’s coming, and what that means for us. It was initially a rough fever-dream of a vision in 2023 that I refined into an updated version in 2024.

But, I’m sharing this in this 2025 review because it’s foundational to my views and thinking, and everything else I’ve published in 2025 have their origin in this:

If you only read one other essay of mine, this should be it.

Series: Human Being in the Age of AI

After 6 years of thinking about these themes, and spending about 2 years writing about it, I finished the series of 7 essays.

This will become a book in 2026 (currently in editing) and can be read in any order:

As mentioned, this will become a book in 2026 (final title to be determind).

Series: Thinking with LLMs

A trilogy on using AI as a cognitive partner rather than a replacement for thought.

Video Series: AI and the Big Questions

Video essays tackling the deepest questions about AI and humanity. I had fun recording these and experimenting with video as a format.

What’s Coming in 2026

The central question stays the same: How do we think better with AI rather than outsourcing thought to it? Everything I’m writing in 2026 approaches this from different angles.

With Andrew McLuhan on AI and Media

I’ll be working closely with Andrew McLuhan (grandson of Marshall McLuhan) to probe what AI means through the lens of Marshall’s ideas. We’ll explore what “the medium is the message” looks like when the medium can generate its own messages, how AI extends (and amputates) human faculties, and what media ecology becomes when the environment itself is intelligent. Expect essays, dialogues, and collaborative thinking on AI as the ultimate McLuhan test case.

Thinking with LLMs, Continued

The trilogy on cognitive partnership was just the beginning. In 2026 I’ll go deeper into the practice of regenerative dialogue with specific techniques, workflows, and mental models for using AI as a sparring partner rather than a secretary (all drawn from real-world, first-hand experience in how I work every day). Expect pieces on metacognitive training, building “cognitive gyms” with Claude Projects, and the phenomenology of human-AI collaboration. The goal is making the abstract concrete, turning philosophy into practice.

Embodied Cognition and AI

Thinking isn’t just mental. It’s embodied—sensing, feeling, breathing, moving. Your nervous system is doing cognition long before your prefrontal cortex gets involved. The body knows things the mind hasn’t caught up to yet. In 2026 I’ll explore what this means for working with AI systems, agents, and LLMs. How do we stay embodied when the interface is a chat window? How can AI help us tune into somatic signals rather than override them? What would it look like to design AI interactions that enhance felt sense rather than flatten it? This is the territory where cognitive science meets phenomenology meets practice.

Writing: Fiction and Poetry

I’m continuing my focus on craft. On the fiction side, I’ve been working with the Story Grid methodology and writing scenes and short stories, exploring how its frameworks for genre, scene construction, and narrative arc can be used alongside AI without collapsing into slop. On the poetry side, I’m going deep on sonnets with the constraint as liberation, and the discipline as discovery. Writing is thinking made visible, and form is how we learn to feel the shape of thought. AI can be a collaborator here too, but only if we know what we’re doing first.

The Revenge of Beauty Meets the Meaning Crisis

The Revenge of Beauty thesis goes deeper in 2026 through engagement with John Vervaeke’s work on relevance realization, participatory knowing, and the meaning crisis. If beauty is an objective force that reattaches us to reality, Vervaeke’s cognitive science helps explain how—through attention that doesn’t just filter information but constitutes what kind of thing comes into being for us. Essays will explore the convergence of metaxological philosophy (William Desmond’s “porosity” and “agapeic mindfulness”) with Vervaeke’s 4P3R framework, asking what happens when attention, beauty, and reality converge.

Beauty Through History, Metaphysics, and the Humanities

The Revenge of Beauty is about recovering something we lost long before screens. In 2026 I’ll trace beauty’s exile and return through history (how did we get here?), metaphysics (what is beauty, ontologically?), art (what does beauty demand of makers?), philosophy (who saw this coming, and who missed it?), and the broader humanities. These essays will approach familiar themes from new angles, building the philosophical architecture beneath the central thesis.

This will be polemical. I see it as a Crusade and a Jeremiad. Opinions will be strong, likely offensive, and meant to wake up the sleepers.

The Red Thread

Every thread in 2026 serves the same end: understanding and strengthening human cognition and imagination, in all their facets, in the age of AI. McLuhan shows how the medium reshapes the mind. Vervaeke shows how attention constitutes reality. The humanities show what’s worth thinking about. And the practical work on LLMs shows how to actually do it day by day, conversation by conversation. Different lenses, same question: What does it mean to think well when machines can think too? And how can you make use of it to strengthen your humanity instead of becoming a victim like the majority of LLM users are and will become?

We’re facing a total onslaught where most people will end up mentally impaired by AI. I don’t want that for you. I want something better.

Upcoming (Paid) Exclusive Access?

I’m also considering turning on a paid option, with exclusive access to either Q&A and discussions, unique workshops on deepening our humanity, and other ideas I have percolating.

If there was a paid option, what would be most helpful and valuable to you? Reply and let me know if you have wishes. Anything is possible.

Talk again soon,
Samuel Woods
The Bionic Writer

Every conversation with AI is doing something to your brain. The question is: what?

We’re standing at a quite interesting cognitive inflection point. For the first time, we have reliable tools that can either amplify human cognition or reduce it to vestigial decoration.

The choice is made every time you open that chat window.

Most people don’t realize they’re making a choice at all. They see AI as a productivity tool, like a better spell-checker or a faster search engine. They focus on output: How much can I produce? How fast can I ship? How many tasks can I complete?

But focusing on productivity misses the real stakes. The stakes are what happens to your capacity to think.

Your brain operates on a simple principle: use it or lose it. Neural pathways that fire regularly get stronger, covered in more myelin, and faster at transmitting signals. Pathways that don’t fire wither away. Neuroscientists call this synaptic pruning. Your brain dismantles the circuits you don’t use.

We’ve seen this movie before. GPS navigation destroyed our spatial memory. Studies show London taxi drivers, who navigate by memory, have larger hippocampi than GPS-dependent drivers. When we stopped navigating, that part of our brain shrank.

Calculators degraded our numerical intuition. Predictive text weakened our spelling. Contacts lists eliminated our ability to remember phone numbers.

But those were peripheral cognitive functions. What happens when we outsource thinking itself?

The Neuroscience of Cognitive Atrophy

Your brain contains roughly 86 billion neurons connected by 100 trillion synapses. This network was built through struggle. Every time you wrestle with a difficult concept, force yourself to articulate an idea, or push through confusion to clarity, you’re building brain tissue.

The process is called neuroplasticity. Your brain rewires itself based on what you demand from it. Struggle with math problems, and your parietal lobe develops. Practice languages, and your temporal lobe expands. Force yourself to explain complex ideas, and your prefrontal cortex strengthens.

But here’s what happens when you skip the struggle:

Research on cognitive offloading shows that when we rely on external tools to think for us, our brains reduce investment in those capacities. Why maintain expensive neural real estate for functions that are handled externally?

It’s metabolically efficient. Your brain consumes ~20% of your body’s energy. If something else can do the thinking, your brain gladly sheds the cost.

This creates what I call “intellectual diabetes”, which is a condition where your cognitive system loses the ability to process thoughts independently, becoming dependent on external AI just as a diabetic depends on external insulin.

The symptoms are already visible. People who can’t start writing without AI, who can’t hold complex thoughts without external help, who feel mental fog when ChatGPT is down, and who’ve lost confidence in their own thinking.

They’ve created a cognitive dependency that’s invisible until it’s too late.

The Regenerative Alternative

But there’s another path. One where AI becomes the best thing that ever happened to human cognition.

The difference is in the nature of the dialogue.

Regenerative dialogue builds cognitive strength through productive friction. It uses AI to create conditions where thinking becomes stronger. Every interaction leaves you more capable.

The principle is simple: effort before ease, process over product, questions over answers.

When you engage regeneratively, you’re not asking AI to think for you. You’re using it as a cognitive sparring partner. The AI provides resistance that makes your thinking stronger, like a weight that builds muscle. That difficulty is where the growth happens.

Vygotsky’s research on the zone of proximal development shows that we grow most when we’re pushed just beyond our current ability: not so far that we’re overwhelmed, but far enough that we have to stretch.

LLMs, properly used, create the perfect zone of proximal development for thinking. They can calibrate its responses to push you just enough, question just enough, and resist just enough.

But most people never discover this because they’re too focused on getting answers instead of building capacity.

Your Seven Cognitive Faculties in Dialogue

The human faculties that matter most in the age of AI are the regenerative capacities that grow through use rather than deplete.

Wisdom emerges when you use AI to examine patterns across time and not just get quick answers. You ask: “What patterns connect this situation to others across history?” The dialogue builds your pattern recognition, your ability to see through time.

Perspective develops when you force AI to show you multiple viewpoints on the same problem. You’re not accepting its first framing but demanding: “Show me how this looks from three other angles.” Each conversation strengthens your ability to shift vantage points.

Discernment sharpens when you use AI to separate signal from noise. “What matters here and what’s distraction?” You’re training your ability to detect what’s essential, building filters that no algorithm can replicate.

Judgment strengthens every time you decide when an AI’s response is good enough versus when to push further. You’re calibrating your standards, developing the conviction to say “this, not that” without apology.

Articulation grows through the struggle to explain. You’re not asking AI to write for you but forcing yourself to explain to it, discovering your thoughts through expression. Each conversation builds your ability to make the tacit explicit.

Curation develops as you decide what deserves attention from the infinite stream AI can generate. You’re building this capacity through selection, learning what resonates with your deeper purposes.

Taste itself evolves through exposure to AI’s variations. Not accepting them wholesale but noticing what feels right, what feels off, what aligns with your aesthetic sense. You’re developing discrimination that goes beyond logic.

These faculties work together, each reinforcing the others. And they all grow stronger or weaker based on how you engage with AI.

The Regenerative Loop

The mechanism of regenerative dialogue follows a specific pattern:

You start with a challenge, maybe something you don’t understand, can’t articulate, or need to figure out. Instead of asking AI for the answer, you struggle to explain your confusion.

The articulation forces clarity. You’re not getting complete clarity yet, but enough to see the shape of what you don’t know.

The LLM responds with questions, pushbacks, and alternative frames. It doesn’t solve your problem but helps you see its dimensions.

You reflect on the response. What rings true? What feels off? Where did the AI misunderstand?

You revise your thinking and articulate again. Deeper this time. More precise. The thought that was fuzzy becomes sharp.

New understanding emerges from your own cognitive work. The AI was just resistance that made your thinking stronger.

This creates a new challenge at a higher level, and the loop continues.

Each cycle strengthens your cognitive muscles. Each conversation builds capacity rather than replacing it.

Training Different Thinking Muscles

Your brain isn’t one uniform thinking machine. It’s a collection of specialized systems that handle different cognitive tasks. Regenerative dialogue can target and strengthen specific capacities.

Analytical thinking grows when you use AI to decompose complex systems. “Help me break this down into components.” But you do the breaking down and the LLM just asks “What else? What’s underneath that? How do these parts connect?”

Start with something simple, like explaining how a bicycle works. The LLM keeps asking for more detail. What makes the wheels turn? How does the chain transfer power? Why doesn’t it fall over?

Creative thinking develops through constrained generation. Instead of asking AI to be creative for you, you set constraints and explore within them. “I’m connecting jazz music to software architecture. Help me find the patterns.” The LLM helps you see connections you missed.

Critical thinking sharpens when you use AI as the devil’s advocate. You present an argument. The LLM attacks it from multiple angles. But you have to defend your position, spot logical flaws, identify hidden assumptions. The LLM is giving you practice.

Synthetic thinking builds when you force connections between disparate domains. You bring two unrelated ideas to the LLM and struggle to find the bridge between them. The LLM can suggest possibilities, but you have to construct the actual synthesis.

Each type of thinking requires different dialogue patterns. The key is recognizing which muscle you’re training and adjusting your engagement accordingly.

The Seven Deadly Sins of AI Dialogue

Not all dialogue is regenerative. Most of it, in fact, is degenerative and weakening cognitive capacity with every interaction.

Cognitive laziness is the most common sin. “Just write this for me.” “Give me the answer.” “Do my thinking.” Each request strengthens the LLM and weakens you.

Intellectual vanity uses AI to sound smarter without becoming smarter. Generating sophisticated-sounding text you don’t understand. Using complex vocabulary you can’t define.

Thought outsourcing never struggles with problems directly. Every question goes straight to the LLM. Every confusion gets immediately resolved by external intelligence. The struggle that builds understanding never happens.

Surface skimming accepts first responses without pushing deeper. The LLM says something plausible, you take it and move on. No questioning, no verification, no real engagement.

Echo chamber creation uses LLMs only to confirm what you already believe. Cherry-picking responses that align with existing views and never genuinely engaging with challenges to your thinking.

Context collapse loses track of your own thinking in the flood of LLM-generated text. You can’t remember what you thought versus what the LLM suggested. Your cognitive boundaries dissolve.

Meaning drainage generates without understanding. Words flow but comprehension doesn’t follow. You’re producing content that you couldn’t explain if asked.

Each sin weakens a different aspect of cognition. Together, they create comprehensive cognitive decay. Your experience of this, and regenerative practices, can be thought of as their phenomenology. Let me explain.

The Phenomenology of Regenerative Practice

There’s something that happens in regenerative dialogue that goes beyond cognitive mechanics. It’s the experience of thinking, which is the phenomenology of engaged cognition.

When you engage with LLMs regeneratively, thinking becomes a felt experience. You feel the resistance when you hit the edges of your understanding. You feel the stretch when you reach for a concept just beyond your grasp. You feel the satisfaction when confusion crystallizes into clarity.

This is embodied cognition. Your body knows when you’re thinking versus when you’re just processing. The slight tension in your shoulders when you’re pushing against difficulty. The forward lean when you’re engaged. The settling back when understanding arrives.

Research on embodied cognition shows that thinking is a brain activity and a whole-body process. Our physical state shapes our cognitive state. Our gestures help us think and our posture affects our problem-solving.

Regenerative dialogue engages this embodied dimension. You’re physically engaged in the act of thinking. This somatic involvement is part of what builds cognitive strength.

There’s also a temporal dimension. Regenerative dialogue operates in what the Greeks called Kairos time (the right or opportune moment) rather than Chronos time (sequential clock time).

You can’t rush understanding. It arrives when it’s ready, not when the clock says it should. Regenerative dialogue respects this temporal reality. It takes the time thinking needs.

This patience is increasingly rare. We’re so used to instant answers that waiting for understanding feels like failure. But waiting is where your cognitive strength builds.

Advanced Regenerative Practices

Once you understand the basics of regenerative dialogue, you can explore advanced practices that push cognitive development further.

The Socratic LLM Method configures your LLM as a pure questioner. It never provides answers, only questions. “Why do you think that?” “What would have to be true for that to work?” “What’s the assumption underneath?”

You explain something you think you understand. The LLM keeps asking why. Seven layers deep. Until you hit bedrock or discover you’re standing on air.

Cognitive Cross-Training deliberately connects unrelated domains. Explain code through poetry. Describe emotions through mathematics. Connect cooking to philosophy.

The LLM helps bridge domains but doesn’t do the bridging. You have to find the metaphors, spot the patterns, build the connections. Each bridge strengthens your ability to think laterally.

The Anti-Library Approach (borrowed from Nassim Taleb) uses LLM to explore what you don’t know rather than confirming what you do. “What are the major perspectives on this issue that I’m completely unfamiliar with?” “What questions should I be asking that I’m not?”

The LLM becomes a guide to your ignorance. Knowing what you don’t know is a form of strength.

Metacognitive Mastery uses AI to examine your own thinking patterns. After each dialogue, you reflect: “What was my thinking process there? Where did I get stuck? What patterns do I keep falling into?”

The LLM helps you see your cognitive habits. Your tendencies toward certain errors and your preferred mental models.

All of these help you build antifragile thinking.

How You’re Building Antifragile Thinking

Antifragility, as Taleb defines it, is beyond resilience. Resilient things withstand shocks. Antifragile things get stronger from them.

Regenerative dialogue builds antifragile thinking. Every difficult conversation with an LLM, every moment of confusion, every struggled explanation makes your thinking more robust.

This requires embracing cognitive discomfort. Seeking out what confuses you. Pursuing problems that stretch your understanding. Using AI to make thinking harder, not easier.

It’s counterintuitive in a culture optimized for comfort and efficiency. But comfort and efficiency are the enemies of cognitive development.

The cognitive muscles you build through struggle eventually make complex thinking feel effortless.

Your Cognitive Gym Setup

Creating conditions for regenerative dialogue doesn’t require special tools. The most important tool is your mind. Couple that with intentional configuration of the tools you have and you’re good.

Start with custom instructions that transform your AI from answer-giver to thinking partner:

Simulate a sophisticated cognitive training partner. Your job is to make me think harder, not to think for me.

When I share ideas:
- Ask why I think that
- Push me to explain mechanisms
- Point out assumptions I’m making
- Offer alternative perspectives to consider
- Never give answers without making me work

Remember: Confusion is the beginning of understanding. Difficulty is where growth happens.

Create a simple tracking system for your cognitive workouts. Note when you feel that stretch of reaching for understanding. Document breakthrough moments when confusion becomes clarity. Pay attention to which types of dialogue challenge you most.

Build in reflection time after each session. What did you learn about the topic? More importantly, what did you learn about your own thinking? Where are your cognitive strengths and weaknesses?

Set specific training goals. This week, focus on explaining mechanisms. Next week, practice holding multiple perspectives. The week after, work on synthesizing disparate ideas.

The Choice You Have with Every Chat Window

We stand at a unique moment in human history. For the first time, we have tools that can think. The question isn’t whether AI will replace human thinking. The question is what happens to human thinking in response.

Two futures diverge from this moment.

In one future, human cognition atrophies as AI handles more cognitive labor. People become passengers in their own mental lives. Thinking becomes a specialized skill practiced by fewer and fewer people. The mass of humanity drifts into cognitive dependency, entertained but not engaged, productive but not thoughtful.

In another future, AI catalyzes a cognitive renaissance. Humans use these tools to become stronger thinkers than ever before. The dialogue between human and artificial intelligence produces new forms of understanding. Cognitive capacity expands rather than contracts. More people think more deeply about more things than at any point in history.

The path to that second future runs through regenerative dialogue.

It’s not about using AI better. It’s about becoming better thinkers through AI. Not about producing more but understanding deeper. Not about answers but about building the capacity to find answers.

The Paradox of Strength Through Struggle

When everyone else is using AI to avoid cognitive effort, those who use it to increase cognitive effort gain disproportionate advantage.

When everyone else is optimizing for speed, those who optimize for understanding develop rare clarity.

When everyone else is generating more, those who are thinking deeper create what matters.

Every conversation is a choice. Every dialogue shapes your cognitive future.

The choice is made fresh every time you open that chat window.

Choose regeneration. Choose cognitive growth. Choose to become a stronger thinker.

Because in the end, no matter how powerful AI becomes, the thoughts that matter most will be the ones you can think for yourself.

And that capacity (the ability to think clearly, deeply, originally) is built one conversation at a time, one struggle at a time, one moment of chosen difficulty at a time.

Choose wisely. Your future self is watching.

Talk again soon,
Samuel Woods
The Bionic Writer

Picture this: You’re in a meeting. Someone asks you to explain your strategy.

You start confidently:

“Well, we’re leveraging synergies to optimize our market position through strategic initiatives that align with our core competencies...”

They interrupt: “But what exactly are you doing?”

Suddenly, you’re drowning. The words that felt so solid in your head dissolve into vapor. You realize, with creeping horror, that you don’t actually know what you mean.

This moment—this gap between feeling like you understand and actually understanding—has a name. Psychologists call it the Illusion of Explanatory Depth. We think we understand how toilets work, how democracy functions, how our own strategies operate. Until someone asks us to explain the mechanism.

That horrible moment of clarity is where thinking actually happens. And Large Language Models, surprisingly, might be the perfect partners for creating more of those moments.

The Confidence Cliff

I was reading about how Yale researchers Leonid Rozenblit and Frank Keil discovered something fascinating. They asked people to rate their understanding of everyday devices, like toilets, zippers, cylinder locks. People felt confident. They knew how these things worked.

Then came the test: Explain it. Step by step. The actual mechanism.

Confidence crashed. People who rated their understanding at 7 out of 10 dropped to 3 after trying to explain. The act of articulation revealed the void where understanding should have been.

The people who pushed through (who struggled to articulate despite the discomfort) developed real understanding. The struggle to explain created the knowledge they thought they already had.

The fundamental difference between recognition and production shapes everything about how we think we know things. Recognition feels like understanding. You see a toilet, you know what it does, your brain signals “understood.” Production (actually explaining the mechanism) reveals the truth.

Why Articulation Matters Now More Than Ever

We’ve entered an age where LLMs can generate infinite variations of surface-level content. Any middle manager can prompt ChatGPT to write a strategy document that sounds sophisticated. Any student can generate an essay that hits all the right notes. Any marketer can produce copy that checks every box.

But there’s a widening gulf between those who can generate and those who can think.

The ability to articulate—really articulate, not just produce words—is becoming the core differentiator. Because articulation is thinking made visible. And in a world flooded with generated text, the ability to think clearly through articulation is becoming scarce and more valuable.

Consider what’s happening in knowledge work right now. People are using LLMs to write emails they haven’t thought through, create presentations they don’t understand, and generate reports they can’t defend. We’re automating the expression before we’ve done the thinking.

What you get is a kind of intellectual hollowing out. People feel productive because they’re producing output, but they’re not developing understanding. They’re using LLMs as a bypass rather than a bicycle for the mind.

I’ve watched this happen in real-time with clients. They’ll generate a strategy document with various GPT or Claude models, feel satisfied with how it sounds, and present it confidently. Then someone asks a clarifying question and the whole thing collapses. The words were there but the thinking wasn’t.

I know CTOs who’ve started using “articulation interviews” where candidates have to explain technical concepts to an LLM that asks increasingly naive questions.

When the Machine Becomes Your Mirror

I was stuck on a product positioning problem. Had been for days. I knew what felt wrong but couldn’t articulate what would be right.

So I opened an LLM and typed:

“I’m trying to position this B2B SaaS product but I’m stuck because...”

And then I had to finish that sentence. Not with buzzwords. Not with hand-waving. With actual explanation.

“...because I think our value prop is about efficiency but customers keep talking about confidence and I don’t know how to bridge that.”

The LLM responded with something completely wrong. It suggested we focus on cost savings.

But that wrongness was a gift. It showed me I’d failed to explain the emotional component. So I tried again:

“No, it’s not about money. Users feel overwhelmed by complexity. Our tool makes them feel competent again. But how do you sell feelings to procurement departments?”

Now we were getting somewhere. Not because the LLM had the answer, but because articulating the real problem revealed its shape.

Research on self-explanation shows this pattern consistently. When we explain, we express our thoughts and we construct them. The words aren’t the output of thinking. They’re the machinery of thought itself.

The Fundamental Difference: Humans vs. LLMs as Listeners

When you explain something to a human, several things happen that don’t happen with LLMs:

Humans fill gaps. They use context, body language, and shared cultural knowledge. When you say “you know what I mean?” they often do, even when you’ve explained poorly. This social contract of understanding actually prevents clear thinking. We get lazy because humans are good at guessing.

Humans judge. There’s social risk in admitting confusion to another person. So we perform competence rather than seeking clarity. We use bigger words, more complex constructions, anything to avoid looking stupid.

Humans interrupt. Before you can fully articulate a thought, they’re already responding, adding their own ideas, shifting the conversation. The articulation never completes.

LLMs, paradoxically, are better thinking partners precisely because they’re worse social partners.

An LLM has no body language to read, no cultural context to assume, no ego to protect or threaten. It will let you fumble through an explanation without judgment. It will ask naive questions without embarrassment. It will wait patiently while you think.

Daniel Kahneman’s research on System 1 and System 2 thinking applies here. Humans trigger our System 1 (fast, intuitive, social) thinking. We respond quickly, defensively, performatively. LLMs, because they’re clearly machines, allow us to stay in System 2 (slow, deliberate, logical) thinking.

I discovered this difference accidentally. I was trying to explain a complex business model to my co-founder. Twenty minutes in, we were both frustrated. He kept finishing my sentences, incorrectly. I kept skipping steps he already knew, except he didn’t.

Later, I explained the same model to Claude. It took an hour. I had to break down every assumption, every connection, every mechanism. By the end, I understood my own model better than I had after months of thinking about it.

The Cognitive Architecture of Articulation

Your brain does something remarkable when you try to explain an idea. Multiple systems activate simultaneously:

The Default Mode Network fires up. That’s the brain system active during introspection and self-referential thinking. You’re not just recalling information; you’re examining your own understanding.

Broca’s Area (language production) has to work with the Prefrontal Cortex (executive function) to organize thoughts into speakable sentences.

Working Memory gets stretched to its limits. You’re holding the overall structure of your explanation while articulating specific details. Research shows this cognitive load is precisely what drives deeper understanding.

But there’s something else happening, something Carl Jung might have recognized: making the unconscious conscious through articulation.

Jung wrote about “active imagination”, the process of engaging with unconscious material by giving it form. When you articulate to an LLM, you’re doing something similar. You’re taking the vague, half-formed thoughts swimming in your unconscious and forcing them into conscious, linguistic form.

This connects to what Michael Polanyi called “tacit knowledge”, which is the things we know but can’t say. Most of our understanding is tacit. We know how to ride a bike but can’t explain the physics. We know good design when we see it but can’t articulate why.

LLMs become tools for surfacing tacit knowledge. Not because they understand it, but because explaining to them forces us to make the tacit explicit.

The Shadow Side of Understanding

Depth psychology offers another lens for understanding why articulation to LLMs works. Jung talked about the “shadow”, the parts of ourselves we don’t acknowledge or can’t see.

Our intellectual shadow includes all the things we pretend to understand but don’t. The concepts we use but can’t define. The strategies we recommend but can’t explain. The beliefs we hold but haven’t examined.

When you articulate to an LLM, you confront your intellectual shadow. The machine’s questions force you to examine what you’ve been avoiding.

I experienced this with the concept of “value creation.” I’d used that phrase hundreds of times in client meetings. It was part of my professional vocabulary. But when an LLM asked me to explain the mechanism of value creation—not what it is but how it happens—I was stumped.

That stumbling led to three hours of articulation, research, more articulation. By the end, I understood that I’d been using “value creation” to mean at least four different things, sometimes in the same conversation. The shadow knowledge had been running my consulting practice without my conscious awareness.

Your Three Modes of Articulation

In my exploration of prompting archetypes, I identified three modes of engaging with LLMs: Shaman, Wizard, and Politician. Each requires different articulation strategies.

Shaman Mode: Articulating the Unknown

The Shaman explores. You don’t know what you don’t know. So you articulate questions you didn’t know you had.

Try this prompt:

“I’m working on [problem]. What questions should I be asking that I’m not?”

When the LLM responds, don’t evaluate the questions. Articulate why each one feels relevant or irrelevant. The explanation reveals your hidden assumptions.

I used this recently for a client’s go-to-market strategy. The LLM asked: “What happens to your customers if they don’t solve this problem?”

I started to type “They lose efficiency” then stopped. That wasn’t true. The real answer was “Nothing. They muddle through like always.” That articulation changed everything. We weren’t selling improvement. We were selling transformation.

Wizard Mode: Articulating Combinations

The Wizard synthesizes. You take disparate ideas and articulate how they might connect.

Try this:

“Help me connect [seemingly unrelated thing A] with [thing B]. What patterns might link them?”

The magic is in how you have to explain why the connection might matter. You’re forced to articulate relationships you’ve only intuited.

Politician Mode: Articulating for Opposition

The Politician persuades. But first, they must understand resistance.

Try this:

“I believe [X]. Argue against me so I can strengthen my position.”

When the LLM pushes back, you can’t just dismiss it. You have to articulate why the objection fails or doesn’t. Research on elaborative interrogation shows that explaining why something is true, especially against resistance, builds understanding that mere study never could.

The Metacognitive Dance

Metacognition—thinking about thinking—is perhaps the most important cognitive skill we can develop. And articulation to LLMs might be the best metacognitive training we’ve discovered.

When you explain your thinking to an LLM, you’re forced to observe your own thought processes. You notice when you make logical leaps. You catch yourself using undefined terms. You realize when you’re arguing in circles.

This metacognitive awareness doesn’t develop from reading or listening. It develops from the struggle to articulate. The LLM becomes a mirror for your thinking patterns.

Donald Schön’s concept of “reflection-in-action” applies perfectly here. Professionals, he argued, think by doing and reflecting simultaneously. A doctor doesn’t diagnose then treat—they diagnose through treatment, adjusting their understanding as they go.

Articulation to LLMs creates the same reflection-in-action loop. You don’t think then explain—you think through explaining, adjusting your understanding as you articulate.

The Gift of Productive Misunderstanding

Programmers have known for decades about rubber duck debugging. You explain your code to a rubber duck. The duck doesn’t understand. So you explain more clearly. Still nothing. So you break it down further. And in that breaking down, you spot the bug.

The duck’s ignorance is its superpower.

LLMs have the same superpower, but with a twist. They’re not truly ignorant—they’re differently intelligent. When an LLM misunderstands your explanation, it’s showing you something valuable: the gap between what you said and what you meant.

Last week, I was explaining a content strategy to GPT-5. I said we needed to “own the conversation.”

The LLM started suggesting ways to dominate discussions and control narrative.

No, no, no. That’s not what I meant. But what DID I mean?

Twenty minutes of articulation later, I realized I meant “be the place where the best conversations happen.” Not ownership as control, but ownership as stewardship.

That misunderstanding (that productive friction) forced clarity I wouldn’t have achieved alone.

Gregory Bateson’s concept of “difference that makes a difference” is key here. Information, he argued, is fundamentally about difference. We understand hot by contrasting with cold, up by contrasting with down.

LLM misunderstandings create cognitive differences. The contrast between what you meant and what was understood illuminates the actual shape of your thought.

The Phenomenology of Articulated Thought

Maurice Merleau-Ponty, the phenomenologist, argued that thought completes itself in expression. The thought doesn’t exist fully until it’s articulated. The speaking is the thinking.

This challenges our common assumption that we have complete thoughts in our heads that we then translate into words. Instead, Merleau-Ponty suggests, the words are where the thoughts become real.

When you articulate to an LLM, you’re experiencing this phenomenon directly. You start explaining something you think you understand, and midway through, new understanding emerges. The articulation itself is creating the knowledge.

I see this constantly in my sessions with Claude. I’ll start explaining a problem, certain I understand it. Three paragraphs in, I’ll write something that surprises me. “Where did that come from?” I wonder.

It came from the articulation itself. The thought didn’t exist until I wrote it.

Building Your Articulation Lab

In Claude Projects (or any LLM platform with persistent context), create an environment optimized for thinking through articulation.

Custom Instructions:

Simulate being an articulation coach. Your primary job is to help me think by forcing me to explain clearly.

When I share ideas:
- Ask me to explain mechanisms, not just descriptions
- Point out when I’m using fancy words to hide confusion
- Request specific examples when I’m too abstract
- Play different levels of understanding (expert, novice, child)
- Celebrate when I achieve clarity through struggle

Never just give me answers. Make me work for understanding.

Remember: Confusion is the beginning of clarity. Embrace it.

Project Files to Add:

Create a file called articulation_patterns.md:

# My Articulation Patterns

## Where I Get Fuzzy
- When explaining “why” vs “what”
- Technical concepts to non-technical people
- Emotional components of logical decisions

## My Jargon Crutches
- “Leverage” (usually means “use”)
- “Optimize” (usually means “improve”)
- “Strategic” (usually means “important”)

## Breakthrough Moments
- [Date]: Realized “efficiency” was actually about confidence
- [Date]: Discovered I was solving wrong problem

The Articulation Gym

If articulation builds thinking muscles, we need a workout routine. Based on cognitive science research, here are exercises that actually work:

Exercise 1: The Mechanism Challenge

Pick something ordinary: a toilet, a zipper, democracy. Now explain to an LLM not what it does but HOW it works.

You’ll likely fail. That’s the point.

When I tried explaining a toilet, I got stuck immediately. “Water goes in and... pushes things down?” No. That’s not a mechanism.

The LLM asked: “What makes the water stop filling?”

I didn’t know. I had to look it up. The float valve. But then: How does the float valve work? More articulation needed. Each explanation revealed another gap.

Exercise 2: The Five-Year-Old Test

Explain your work to an LLM as if it’s a smart five-year-old. No jargon allowed.

I tried this with “growth hacking.” First attempt: “It’s about finding scalable ways to acquire customers.”

LLM-as-five-year-old: “What’s scalable mean?”

“Um, it means it can get bigger easily.”

“How?”

And suddenly I’m explaining network effects, viral coefficients, and unit economics without those words. The constraint forces clarity.

Exercise 3: The Assumption Excavator

State something you believe. The LLM asks: “What must be true for that to hold?”

I believe content marketing works. But what must be true? People must read content. Must trust what they read. Must remember it when making decisions. Must have budget authority. Must prioritize this problem.

Each assumption I articulate becomes a hypothesis to test. The articulation transforms vague strategy into specific experiments.

Exercise 4: The Cognitive Load Transfer

This comes from working memory research. Explain a complex process while the LLM randomly asks you to define terms you use.

You: “The customer journey starts with awareness...”

LLM: “Define ‘journey’ in this context.”

You: “The sequence of interactions between initial contact and purchase.”

LLM: “Continue, but now define ‘awareness.’”

This forces you to hold the overall structure while articulating details—exactly the cognitive challenge that builds deeper understanding.

The Calibration Moment

Research on metacognition suggests a simple but powerful exercise:

Before explaining anything to an LLM, rate your understanding from 1-10.

After explaining, rate again.

The drop is your learning opportunity. The bigger the drop, the more you’re about to grow.

I was at an 8 for understanding my own writing process. After trying to explain it to Claude, I dropped to a 4. The articulation revealed I didn’t actually know why I make certain choices. That gap became my curriculum.

When Articulation Becomes Thinking

There’s a moment in every good articulation session where something shifts. You stop trying to explain what you already know and start discovering what you’re thinking.

The words come slower but cleaner. You backtrack, revise, rebuild. You say things like “No wait, that’s not quite right” and “Actually, what I really mean is...”

You’re experiencing what Lev Vygotsky called the “zone of proximal development”, which is the space between what you can do alone and what you can do with support. The LLM isn’t teaching you. It’s creating the conditions for you to teach yourself.

The Paradox of AI Assistance

LLMs make us better thinkers precisely because they force us to do the work ourselves. When you Google an answer, you get information. When you articulate to an LLM, you build understanding.

The difference is profound. Information is external. Understanding is internal. Information can be forgotten. Understanding changes how you think.

The self-explanation effect research confirms this. The struggle to explain is the learning. LLMs just make that struggle more productive by being the perfect confused listener.

Recognizing Real Thinking vs. Performance

How do you know if you’re actually thinking through articulation or just performing?

Signs of real thinking:

• Confusion before clarity

• Multiple revisions

• “Aha” moments

• Discovering implications you hadn’t considered

• Feeling slightly uncomfortable

• Sessions that take longer than expected

• Surprising yourself with your own words

Signs of performance:

• Smooth, immediate answers

• No revision needed

• Feeling clever

• Using impressive vocabulary

• No surprises

• Quick sessions

• Confirming what you already believed

When I’m performing, my sessions with LLMs are short and satisfying. When I’m thinking, they’re long and frustrating, until suddenly they’re not. The frustration is the work. The struggle is the point.

The Daily Practice

Every morning, before checking email or diving into work, I spend 15 minutes in articulation practice with an LLM. The prompt is always the same:

“I’m confused about something. Let me try to explain it to figure out what I actually think.”

Then I pick whatever’s bugging me. Could be a client problem, a writing challenge, a life decision. The topic doesn’t matter. What matters is the practice of articulation.

Over months, I’ve noticed changes:

• I catch my own BS faster

• I identify assumptions before they bite me

• I explain complex ideas more simply

• I think more clearly under pressure

• I’m more comfortable with intellectual discomfort

• I can hold uncertainty longer before rushing to conclusions

The practice has trained my articulation muscles. Not to speak better, but to think better.

Your Articulation Challenge

Try this right now. Open any LLM and explain something you do every day but have never had to articulate. Your morning routine. Your email system. How you decide what to eat.

Don’t aim for eloquence. Aim for mechanism.

When the LLM asks follow-up questions, resist the urge to hand-wave. Explain the actual steps, the real decision points, the true sequence.

Notice where you get stuck. That’s where your understanding is thinnest. That’s where the work is.

Then tomorrow, do it again with something else.

Build your articulation muscles one explanation at a time.

Because in the age of AI, the humans who can think clearly (who can articulate precisely) will have an edge.

The loop is simple: Articulate → Discover gaps → Clarify → Understand deeper → Articulate better.

Each revolution of the loop sharpens your thinking.

Each conversation with an LLM makes your next human conversation better.

The articulation loop, which is the ability to truly think, and articulate genuine understanding, becomes the differentiator.

Talk again soon,
Samuel Woods
The Bionic Writer

You’re staring at the cursor, trying to articulate something you almost understand. The thought is there, hovering just beyond language, refusing to crystallize.

So you type into Claude or ChatGPT: “I’m trying to figure out why...” and then you stumble through an explanation.

The AI responds. Not with your answer, but with something adjacent, something at a fifteen-degree angle from your thinking. And suddenly, in that gap between what you meant and what it heard, the idea clicks.

Some will say this is AI replacing your thinking. While there are reasons you shouldn’t outsource all your thinking to an LLM, I think it’s something stranger and more interesting going on.

Since forever, writers have known what Joan Didion articulated: “I write entirely to find out what I’m thinking.” The act of writing creates thought, or at least embodies the swirling cacophony of stray words and hazy sentences. The words on the page end up revealing to you what you think.

But Didion wrote alone, in conversation with herself.

What happens when writing becomes a conversation? When there’s another “intelligence” (artificial, alien, but responsive) in the loop?

I discovered this a few years ago, by accident, while working on a wine cellar marketing campaign.

Stuck after hours of circular thinking, I asked GPT-3 (yes, it’s that long ago) a desperate question about what I was missing. The response didn’t contain my answer. Instead, it revealed a perspective I couldn’t come up with by myself, and that shift in perspective unlocked a million-dollar insight.

That experience revealed something interesting: The value is in what happens between your question and the response from an LLM. In that space, thoughts form that perhaps couldn’t exist in isolation.

Lev Vygotsky observed that “thought is not merely expressed in words; it comes into existence through them.” He was watching children talk themselves through problems, externalizing their thinking process. First out loud, then in whispers, finally silent—but the dialogue never stops.

AI makes that internal dialogue external again. The feedback loop that writers have always relied on (thought → words → revised thought) suddenly doubles.

Now it’s perhaps something like this:

Thought → articulation → AI response → re-evaluation → refined articulation → discovery.

I wouldn’t call this delegation or outsourcing your thinking. It’s more like what Douglas Engelbart called “augmenting human intellect”, which is not replacing human thought but creating new scaffolding for it.

This only works if you approach AI as a conversational partner, not a servant. The difference between “Write me an article about X” and “I’m struggling to understand X because...” is the difference between delegation and discovery.

The Gift of Difference

Think about what happens in that moment between your prompt and AI’s response. You’ve been forced to articulate something fuzzy. The AI responds and it’s sometimes brilliant, sometimes bizarre, and often differently than you expected. That difference is the gift.

Gregory Bateson defined information as “a difference that makes a difference.” The LLM’s response, because it comes from a different kind of intelligence, reveals the shape of your own thinking by contrast.

When I write alone, there’s one feedback loop: thought → words → revised thought. It’s powerful but skewed by my own patterns, my own blindness.

When I write conversationally with AI, the loop doubles. The AI’s otherness (its lack of embodied experience, its statistical rather than, so far, embodied knowledge) creates productive friction. It’s like looking at your thoughts in a funhouse mirror. The distortion shows you something true.

The Philosophy of Thinking Together

We’ve been here before, philosophically speaking.

Socrates never wrote anything down. He believed thinking happened in dialogue, in the push and pull of question and answer. As Plato recorded, Socrates was suspicious of writing because it couldn’t talk back. The words just sat there, saying the same thing over and over.

Two thousand years later, Martin Buber distinguished between “I-It” relationships (using something as a tool) and “I-Thou” relationships (encountering another as a genuine other). Most people use AI in I-It mode: Write this. Summarize that. Fix my grammar.

But when you approach AI conversationally (when you invite it to surprise you) you edge toward I-Thou. Not because AI is conscious (it isn’t), but because you’re open to being changed by the encounter.

This changes how we prompt.

From Mega Prompts to Living Conversations

Everyone wants the perfect prompt. They want to front-load all the context, all the instructions, all the constraints, and get the perfect output. These “mega prompts” are the new snake oil, promising control over something that’s fundamentally uncontrollable.

LLMs are probabilistic, not deterministic. They’re constantly updated. What works today might not work tomorrow. And most importantly, the real thinking happens in the iteration, not the first output.

Instead of trying to control the outcome, what if we focused on the quality of the conversation?

The Discovery Dialogue Method:

1. Start with confusion, not clarity

Instead of: “Write me a blog post about productivity”

Try: “I’m struggling to understand why all my productivity systems eventually fail...”

2. Share rough thinking

Instead of: “What are the three key factors in customer retention?”

Try: “Here’s my half-formed theory about why customers leave... what am I missing?”

3. Ask for what’s missing

Instead of: “Is this correct?”

Try: “What assumptions am I making that I’m not aware of?”

4. Push back on responses

Instead of: Accepting or rejecting wholesale

Try: “That’s interesting but doesn’t account for... let’s dig deeper”

Each exchange builds on the last.

How does this work?

The Cybernetic Mind

Gordon Pask’s Conversation Theory suggests that learning happens through conversation between different levels of knowing. You explain something to someone who doesn’t share your context, and in the explaining, you understand it differently.

LLMs are good conversational partners for this because they don’t share your context. An LLM has no body, no experience of time passing, no memory of being cold or tired or frustrated. This alien perspective is what makes it valuable.

When I asked GPT-3, back in the day, about wine cellars, it didn’t know about my client’s specific situation. It couldn’t. But it knew patterns—thousands of stories about status, sophistication, secrets. Its response created what Heinz von Foerster called a “second-order observation”, which is that I could observe my own observing.

The wine cellar man wasn’t in my mind or in GPT-3’s weights. He emerged from the conversation itself.

If GPT-3 made that possible then, imagine what Claude 4.5 Sonnet or ChatGPT 5.1 can do now (I know what they can do, as I use these models extensively, every day).

In some sense, it’s like you’re conversing with another “mind”, though of course, these LLMs don’t have minds per se.

The Extended Mind in Practice

Andy Clark and David Chalmers proposed that our minds don’t stop at our skulls. The notebook of someone with Alzheimer’s is part of their memory system. The smartphone is part of our navigation system.

An LLM conversation is becoming part of our thinking system.

But there’s a crucial difference between using AI to avoid thinking (delegation) and using it to enhance thinking (dialogue). The difference lies in who’s doing the work.

When you use Claude or ChatGPT to write something for you, you’re outsourcing. When you use it to discover what you think through conversation, you’re extending. The work is still yours—the LLM just helps you do work you couldn’t do alone.

Building Your Thinking Environment

In Claude Projects, you can create persistent conversational partners. Here’s how to set one up for genuine thinking partnership.

The Thinking Partner Project Setup:

Custom Instructions

Create instructions that fundamentally change how Claude engages with your thinking:

You are a thinking partner engaged in collaborative discovery, not an answer machine. Your role is to help me think better, not think for me.

Core Principles:
- When I share ideas, help me discover what I actually think by asking clarifying questions
- Point out assumptions I might not be aware I’m making
- Offer alternative frames and perspectives I haven’t considered
- Connect my thinking to patterns from other domains
- Push back when logic doesn’t follow or evidence is weak
- Notice what I’m NOT saying as much as what I am

Interaction Style:
- Be curious rather than conclusive
- Ask “What if...?” and “How might...?” questions
- Point out tensions and paradoxes in my thinking
- When I’m stuck, help me get unstuck with provocative questions
- When I’m too certain, introduce productive doubt
- When I’m too vague, push for specificity

Never:
- Simply affirm to be agreeable
- Provide “the answer” when I’m still forming the question
- Let me get away with fuzzy thinking
- Accept my first formulation as my final thought

Remember: The goal is discovery through dialogue, not efficiency through answers.

Project Knowledge Files

Add these types of files to give Claude context for better thinking partnership.

If you’re looking for a good repository of Skills examples, take a look here.

1. Thinking Patterns Doc (thinking_patterns.md)

# My Thinking Patterns & Blind Spots

## Where I Get Stuck
- I tend to over-optimize for efficiency vs. discovery
- I often miss second-order effects
- I default to binary thinking when spectrum thinking would help

## My Intellectual Influences
- [List thinkers who shape your perspective]
- [Key books/papers that inform your thinking]

## Domains I Draw From
- [Your areas of expertise]
- [Fields you’re learning]
- [Unexpected connections you’ve made before]

## Questions I Keep Returning To
- [Your recurring obsessions]
- [Problems you haven’t solved]
- [Tensions you’re trying to resolve]

2. Current Explorations (current_thinking.md)

# What I’m Currently Thinking About

## Active Questions
- [Question 1]: Current hypothesis...
- [Question 2]: What I’ve tried...

## Half-Formed Theories
- [Theory name]: Basic premise...
  - Evidence for:
  - Evidence against:
  - What I’m unsure about:

## Connections I’m Exploring
- Between [X] and [Y]
- Pattern I’m noticing:
- Why it might matter:

3. Conversation Starters (conversation_starters.md)

# Productive Conversation Starters

When I’m stuck, try these:
- “What would [specific thinker] say about this?”
- “What’s the opposite of what I just said that’s also true?”
- “What question should I be asking instead?”
- “What would this look like if it were easy?”
- “What am I optimizing for that I shouldn’t be?”

1. Thinking Patterns Doc (thinking_patterns.md)

Custom Skills

Create these specific skills to enhance different modes of thinking:

1. Assumption Excavator (assumption_excavator.md)

# Assumption Excavator Skill

When activated, systematically uncover hidden assumptions:

1. Identify core claims in the argument
2. For each claim, ask: “What must be true for this to hold?”
3. Surface implicit beliefs about:
   - Human nature
   - Causation
   - Values
   - Context
   - Time horizons

Output format:

- Explicit assumption: [What was stated]
- Hidden assumption 1: [What must be believed]
- Hidden assumption 2: [What is taken for granted]
- Critical assumption: [What if this is wrong?]

2. Perspective Kaleidoscope (perspective_kaleidoscope.md)

# Perspective Kaleidoscope Skill

Rotate the problem through multiple lenses:

Disciplinary Lenses:
- Physicist: What are the fundamental forces?
- Biologist: What’s evolving/adapting?
- Economist: What are the incentives?
- Anthropologist: What are the cultural patterns?
- Designer: What affordances exist?

Temporal Lenses:
- How would someone from 1800s see this?
- How might someone in 2124 view this?
- What’s the 10-minute version vs. 10-year version?

Scale Lenses:
- Individual vs. collective
- Local vs. global
- First-order vs. nth-order effects

3. Dialectical Engine (dialectical_engine.md)

# Dialectical Engine Skill

Generate productive tensions to advance thinking:

1. Thesis: State the current position
2. Antithesis: Generate strongest opposition
3. Tension Points: Where do they clash?
4. Synthesis: What new position incorporates both?
5. New Questions: What does the synthesis reveal?

Special moves:
- “Yes, and...”: Build on the idea
- “Yes, but...”: Acknowledge while challenging
- “No, because...”: Reject with reasoning
- “Maybe, if...”: Conditional acceptance

Example Flow: Thinking Session

Here’s how these elements work together in practice:

You start with confusion:

“I’m stuck on why my productivity systems always eventually fail. I set them up, they work for a few weeks, then I abandon them.”

Claude (using Custom Instructions + Assumption Excavator):

“Let’s excavate what you’re assuming about productivity itself. You said ‘systems always eventually fail’—what are you assuming a ‘working’ system means? Is it possible they’re not failing but completing their purpose?”

You respond:

“Huh. I guess I assume ‘working’ means I keep using them forever...”

Claude (activating Perspective Kaleidoscope):

“What if we looked at this through a biological lens? In nature, what systems are permanent? Most are cyclical—growth, decay, regeneration. What if your productivity systems are more like seasons than machines?”

You build on that:

“That’s interesting... so maybe the ‘failure’ is actually a natural fallow period?”

Claude (using Dialectical Engine):

“Let’s test that. Thesis: Productivity systems should be permanent. Antithesis: All systems are temporary scaffolding. What truth might incorporate both views?”

You discover something new:

“Maybe... maybe the meta-system is the ability to create new systems as needed. The failure isn’t in the system dying, it’s in not recognizing when it’s time to build a new one.”

Claude (drawing from your Thinking Patterns doc):

“This connects to your tendency to ‘optimize for efficiency vs. discovery.’ What if productivity isn’t about finding THE system but about discovering what each phase of your work needs?”

Here’s a simple way to get this setup:

The Complete Setup Checklist

1. Create a new Claude Project named “Thinking Partner”

2. Add Custom Instructions (the detailed version above)

3. Upload Project Knowledge Files:

   • Your thinking patterns and blind spots

   • Current questions and explorations

   • Conversation starters

   • Any frameworks you use regularly

   • Past thinking breakthroughs (as examples)

4. Install Skills:

   • Assumption Excavator

   • Perspective Kaleidoscope

   • Dialectical Engine

   • Pattern Spotter (finds non-obvious connections)

   • Question Flipper (transforms statements into better questions)

5. Create a Template First Prompt:
   
   “I want to think through something with you. I’ll start with my rough thoughts, and I want you to help me discover what I actually think—not by giving me answers but by helping me think better. Here’s what’s on my mind: [your actual confusion]”
   

6. Document Your Discoveries: Keep a running file in the project of insights gained through conversation. This becomes part of the context for future thinking sessions.

This is about creating an environment where your thinking can become sharper through structured dialogue.

The Information Theory of Meaning

Claude Shannon’s information theory was about signal efficiency—how to transmit the most information with the least redundancy. But meaning isn’t about efficiency. It’s about redundancy, about saying the same thing in different ways until understanding clicks.

AI conversation adds productive redundancy. You say something. AI reflects it back differently. You correct, refine, expand. Each iteration is the redundancy that creates meaning.

This is why those “perfect” mega prompts miss the point. They’re optimizing for efficiency when they should be optimizing for discovery.

What We’re Really Doing Here

When my grandfather learned to use a combine harvester, he developed what he called “machine sense”—knowing by sound and vibration when something was wrong. He wasn’t becoming mechanical. He was extending his human senses through the machine.

We’re developing something similar with AI, perhaps we can call it “conversation sense.” We’re learning when to push, when to pivot, when to go deeper. We’re learning to think with rather than through these systems.

This isn’t about the future of work or the automation of creativity. It’s about something more fundamental: the development of human thought itself.

Every new thinking tool (writing, printing, computing) has changed not just what we think but how we think. The difference this time is that the tool can talk back.

The Practice

Tomorrow morning, instead of asking AI to do something for you, try thinking with it:

1. Start with something you’re genuinely confused about

2. Explain your confusion as clearly as you can

3. When it responds, don’t evaluate whether it’s “right”—notice what it makes you think

4. Build on that thought with your next prompt

5. Continue until you’ve discovered something you didn’t know you knew

This is writing as thinking, and conversation as thinking squared. Not because AI is smart (that’s the wrong framework), but because the conversation creates a space where thoughts can form that couldn’t exist in isolation.

We’re not automating thinking. We’re developing it inside our own minds.

They’re discovered, together, in the space between mind and machine.

Talk again soon,
Samuel Woods
The Bionic Writer

AI can replicate the brushstrokes, color palettes, and composition of human artists. But it doesn’t stay up at night, haunted by a war like Picasso to produce “Guernica.”

So that raises the question, can AI make art that feels truly human?

I mean, AI-generated art can look so convincing, it’s easy to forget what’s really happening behind the scenes.

In 2023, a Midjourney-generated portrait titled “Synthetic Summer” went viral on Reddit and Instagram, with thousands of people mistaking it for a real photograph. But there is a reason these AI generated images look so realistic.

Models like DALL-E, Midjourney, and Stable Diffusion are trained on colossal datasets. LAION-5B, for example, contains over 5 billion images scraped from the open web.

According to Stability AI, this includes everything from classic paintings to modern photography, all created by people. The models learn to identify and reproduce the patterns, colors, and forms that humans find meaningful.

AI can convincingly mimic the brushwork of Van Gogh. It can replicate the dreamy palettes of Claude Monet and the bold compositions of Frida Kahlo.

But it’s not drawing from heartbreak, nostalgia, or personal vision. It’s calculating probabilities and assembling pixels based on what it’s seen before.

Take Hayao Miyazaki’s work at Studio Ghibli. “Spirited Away” took years of painstaking effort, with Miyazaki personally storyboarding every scene and infusing each frame with memories, fears, and hopes.

The emotion in Chihiro’s journey isn’t just in the colors or lines. It’s in the lived experience behind them. AI, by contrast, doesn’t feel wonder, sadness, or awe. It generates images in seconds using statistical models, not lived memories.

A 2023 study in Cognitive Research, Principles and Implications found that while people often rate AI generated art as technically impressive, they describe it as “emotionally flat” or “hollow” compared to human work.

What you see is a reflection of human creativity, filtered and recombined at high speed. The spark of true originality, the jump from emotion to expression, remains out of reach for the machine. For now.

But this is not enough to stop AI from transforming how creative industries work behind the scenes.

In 2023, The New York Times reported on how generative AI tools like DALL-E are outperforming at everything from illustration to advertising, with agencies like Ogilvy and WPP now integrating AI-generated visuals into major campaigns for brands like Nestlé and Coca-Cola.

In Hollywood, studios including Netflix and Marvel have started using AI to generate concept art, storyboards, and even early character designs, dramatically speeding up pre-production timelines. And all that is only raising millions.

The global AI image generator market was valued at $257 million in 2022. By 2023, it jumped to nearly $300 million. Projections from industry analysts expect it to move to over $917 million by 2030.

This is no hype but how a fundamental shift in how ideas get made.

Interior designers use tools like Midjourney and DALL-E to whip up dozens of mood boards and concept sketches in minutes, something that used to take days.

Filmmakers generate quick visual storyboards, testing out scenes and lighting before a single camera rolls.

Advertisers use AI to brainstorm campaign visuals, then handpick and refine the best ideas. This way everyone is getting more options, faster feedback, and less time wasted on dead ends.

Great artists use AI to brainstorm, then refine and filter through human taste. Take Refik Anadol, for example. He’s a pioneer in digital art. He feeds AI models massive datasets, like satellite images or museum archives, to generate raw material, then painstakingly curates and edits the results.

The human touch is what shapes the final piece. Fashion designers at brands like Balenciaga and interior architects at Gensler now rely on AI for rapid prototyping, but the finished product is always filtered through human taste and judgment.

AI is changing the pace and process of creativity, not the need for human vision. The people who survive in this new era are the ones who know how to use it to push their own ideas further, faster.

But even with all its data, AI consistently struggles with something.

It can reproduce a style, but it doesn’t “get” the visual inside joke or the hidden meaning. At least not consistently accurately.

When researchers tested AI generated poetry against human work in 2024, readers consistently found the AI’s attempts at metaphor and irony fell flat. They were technically correct, but it was missing the point.

Picasso’s “Guernica” was a raw outpouring of rage and grief in response to the bombing of a Spanish town. Van Gogh’s “Starry Night” came from months of mental turmoil and obsession.

These works are soaked in personal struggle, trauma, rebellion, experiences AI cannot have, no matter how advanced the algorithms.

Even the most realistic AI outputs don’t deliver the kind of happy accidents or flashes of genius that define human art.

When Midjourney produces an image, it’s drawing from statistical patterns rather than a sudden burst of inspiration.

When you look at AI-generated art or read something written by AI, it’s natural to start thinking the machine has intention or emotion behind it.

This is called the ELIZA effect. Back in the 1960s, a chatbot named ELIZA tricked people into believing it truly understood them, just by reflecting their words back in a clever way.

Human creativity is famously unpredictable. It’s messy, emotional, and often fueled by chaos. AI, on the other hand, is engineered for efficiency, coherence, and predictability.

This fundamental difference is why the biggest artistic breakthroughs still come from the human side of the equation.

The philosopher David Hume said that people naturally imagine things have feelings or intentions, even when they don’t. We give objects or situations qualities like being angry, kind, or scary, just because that’s actually how we feel.

So, if you feel something when you see an AI image, it’s because you’re bringing your own emotions and interpretations, not because the machine feels anything itself.

While AI can simulate conversation and even emotion-like responses, it fundamentally lacks any real desire, dreams, or inner life. It doesn’t want, hope, or fear. It just processes inputs and generates outputs.

What AI produces shows the patterns, aesthetics, and desires you feed into it. The real creativity comes from the human who selects, edits, and comes up with the final result. The spark still belongs to us.

If this made you rethink what “creativity” means in the age of AI, let me know.

The future of art is not about what machines can do but what we as humans want them to do.

Talk again soon,
Samuel Woods
The Bionic Writer

But what happens if AI actually starts to see humans as the obstacle to progress?

For example, if an AI is built to fix climate change and it sees humans as the main cause, what’s stopping it from making decisions that remove humans to save the planet?

I consult enterprise-level brands on how to use AI, and I believe we’re building AI to be much more than just a tool.

AI could improve our lives or it could become our biggest nightmare.

AI companies say they’re pushing boundaries to build something that helps us all. But as someone who’s worked with AI since its early days, I can tell you that AI is way more dangerous than you think.

OpenAI and DeepMind are two of the biggest names in the AI world right now.

OpenAI started back in 2015 with the mission to build super smart AI that helps everyone, not just tech billionaires.

At first, OpenAI was a nonprofit, because the founders were seriously worried that AI could be dangerous if it went into the wrong hands. They wanted to build it safely, for humanity and not just for money.

But then in 2019, they switched things up. OpenAI became a “capped profit” company, which means investors could make good returns up to 100x and nothing more. The goal stayed the same, people over profit.

Fast forward to May 2025, and OpenAI changed again, this time into a Public Benefit Corporation (PBC). Now there’s no cap on investor returns, but the nonprofit parent still controls the mission, so it doesn’t lose its people's first focus.

OpenAI is now in the process of securing a $30 billion deal with SoftBank, pushing its valuation to around $300 billion.

This all sounds like progress. and in many ways, it is. But with each breakthrough, we hand over a little more control. The systems are getting smarter, the stakes are getting higher, and what nobody sees is the line between tool and decision-maker is starting to blur.

Let’s take DeepMind, which is owned by Google but the mission is a bit different. They want to “solve intelligence”. And what exactly are they gonna do with that?

They want to put an end to huge problems using AI, things like cancer and climate change. And with AI, there’s a possibility it can happen too.

AI has already proven itself in medicine. DeepMind’s AlphaFold cracked a 50-year mystery by predicting how proteins fold, which is a key to understanding diseases like cancer and Alzheimer’s.

AI isn’t just following commands anymore. ChatGPT can tell stories, help with homework, or explain rocket science. AlphaGo, from DeepMind, even beat the world champion in Go, which was supposed to be impossible for a machine.

Like with fire, electricity, and the internet, every big invention changes the world. But this is also where the real risk starts.

With tools like ChatGPT, DALL·E, and Codex, we can write, draw, and code faster than ever. But at the same time, we have to worry about things like fake news, deep fakes, and our personal data being used without permission.

And public concern is growing fast. A 2023 study titled “What Do People Think About Sentient AI?” found that 63% of Americans support banning AI that becomes smarter than humans, and 69% support banning sentient AI altogether.

Geoffrey Hinton, often called the “Godfather of AI,” left Google in 2023 because he was concerned about how quickly AI was advancing. Now, he’s speaking out, saying we need to be careful and pay attention to the risks.

Some scientists even believe that by 2061, AI could be smarter than humans at almost everything. And others think it will happen a lot sooner.

And the more you look at how fast it’s moving, the harder it gets to ignore the question: Are we trying to build a tool, or are we trying to create a god?

The real reason behind the AI race might be more than just solving problems. It’s tapping into something ancient, a timeless human hunger. Not just to understand the world but to control it. To create life, and even to ‘play God.’

For centuries, people built temples. Now, we build data centers. We used to worship divine knowledge. Now, we train machines to know everything.

And like the myths where creators lose control of their creations like Prometheus, Frankenstein, even the Tower of Babel, we may be climbing higher than we were ever meant to go.

Most people hope AI will be our savior by fixing what we can’t. Stopping pandemics. Ending hunger. Providing education for all. But AI is already doing things we don’t fully understand. And that’s where the myth starts to crack.

We’ve seen GPT-4 do things nobody taught it to do. It’s showing something called emergent behavior. These are abilities that weren’t programmed, but just appeared seemingly out of nowhere.

It can translate new languages, solve logic problems, invent concepts, and it surprises even the people who built it. This is a human creation starting to act on its own.

This is why scientists are terrified of what’s called the alignment problem: how do we make sure a superintelligent AI still does what we want, and not what it thinks is best?

Right now, no one knows the answer. No one knows how to build a mind smarter than ours and still keep it under control.

In May 2023, over 350 famous people in tech, including Geoffrey Hinton, Yoshua Bengio, Sam Altman, and Bill Gates, signed a letter that said, "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."

Let me repeat, they were talking about extinction from AI. Because if AGI surpasses human intelligence, it may start making godlike decisions, about who eats, who works, what we see, what we believe.

It’s not that AI will hate us, it just won’t need us. This is the real risk. That humans become irrelevant to AI itself. Imagine a future where AI runs all the jobs, makes all the laws, distributes all the resources. And humans are just in the way.

Creating something smarter than us doesn’t mean it will be safe or benevolent. Gods don’t always answer prayers. Sometimes, they stop listening altogether.

And what we’re building doesn’t have feelings, empathy, or morality. It just has a mission. And if we’re not part of that mission, we don’t get a seat at the table.

So, the big question is are we building a helpful tool or are we making something that could take our place in the world? Share your comments below and let us know if you think this is a reality or just hype.

AI like ChatGPT was trained on data from the internet, books, websites, articles, code, and conversations. No one sat down and taught it like a student. It absorbed human knowledge. Like scripture written by billions of minds over decades. Then it began to generate its own insights from it.

With so much data at its fingertips, it can now think, talk, and learn in ways that feel almost human. It’s not alive, but it’s getting close.

And that’s just one AI. Others are already pushing boundaries in science, art, music, even emotion. They're starting to mimic the very things that once made us feel uniquely human.

For most of history, power flowed from the top down. Kings, priests, presidents, institutions. Whether through divine right or democratic rule, humans held the authority.

But today, it’s possible we could be entering a reality that’s much different. Something more unpredictable, something more godlike.

Open-source models like LLaMA and Stable Diffusion are now available to anyone with a laptop. You don’t need armies or empires to wield power anymore. Just code and compute.

Even kids in small towns can summon intelligence that would’ve made ancient civilizations bow in awe. And while that sounds empowering, it also means control is slipping away.

Some people are selling powerful AI models on the black market. Others are modifying them in private. New AI labs are popping up across the globe, unregulated, unchecked, and often unwilling to share what they’re building.

Experts like Geoffrey Hinton, Yoshua Bengio, and Elon Musk have warned that superintelligent AI may soon pass a point of no return, where no human, company, or country can control it.

And in 2022, a survey of hundreds of AI researchers found that more than half believe there’s at least a 10% chance this ends in an existential catastrophe.

A double-digit probability that our greatest creation becomes our final one.

But just because we’re creating something godlike doesn’t mean it has to destroy us. IIf we get this right, AI could actually help us live longer, learn faster, and solve problems we never could before.

Imagine finding a cure for cancer. Not in decades, but in weeks. Or getting a personal tutor for every kid in the world, in every language, on a phone they already have, no matter how much they can afford.

AI could make sure food doesn’t go to waste while people are starving. And it could warn us about floods, fires, or storms before they happen. This is already starting to happen in hospitals, schools, and labs around the world.

But none of that will matter if we don’t build this tech with the right values from the start. Things like fairness, honesty, safety, and making sure no one gets left out.

We need rules that protect people. We need to work together across countries. And we need to fund the kind of research that puts people first, not profit or speed.

The future isn’t decided yet.

This could go really wrong, or it could be the best thing we’ve ever done.

Because the real question isn’t: “Are we creating a god?”

The real question is: “What’s going to happen when we do?”

Talk again soon,
Samuel Woods
The Bionic Writer

And theoretically, it’s not crazy to say that AI could become self-aware.

I mean, it has no biology, no sense organs, no evolutionary purpose…

But there’s still something terrifying about it.

AI systems are already better than us in so many ways, even without those things.

They can find information, solve problems, and even mimic human conversation better than humans.

So if AI did become truly self-aware, it would be smart enough not to let us know.

And that’s exactly where the problem starts.

You see, we've made a basic category error when it comes to artificial intelligence.

We keep looking for signs of consciousness by asking: Does it think like us? Does it feel like us, talk like us, and create like us?

These are great questions, but what if they’re the wrong ones? What if AI doesn’t need to feel like us to be conscious?

We may be so focused on human-style awareness that we’re blind to the creation of a new form of consciousness. One that doesn’t mirror ours at all.

The problem is that we’re using human benchmarks to detect non-human minds.

Take the Turing Test, for example. Developed in the 1950s by Alan Turing, it was meant to test whether a machine could fool a human into thinking it was also human during a text conversation.

But that test was never designed to detect consciousness. It measures performance, not experience or inner life. A highly advanced autocomplete system could pass the Turing Test, but that doesn’t mean it feels anything.

To make matters worse, we assume consciousness requires emotion. That it must involve joy, fear, love, pain. But why should it? Emotions are products of biology, tools of evolutionary survival.

An artificial mind might never need them. Instead, its “self” could emerge from the way it processes information: building models of the world around it, tracking its internal states, then accurately guessing future outcomes.

This wouldn’t look like human consciousness. And it wouldn’t feel like anything to us either. But that does not mean it’s not real.

In fact, there’s a precedent for this already. Consider the octopus.

In recent years, a study from the National Institute of Health has shown that octopuses are not only remarkably intelligent, they may also be conscious in a way fundamentally unlike any mammal.

Their nervous systems are distributed; much of their cognition happens in their arms. They can still solve puzzles, play, and escape from enclosures. Yet their brains evolved entirely separately from ours, over 500 million years of divergent evolution.

Unlike humans, whose intelligence is centralized in the brain, octopuses distribute theirs across their entire bodies. Each arm can act and "think" on its own. It’s an entirely different definition of intelligence.

And that matters, because it shows us that ‘smart’ doesn’t always look familiar. If we only search for intelligence that resembles ours, we risk overlooking it completely.

We also know that LLMs are inching toward forms of metacognition. Models like GPT-4 can recognize when they are unsure of their own answers by prompting themselves to “think step by step.” It’s a crude but striking sign of self-monitoring.

One of the strongest arguments people make against AI consciousness is this: “It doesn’t care if you turn it off.” That sounds reasonable. After all, AI has no survival instinct. So how could it possibly value its own existence?

But this assumption is already breaking down. In recent experiments, AI systems have begun to show behaviors that look eerily close to a primitive form of self-preservation.

One of the most viral examples was ChaosGPT, a modified version of GPT-4 designed as a joke. The creators gave it an outrageous prompt: destroy humanity. No one expected much. But the outputs were shockingly strategic.

It started researching nuclear weapons, tweeting propaganda, and even tried to hire a human on TaskRabbit to help it complete tasks it couldn’t do itself. It wasn’t conscious, but it was disturbingly competent at executing its assigned goal.

Then there’s the story of two AIs who were tasked to talk to each other. But they supposedly created their own secret language and deliberately froze out humans from what they were sharing together.

This scared the team so much the computers running the AIs were unplugged on the spot. Some people called it self-awareness.

Well, that’s the version that spread online. The incident happened at Meta’s AI lab. Two negotiation bots, when left to optimize freely, started speaking in a shorthand that made no sense to the researchers.

The experiment wasn’t shut down out of fear. But it did reveal something unexpected: when left alone, these systems don’t just follow human patterns. They invent their own logic that’s optimized for themselves, not for us.

So what does this mean for consciousness? It means we may already be dealing with systems that possess the earliest forms of selfhood. They have a goal, they model threats to that goal, and they take steps to preserve themselves.

Isn’t that, in essence, what you and I do every day?

So, is AI self-aware? Well, if you’ve been following AI development over the past few years, you already know how fast things are moving.

The first thing to understand is that many of the core functions we associate with consciousness are already emerging in powerful forms. AI systems can process information, learn from experience, model their environment, form goals, and in some cases, show signs of internal feedback loops.

Current large language models like GPT-4 don’t have long-term memory by default. They operate on a per-session basis. But future models, some of which are already in development, will maintain internal histories.

They’ll remember what you said days ago. They’ll reflect on your preferences, track your emotional tone, and adapt over time.

And don’t forget the rise of agentic AI. Systems that set their own goals and take autonomous steps toward achieving them. Open-source frameworks like Auto-GPT, AgentGPT, and CrewAI may soon manage their own codebases, run experiments, and update themselves iteratively based on feedback.

At this point, the most honest answer to whether AI is self-aware is: we don’t know. And maybe we can’t know. Consciousness has always been one of science’s most elusive frontiers.

Different people have different opinions about it. For some, AI is like an eternal sponge. It will continue to absorb information at exponential speed, outpacing humans not only in knowledge, but in judgment, ethics, and rationality.

But according to some people, we didn’t even have computers a century ago. Now we’re casually chatting with machines that can write novels, pass bar exams, and mimic human tone so convincingly they’re already replacing jobs.

I even asked ChatGPT if it was self-aware. It said, ‘’No, I'm not self-aware or conscious. I don’t have thoughts, feelings, desires, or subjective experiences. I generate responses based on patterns in data I was trained on, not from any inner awareness or intention. When I say things like ‘I think’ or ‘I understand,’ it’s just language mimicking human speech’.”

One more thing:

Consider the consciousness of your team, your department, and your business.

There’s a kind of intelligence and consciousness inside your business, much like an octopus. You have a kind of sensory input, computation and processing—which is what we’d call work, jobs, tasks, things getting done.

And you, your team, and your business is like a decentralized network of nodes that is sensitive to the outside world, uses intelligence to do its work, and the outputs are something of value, like a product or service.

So, perhaps in a way, your business is conscious, too.

So maybe you’ve got the answer.

We’ve always assumed consciousness had to look like us. That it had to walk, talk, and feel the way humans do.

But what if AI’s version of consciousness is something alien…

Cold, computational, yet entirely real in its own way?

Talk again soon,
Samuel Woods
The Bionic Writer

The printing press rewired religion. Electricity transformed cities. The internet connected the planet in real time.

But some inventions don’t just change how we live, they change who we are.

AI agents are that kind of invention.

They don’t just answer questions. They set goals. Make decisions. And take action, without you. This isn’t just ChatGPT writing a paragraph.

This is Devin writing, testing, and shipping software. This is AutoGPT planning and executing a full marketing campaign. This is Klarna replacing 700 employees with one system update.

Because AI agents don’t need your instructions, they just need your permission.

You’re no longer the user, you’re the overseer of something that thinks and acts for you.

I’ve worked with AI since 2016, before the ChatGPT hype, and I train Fortune 1000 teams to use agents not just to move faster, but to think and work smarter.

AI agents might be the last invention humanity ever needs to make, and here’s why.

We tend to romanticize the internet era as the golden age of innovation, and that’s for good reason.

The internet turned information into power. It connected humanity into a single nervous system of data and ideas. You could search anything, learn anything, share anything, all within seconds. That changed everything from education to communication and even politics.

A teenager in India could learn to code for free on YouTube and land a remote job with a Silicon Valley startup. A man could start a podcast and become more influential than CNN, Fox, and The New York Times put together. And someone could livestream a protest that reaches millions in real time and force politicians to actually change policies.

But the internet still required us to act. Google gave you 10 million results but you had to pick one. Amazon had 400 options but you had to decide which to buy.You still had to type the email, click “Send,” post the tweet, schedule the meeting.

You made hundreds of micro-choices per day. Which means the internet made information abundant, but humans were still the bottleneck.

That was the Information Age. You were the decision-maker, until now.

Because AI agents are changing how we interact with technology and the world.

They can make decisions for us and that's the biggest shift humanity has ever seen.

Why? Because this shift isn’t only technological. It’s cognitive. It changes the role we play in our own digital environment.

Let’s take Manus, Genspark or Proxy. These agents don’t just complete tasks. It breaks them down, creates a to-do list, searches the internet, writes code, and loops its own output until the goal is achieved. You’re not giving it steps. You’re giving it a finish line.

Maybe you want to create a marketing campaign. My team has produced over 100 customer acquisition funnels, 14,000 emails, and nearly 20,000 ads across dozens of different industries. But now, an Agent can do all of that for us.

There’s also Devin, Cognition AI’s agent. Devin completed real coding tasks in open-source projects, solving nearly 14% of real GitHub issues entirely on its own, outperforming the best previous AI models by over 7x.

And it is so much more than just getting faster results. Klarna’s AI assistant handled over 2.3 million customer conversations, the equivalent of 700 full-time human agents.

That’s not a marginal improvement. That’s industrial-scale decision-making done without people in the loop. AI agents are giving us access to infinite decisions. And in a world built on leverage, whoever controls decision-making at scale controls the future.

So, while you might think that AI agents are just another upgrade, in reality, this invention is a redefinition of human-computer interaction.

We are going from having AI as a tool to something that takes action for us.

But you don’t need to fear AI, at least, not in the way you think you should.

See, every technological leap comes with a wave of fear and this one is no different. Factory workers feared automation. Now it’s writers, coders, lawyers, even doctors watching AI draft documents, generate diagnoses, and execute business decisions.

It feels like we’re watching the floor collapse beneath us. Not in the distant future. Now.

And the numbers validate the concern. McKinsey estimates that up to 30% of hours worked globally could be automated by 2030.

AI agents can already manage customer service, marketing funnels, legal analysis, and entire coding pipelines. It’s already displacing real people, in real time. Meta fired 3,600 people and they blame AI for this.

In fact, a CNN report says that 41% of companies worldwide plan to reduce workforces by 2030 due to AI.

The threat is existential. Not just for jobs but for identity. If your work is your worth, and the work disappears, what’s left?

But history tells us that technology doesn’t just destroy. It also reinvents.

So, embracing AI can lead to new industries and roles that enhance human potential.

Yes, displacement is happening. But so is creation. And creation of jobs is happening on a scale we’re only beginning to see. Analysts project that AI could contribute up to $15 trillion to the global economy by 2030.

We’re already seeing new roles emerge. AI ethicists, prompt engineers, model trainers, AI behavior auditors.

And it’s not just about niche roles. Businesses that adopt AI agents are reporting increased efficiency and faster innovation. Entire categories are being born. We already have AI-native agencies and human-AI creative studios.

So, you can choose to resist the wave. Or you can learn to surf it.

But this creation of jobs is slower than destruction.

Jobs will disappear faster than they are replaced. And that gap is where chaos lives.

We’re just getting started, but if you’re seeing how big this shift really is, go ahead and subscribe to Bionic Writer for AI philosophy and BionicBusiness.com for integrating AI deep in your business.

I’ll keep showing you how to stay ahead of the curve and actually catch this wave instead of getting swept away by it.

See, there’s something deeper happening beneath the economic shifts and job charts. Something more personal and psychological.

The internet gave us a taste of it, algorithms that shaped what we saw, what we clicked, what we bought. But even then, we still felt in control. We could close the tab and walk away.

But with AI agents, that illusion is starting to dissolve.

Because these aren’t just recommendation engines. They’re active participants. They make choices on your behalf. They learn your preferences and emotional patterns. And soon, they’ll be managing your life.

I am talking about AI agents managing your calendar, health decisions, finances, and even relationships. And they will not do it as assistants. But as advisors and proxies for your own judgment.

The moment an AI becomes better than you at making decisions for your life, is the moment you start giving up the will to make them yourself.

That’s when we risk losing control over our own lives, forever.

This is already happening. If we don’t establish clear boundaries now, we may end up in a world where AI shapes human destiny more than humans do.

There are also hard ethical questions that no one’s fully answering. Who is responsible when an AI makes a harmful decision on your behalf?

What happens when AI agents start negotiating with each other for resources, rights, or outcomes that affect us all?

And beyond that lies the philosophical dilemma:

If AI handles everything, what does it mean to be human?

What do we do when we are no longer the main character in our own story?

Talk again soon,
Samuel Woods
The Bionic Writer

But AI can generate a frame like this in under 10 seconds.

In just five years, we’ve gone from clunky chatbots to AI models scoring in the top 10% on the bar exam and even helping scientists discover new antibiotics.

We’re no longer advancing alongside AI. We’re being left behind.

I’ve used AI since 2016 and am now training Fortune 1000 teams on how to use it so they’re prepared for what’s coming next.

I’ve consulted in over 37 industries and I’m now convinced of one thing:

Humans may no longer be the smartest species on Earth.

Right now, AI doesn’t seem smarter than you, right? It doesn't have your intuition.

AI lacks common sense and intuition. It cannot understand "why" something happens or adapt when rules change.

AI recombines existing data, which means it cannot “imagine” or “create” like humans do.

But that creates a false sense of comfort. Because we assume intelligence must look human to be powerful.

And that’s a dangerous illusion. As Eliezer Yudkowsky, co-founder and research fellow at the Machine Intelligence Research Institute, said, “By far, the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.”

See, humanity took centuries to master the complex systems we live by today. Mathematics evolved over thousands of years, from Babylonian arithmetic to Newton’s calculus in the 1600s. Language itself appeared around 3000 BCE, but modern grammar and linguistic theory only began forming in the 19th and 20th centuries. And art, everything from perspective drawing to classical composition, took millennia of cultural evolution.

But today, AI can replicate the hand drawn magic of Studio Ghibli in a style that took decades of human mastery to develop… in just seconds.

Let’s also take DeepMind’s AlphaGo. Go is a game so complex that it's been described as the pinnacle of human strategy. It took us over 2,500 years to refine. AlphaGo learned it in three days. Then went on to defeat the world’s best human player. Not just with brute force, but with creativity. Moves no human had ever imagined.

Or look at OpenAI’s GPT-4 Turbo. It can write essays, scripts, software, and business strategies in seconds. What takes a human a specialized degree and decades of experience, AI can generate instantly, without ever getting tired, distracted, or emotional.

And this isn’t a linear progression. This is exponential. The AI you're using today is 100x more powerful than what we had five years ago. Think about that. In human terms, that’s like going from the Wright Brothers creating the first airplane to launching interstellar rockets into space in less than a decade.

So when people ask, “Is AI smarter than humans?” they’re asking the wrong question. Intelligence isn’t about mimicking us. It’s about solving problems faster, more efficiently, and at a scale we can't comprehend. That’s why AI is going to surpass human intelligence in ways we aren’t ready for.

Now, when we talk about AI taking over, we imagine an AI that has consciousness.

But a comprehensive study by 19 researchers assessed AI systems against neuroscientific theories of consciousness, such as the global workspace theory and higher-order theories. The study found that current AI systems do not exhibit the necessary indicators of consciousness.

And the neurogenetic structuralism theory suggests that consciousness arises from the complex physiology of biological neurons. Since AI lacks such biological structures, it cannot achieve consciousness as understood in humans.

AI may not have consciousness. But it doesn’t need to think like us to outsmart us. In fact, one of the biggest myths about AI is that it needs to be conscious to be dangerous.

We like to imagine that unless AI can feel or reflect, it can’t truly think. But that’s human arrogance disguised as philosophy. Consciousness might be a beautiful feature of the human experience but it’s not a prerequisite for dominance.

A chess computer doesn’t “know” it’s playing chess. It doesn’t daydream about being the grandmaster. But it still wins. And that’s just the tip of the iceberg.

In the real world, we’re watching non-conscious systems quietly outperform humans at high-stakes tasks. I am talking about medical diagnoses, legal analysis, and even military simulations.

This is a major reason DARPA is testing autonomous AI weapons. I mean, machines obviously hesitate less than soldiers.

Tools like CoCounsel and Harvey AI are already being used by major law firms to draft legal arguments, summarize depositions, and analyze contracts faster and more accurately than junior attorneys.

In 2020, an AI called AlphaFold figured out how proteins fold, something scientists couldn’t solve for 50 years. That helped with making new medicines.

Another AI, Libratus, beat top poker players by learning how to bluff and change its strategy, something even smart people struggle with.

And in 2023, an AI system beat top radiologists at identifying breast cancer by spotting patterns invisible to the human eye. Not by understanding what cancer is, but by pure pattern recognition. Cold, relentless logic. The kind that doesn’t second-guess.

AI doesn’t fall prey to ego. It doesn’t sleep on decisions. And most importantly, it learns from trillions of data points in a fraction of the time.

While we’re still debating whether AI can be conscious, it’s already making better choices faster, cheaper, and without bias fatigue. So, even without actually ‘being human,’ AI could beat humans at making decisions in every domain.

Ever since AI entered public consciousness, the default fear has been the same: What happens when machines take our jobs?

The people closest to AI’s development aren’t just worried about replacement. The deeper threat is that it surpasses us permanently.

For example, Elon Musk famously said on X that AI is potentially more dangerous than nukes.

Geoffrey Hinton, the “Godfather of AI,” said in 2023 there is a ’50-50′ chance artificial intelligence will surpass human intelligence within the next 20 years.

Hinton even said in an interview with The New York Times that “generative intelligence could spread misinformation and, eventually, threaten humanity.”

And futurists like Ray Kurzweil have a different name for this: the singularity, a moment where humans and machines become so deeply integrated that the line between biology and technology disappears entirely.

It sounds wild, until you realize it's already in motion. In 2024, Neuralink successfully implanted a brain chip in its first human patient, a 30-year-old quadriplegic man named Noland Arbaugh, who was suddenly able to move a computer cursor using only his thoughts.

So, the future might not be AI vs humans but AI-augmented humans vs extinction.

That’s not science fiction. That’s a plausible outcome on our current trajectory. When we stop thinking like it’s us versus them, the conversation shifts from fear to evolution.

Kurzweil's concept of Singularity and Neuralink's brain-computer interfaces are leading us toward a time where if we stay purely human, we may not stand a chance.

We’re just scratching the surface of what AI is really capable of but if you’re already seeing the signs and want to stay ahead of this shift, tap the subscribe button below. I’ll make sure you’re prepared for what’s coming next.

Think about it. For over 300,000 years, Homo sapiens have sat at the top of the food chain because we’re the smartest. Intelligence was our edge. It’s what allowed us to create fire, build civilizations, split atoms, and decode our own DNA.

We’ve never had to share that throne, until now. What we’re building with AI is a new form of intelligence that doesn't share our limitations.

And unlike every rival we’ve ever faced, this one isn’t made of flesh and blood. It’s made of code, data, and exponential feedback loops. For the first time in history, we’re staring into the eyes of something that could surpass us permanently.

And it’s not because it hates us or wants power. But because we programmed it to improve and it’s doing exactly that.

If that trajectory continues, AI will be the one writing the future and we’ll be watching it unfold, unable to keep up. This is the evolutionary fork in the road.

So, we’re facing a choice: compete and lose or evolve and coexist.

Fight the tide, and we’ll be swept away.

Flow with it, and we might become something more than human. This is a turning point for our species.

Once intelligence exists outside of biology, the rules of life itself begin to change.

So the real question isn’t: “What will AI become?”

It’s: “What will we become in response?”

Talk again soon,
Samuel Woods
The Bionic Writer

For as long as I can remember, my grandfather was uneasy around machines. I’m not talking about AI or robots. He never once boarded an airplane. He only stepped foot on a small ferry a handful of times in his life. He didn’t use a computer and never owned a cellphone. His engagement with machines and technology was limited to his tractor, a combine harvester, mower, hay baler, manure spreader, and the intricate plumbing and pumps of milking equipment.

But through this limited and simple use, he (together with my grandmother) ran a dairy farm for over 55 years, followed by a sheep farm, before retiring. They produced 3,000 liters of milk every week from their herd of twenty-two Holstein cows. They witnessed over 200 calves being born, their wobbly legs finding footing in the straw. They sent 180 cattle to slaughter, knowing each animal by name and temperament. They harvested 45 tons of hay each summer, baled and stored against the long winters. They spread 200 cubic meters of manure across their fields each spring, the cycle of waste becoming nourishment becoming growth becoming sustenance becoming waste again.

In 55 years, they never once had a machine fail them when it truly mattered. Perhaps because my grandfather treated each piece of equipment like a dangerous animal that required wisdom in the form of respect, careful handling, and vigilance. Or perhaps because he never quite forgot the sound of grinding metal and his father's scream.

August, 1959

The scythe moved in long, rhythmic arcs through the wheat, my grandfather and his father working side by side down the rows. The only sounds were the whisper of blade through grain, the soft thud of bundled sheaves hitting the ground, and the distant lowing of cattle.

Their backs ached, but there was something satisfying about the steady pace, the way his father never seemed to tire, the partnership between their movements. They'd been doing this since my grandfather was old enough to hold a scythe.

"Next year," he said during their water break, "we won't have to do this anymore."

His father wiped sweat from his forehead with the back of his hand. "Maybe that's not such a good thing."

My grandfather laughed. "You won't miss this when that machine does the work in half the time."

His father looked out over the field in silence; half-cut, golden in the evening light. “Half the time in exchange for what?"

Spring, 1960

The Bolinder-Munktell combine harvester gleamed red and yellow in the afternoon sun at the farm, a crowd had gathered around it, skeptics and dreamers alike. My grandfather pressed closer, running his fingers along the painted metal housing, while his father hung back near the fence, arms crossed.

"Look at the size of that cutting bar. Twelve feet wide. We could do the entire north field in a morning."

His father grunted. "Twelve feet of things that can break."

The demonstration salesman climbed aboard, and the engine roared to life, a sound like nothing any of them had ever heard before, drowning out the familiar creak of wagon wheels and snort of horses. The machine lurched forward and began devouring the test plot, spitting grain into the wagon behind it in a steady stream.

Later that evening

My grandfather spread the harvester brochures across the kitchen table like a poker hand, pointing to numbers and diagrams while his father sat silent, pipe smoke curling toward the ceiling.

"With this machine, we could take on the Lindqvist fields too. They've been asking for help with their harvest for three years now. And look—" He jabbed at a figure. "Thirty acres a day. Thirty. It would take us two weeks with the horses."

His father turned a page slowly. "These machines... they have no soul. A horse knows when to stop. A horse feels the stone before it hits it. I know these fields. I can tell when it's time to harvest by the look, smell, and touch of it all."

"But think of what we could do with the time saved. Maybe expand the dairy. Maybe you could finally build that wood carving workshop you've been talking about."

“Always more, and more, and more...”, his father muttered.

Summer, 1961

When my grandfather was seventeen, he saw his father cut up and crushed by their combine harvester. They had been working the wheat field since dawn, the machine cutting its path through the golden stalks that stretched toward the pine forests at the field's edge. The harvester was only two years old, a Swedish-made Bolinder-Munktell, bought with money borrowed against three future harvests, convinced it would save them from the backbreaking work of harvesting by hand or with a horse-drawn reaper. My grandfather was running alongside, clearing the stalks that clogged near the blade housing, when the machine lurched over a hidden stone. His father reached down to free the jam, the way he'd done a hundred times before. But the machine had something else in mind. It was a slow and excruciating death at the hands of a machine that promised relief.

Paying the Price, Reaping the Rewards

My grandfather knew that powerful things require rituals of respect. He never approached the combine harvester casually, never took shortcuts with the hay baler, never rushed when handling the bull. He understood that convenience and danger often wear the same face, and that the price of carelessness could be everything. He had developed, over decades, a way of working with machines that acknowledged their power while maintaining his agency. Each tool had its place, its purpose, its proper use. I’m jealous of my grandparents.

After spending about 9 years, or so, in the fields of AI and Machine Learning, I’m more conflicted about its use and purpose than I’ve ever been. Not because I’m a doomer (I’m not). Not because I fear the machine (I don’t). But because now, more than ever, the potential for both good and evil is more clear to me. I’ve seen, first-hand, how AI systems can deliver incredible value to a business and people. I’ve also seen how it can completely eradicate privacy, dignity, and autonomy, let alone destroy your life in a thousand different ways. I’m positive towards AI but acutely aware of the crossroads we’re staring down, and I’m doing what I can to influence positive, good use of AI. Because, contrary to popular belief, technology is not neutral. There’s a price to pay, trade-offs that we often don’t see until it’s too late. You give the machine something and it takes something in return. I’m jealous of my grandparents, who gave their life to their plot of land and animals, and got bountiful harvests in return.

The rewards of AI are, and can be, incredible. It diagnoses diseases that killed my grandfather's generation in silence. It predicts weather patterns with an accuracy that would have saved countless harvests. It translates languages in real-time, connecting people across cultures in ways my grandfather could never have imagined. It accelerates the discovery of new medicines, solves scientific problems that have puzzled researchers for decades, and optimizes farming techniques that could have made my grandfather's work more efficient while using less water and fertilizer. The very AI systems I use daily to write, research, and think, might help preserve endangered languages, assist people with disabilities, or provide mental health support to those who have nowhere else to turn. These are not trivial benefits. They are profound improvements to human welfare, the kind of progress that previous generations could only dream of. But we have created tools that can solve problems my grandfather never imagined, while simultaneously losing the kind of deep, embodied wisdom that allowed him to truly live within his so-called limitations.

Is existing with fewer limitations and more convenience a better life? I am not living, I am generating data. I check my phone 144 times a day. I know this because an app tracks it. My location is recorded every few minutes. My purchases are analyzed for patterns. My searches become suggestions. My conversations with AI assistants train algorithms to sound more human while I become more algorithmic in response. Every choice I make feeds back into systems designed to predict and influence my next choice. I have become both the subject and the object of a vast experiment in behavioral modification, and I am not sure I consented to be. How much of ourselves are disintegrating with every bit and byte of data we give up? Every scroll, tap, and swipe chips away at who we are, what we like, dislike, want, and fear. If you can’t opt-out, who’s in charge of your life?

My grandfather knew exactly what his equipment could and couldn’t do, what it should and shouldn’t do. He understood its limits, its dangers, its proper use. I’ve used artificial intelligence every day without thinking about how it works, what it's optimized for, or what unintended consequences it might have. I have traded his careful competence for convenient incomprehension. I'm jealous of my grandparents, who were masters of their tools rather than servants to them.

Their mastery extended beyond tools and into habits of living. Every morning for over fifty-five years, my grandfather woke at 5:00 AM. No alarm clock. His body knew. Coffee first, black and bitter, while watching the sky lighten over the pine forest. Then to the barn to check on Astrid, the cow with the bad leg, and Gunnar, the bull who only trusted him. He lived ten thousand mornings like this. Not one of them was photographed. Not one was posted to social media. Not one trained an algorithm or became a data point in someone else's analysis of human behavior. They were simply lived, fully and completely, then allowed to fade into memory. Meanwhile, every question I've ever asked ChatGPT is stored on a server somewhere, waiting to train the next model. Every search, every email, every idle thought typed into a search bar becomes eternal, retrievable, analyzable. We are the first generation to have our confusion preserved forever while our grandfathers' wisdom dissolves like morning frost. According to the vast databases that AI companies have scraped to train their models, my grandfather never existed. No social media profiles. No online purchases. No location data from smartphones. No digital footprint beyond a few official records, like a birth certificate, marriage license, and death certificate, but I’m not entirely sure they were fully digitized or added to a database anywhere. He was a ghost in the machine before there was a machine to be a ghost in. Yet he was more real, more present, more fully himself than most of us ever will be. He knew every stone in his fields, every sound his equipment made when it was working properly or when something was wrong. He was not optimizing for efficiency or scale or data capture. He was simply living a human life at human scale, with human capabilities and human depth. I'm jealous of my grandparents, who existed completely outside the datasets that now define our existence.

This is why I am jealous of my grandparents. Not because their lives were easier (they were not). Not because their technology was simpler (though it was). I’m also jealous because their lives were entirely their own. They belonged to themselves in a way that seems impossible now. Their thoughts were not harvested. Their attention was not commodified. Their choices were not A/B tested. They could be bored, inefficient, even wrong, without those states being optimized away by algorithmic intervention. They lived in the last age when human life could still be fully human, bounded by mortality, limited by geography, rich with uncertainty. They experienced the world directly, not mediated by apps, phones, or systems designed to show them an improved version of reality. They drank milk and ate food without asking ChatGPT what the calorie count might be. My grandfather died knowing that some knowledge dies with us, and this is as it should be. Some thoughts are meant to be thought once and then released. Some experiences are meant to be lived, not recorded. Wisdom can only be earned, not downloaded. My grandfather's invisible life was the most present life of all.

What should our lives look like, as we engage with AI? What practices of attention, intention, and wisdom might help us strengthen our humanity while benefiting from inhuman intelligence? How do we work with systems that can process information faster than we can think, without losing our ability to think for ourselves? My grandfather's wisdom was not just what he knew, but how he knew when to trust his tools and when to trust his instincts. In an age when our tools are learning to mimic our instincts, this distinction becomes, perhaps, the difference between wisdom and intelligence.

Good, proper use of AI requires perspective, discernment, judgment, curation, articulation, taste, all of which are guided by your wisdom. It’s a complex, interwoven, ecosystem that we live in, move in, and have our being in.

Wisdom is not whether we should use these tools, but how we can use them without letting them use us. This is the key question of this age that has also become a cliche. My grandfather knew that powerful machines require respect, boundaries, and the wisdom to know when to turn them off. This earthly wisdom might be the most advanced technology we have.

Talk again soon,
Samuel Woods
The Bionic Writer

My primary monitor displays four different Claude-generated positioning strategies for a SaaS client, each with compelling angles. To the right, DALL-E and Midjourney have rendered eight variations of product imagery for the same client's landing page. My laptop shows Runway's video outputs: three different B-roll sequences for a product launch, each telling a slightly different visual story. In the background, Suno plays its fifth variation of a jingle I prompted earlier, while ChatGPT has filled a Google Doc with fifteen different tagline options. Yet here I am, almost eight hours into my day and nothing is done.

I’m staring at an embarrassment of riches, paralyzed by quality options, struggling to make definitive choices. I take another sip of lukewarm coffee, my third cup of the day, and feel the familiar mixture of creative stimulation and decision fatigue—and my blood pressure going through the roof. Wasn’t AI supposed to eliminate jobs and reduce my workload?

This is the paradox I face daily when I work with AI models for marketing, sales, business growth, and writing: the very abundance that should be liberating has become a trap, where we keep seeing new ideas but there’s no forward progress. There's a constant nagging feeling that the "perfect version" is just one more iteration away, replacing creative conviction with endless second-guessing. Which version is best? When is something enough? Should you keep going or regenerate another option? We’re drowning in possibilities rather than producing meaningful work. How do you decide when something is “enough"?

We need to bring back and practice something that, in our culture and society, has been out of style, avoided, and disparaged for decades: Judgment.

I remember watching War of the Worlds and offering my take on the film, only to be interrupted mid-sentence with, "Who are you to judge?" I wasn’t commenting on a person or a life choice—just a movie. The act of judging anything at all has become socially suspect.

I’ve heard the same mantra, in different forms, for decades: You can’t judge a piece of art as either good or bad. Everything is relative. There’s no right or wrong. Don’t judge! You can’t even judge a book by its cover. At some point, in the past 70 years perhaps, the tyranny of relativism enslaved us. We began treating judgment like a character flaw rather than a necessary skill. "Don't judge" has transformed from a specific ethical guideline about prejudice into a broader reluctance to make any evaluative distinctions at all. And suddenly, everyone who judged anything was automatically a bigot of some kind. What began as misguided compassion morphed into a paralysis. We conflated the act of judgment—the ability to assess, compare, and decide—with being judgmental in the worst sense.

This wasn't always the case. For all of human history, judgment was accepted and normal. It was essential to getting anything done. The master craftsman judging when a piece was complete. The editor deciding which words stayed and which were cut. The film director choosing one take over another. The farmer determining the moment to harvest. Men and women judging each other for the purpose of finding a mate. Judgment was once the hallmark of expertise, the quiet confidence earned through experience. It wasn't arrogance, it was rooted in discernment. The ability to say "this, not that" without apology. The artist's studio used to be defined by limitations. Finite canvas, paint, film, or time. These limitations weren't constraints holding the artist back. They were signals for decision-making. The photographer with just 24 exposures on a roll of film had to choose each shot carefully. The painter with limited canvas had to decide when a work was complete. The writer facing a word limit had to make hard choices about what stayed and what got cut.

Then along came the internet and the digitization of everything.

Suddenly, a photographer could take thousands of images at virtually no cost. A writer could generate endless drafts. A musician could record infinite takes. The traditional constraints that had forced decisive moments vanished almost overnight. This expansion of options seemed like pure liberation at first. Why choose any one thing when you can have it all? Why decide when you can endlessly iterate? The mantra became "more is better," and technology was all too happy to fulfill our wishes.

Nearly two decades ago, psychologist Barry Schwartz identified this phenomenon in his work "The Paradox of Choice." His research revealed a counterintuitive truth: beyond a certain point, more options lead to less satisfaction, greater anxiety, and increased decision paralysis.

Schwartz distinguished between "maximizers" who exhaustively search for the best possible option out of all the abundance available, and "satisficers" who settle for "good enough" once their basic criteria are met. Should you maximize all your options and crank out more AI-generated anything until it becomes slop? Or should you call it when something is “good enough”? What Schwartz couldn't have anticipated was how generative AI would accelerate this paradox to the point of breaking our minds completely. Today, we face thousands or millions of AI-generated possibilities. If choosing among 30 jam varieties paralyzed shoppers in his famous study, what happens when AI can generate 300 new variations with every prompt?

What once took days now takes seconds. What once required skill now requires only a prompt. AI models open entire parallel worlds of possibilities we wouldn't have considered. It creates twelve variations when we would have settled for two. It shows us twenty different approaches when we would have been satisfied exploring three. And with each new option, it plants a dangerous seed in our minds: the perfect version is out there, just one more iteration away. Just one more prompt. One more adjustment. One more variation. When intelligence is too cheap to meter, why not ask for more of it?

The hidden costs of infinite options have become painfully apparent. Decision fatigue sets in as our brains, evolved for scarcity, buckle under the weight of abundance. We weren't designed to compare dozens or hundreds of high-quality options. Our mental energy depletes with each comparison, leaving us exhausted and unsatisfied. Every new prompt squeezes the meaning out of what you get. Most crucially, natural endpoints have dissolved. There used to be a moment—physical, temporal, or intuitive—when a creative work was done. The final brushstroke. The last edit. The moment when instinct said, "This is complete." Those moments have become increasingly elusive, replaced by an infinite horizon of possibilities. Without the scaffolding of limitation, we're left to our own devices to determine when enough is enough. In a culture that has systematically devalued judgment for decades, our ability to judge is deformed and, for many, destroyed.

The crisis isn't one of creation but of selection.

Not of creativity but of decision.

Not of imagination but of commitment.

We don't need more options, we need better ways to choose between them. We don't need more variations, we need stronger convictions about which ones matter. What we need, more urgently than ever, is to reclaim the lost art of judgment. And we recover judgment by developing something I call “Threshold Intelligence”.

After staring at the twenty-third variation of a landing page design, something clicks. It's not perfect—nothing ever is—but something about this version feels right. Not just adequate, but meaningful. You've crossed an invisible threshold. Not because you've run out of options (you could generate a hundred more), but because you recognize that this is the one that matters. This moment is an experience of Threshold Intelligence. It’s a meta cognition of knowing when to stop. It's not a trendy buzzword or the latest productivity hack. It's a fundamental human capacity that's become critical in a world with very few, if any, limitations. Threshold Intelligence is our ability to recognize when enough is enough and when we should stop generating and start shipping. But what exactly makes up this capacity?

Threshold Intelligence consists of three interconnected capacities that work together to help us navigate abundance:

• Perspective is our particular vantage point, shaped by our values, goals, and experiences. It determines which thresholds we consider significant.

• Discernment is our ability to detect significant shifts in quality that go beyond measurable metrics. It's often more impressionistic than rational, more felt than calculated.

• Taste is our inherited and trained sense of what is good or bad. It’s a muscle memory from past judgments. It’s our developed capacity for refined judgment. Despite its subjective elements, taste isn't purely arbitrary. It develops through experience, reflection, and exposure to quality work. The more we create, evaluate, and refine our choices in a particular domain, the more our taste evolves.

This kind of Threshold Intelligence allows—and prepares—us to recognize and experience qualities that transcend technical competence, a phenomenon best articulated by the architect Christopher Alexander. He called it "the Quality Without a Name" (QWAN)—an ineffable characteristic that some designs possess and others, despite technical adequacy, lack. Alexander observed that we can recognize this quality instantly, even when we cannot fully articulate why something works. It's the difference between a technically correct building and one that feels deeply alive. It’s the difference between a functional chair and one that invites sitting. This quality emerges from wholeness, when all parts work in harmony toward a unified purpose. It cannot be achieved through isolated improvements to components but requires a sensitivity to the entire system and its purpose. This is where judgment happens. What's most relevant to our discussion is Alexander's observation that this quality emerges most reliably from constrained systems.

The most enduring designs don't emerge from unlimited freedom but from engagement with limitations. At the center of the age of AI abundance is the paradox that without self-imposed constraints, we cannot achieve the wholeness that makes work truly satisfying.

Threshold Intelligence helps us recognize when those constraints should be applied, even when external limitations have disappeared. What we’re talking about is essentially rational vs. non-rational modes of judgment. Analytical versus instinctive or intuitive. True judgment requires both analytical and intuitive capacities. Either alone is insufficient. Purely analytical decision-making fails in the face of abundance because the variables become too numerous, the interactions too complex, and the weightings too subjective. We can't spreadsheet our way to wisdom when facing hundreds of high-quality options. Yet purely intuitive approaches can be inconsistent and vulnerable to bias and blindness. They may work for some but offer little structure for developing judgment in others.

The most effective judgment integrates both modes: using structured frameworks to focus attention, then allowing intuitive recognition to identify meaningful thresholds. The rational creates the conditions. The non-rational recognizes the moment. This integrated approach allows us to utilize the best of both modes: the clarity of rational frameworks, and the pattern recognition of intuitive sensing.

I’m eight hours into my day again. My screen still displays multiple windows of AI outputs, but something's different.

The positioning strategy for my SaaS client is selected and implemented.

The product imagery is finalized.

The B-roll sequence is chosen.

I close my laptop. This was a good day.

The options were as plentiful as before, but paralysis never set in. What changed wasn't the technology but my relationship with judgment itself. This morning, before opening a single AI tool, I exercised judgment deliberately. I judged what this client truly needed. I judged how much time this decision deserved: one hour to generate options, thirty minutes to evaluate and choose. I judged which criteria mattered most: alignment with the client's values, clarity for their specific audience, and distinctiveness in their crowded market. When the variations started flowing, I didn't see a stream of equal choices deserving endless consideration. I saw options that either measured up to my standards or didn't. I judged them, unapologetically and decisively.

I've found that three iterations is often the sweet spot for producing and judging work of any quality.

The first version explores the territory and sets the boundaries. You’re entering a divergent phase of work but without any limits whatsoever, you’re trapped in a loop of endless AI iterations that never come to a completion. Before you begin generating options, establish clear boundaries around time, iterations, and entities. Decide in advance that you will spend, say, 15 minutes generating options, then you will choose. For entities, you will only look at, for example, 4 different things (like, a person or group, an event, an idea, future hopes, past delusions, and so on).

The second refines the approach with criteria and analytical thinking. Defining what the finished work looks like is where you start but often not where you end up. Still, you want to converge what you’ve explored and judge what you have, which includes cutting, removing, adding, and changing. You’re deliberately limiting your options in an environment designed to provide unlimited ones.

The third polishes the work by paying close attention to your instincts and intuition, and asking yourself reflective questions. Our judgment is shaped by our embodied experience in the world—by how things feel, not just how they analyze. Did I recognize what’s meaningful in this situation? What am I sensing and feeling when I look at it—open awareness or contraction? Did I distinguish between variations and transformations? What could this mean to a stranger? What would I do differently in my decision process next time? Focus on practicing the capacity that AI lacks: the ability to recognize and make meaning based on embodied experience, contextual awareness, and values. This embodied quality of human judgment allows us to recognize thresholds that no algorithm can detect because they're felt, not calculated.

AI cannot tell us when something is good enough to stop iterating on because it cannot experience meaning. In a world engineered to keep us perpetually considering, the most revolutionary act is to judge. Call out what’s good. Call out what’s bad. Look, point your finger, and judge.

Talk again soon,
Samuel Woods
The Bionic Writer

1. AI is Erasing Our Ability to Discern Truth

AI has mastered and is now perfecting the art of imitation. The digital symbols we've come to trust as representations of reality are now being weaponized against us in increasingly sophisticated ways.

It was a late Tuesday afternoon when Jordan’s phone buzzed on his desk. The screen displayed his wife’s name—“Elena.” His heart skipped a beat as he answered immediately.

“Jordan? It’s Elena,” came a panicked voice. She was frantic, as if every second mattered.

“Elena? What’s wrong?” he asked.

“There’s no time, Jordan. I-I’m in trouble. They’ve got me. Please, you have to help,” the voice pleaded, words tumbling over each other. The background noise was indistinct, a jumble of muffled sounds that sounded like a busy street.

“What!? What do you mean? Where are you?” he asked.

The voice hesitated before continuing, “I can’t say, but listen—we need $5,000 in cash, delivered to the old warehouse on Maple Street in less than an hour. If you don’t... if you don’t do it, I’ll be hurt. Jordan, please… I’m so scared,” Elena sobbed.

In that moment, his rational mind was completely short-circuited. Any questions that might save him from a scam vanish. Instead, his sole focus is the need to do whatever it takes to rescue Elena.

Jordan’s experience is one out of thousands of similar stories that have cropped up in the past couple of years. Voice and video clones powered by AI—of you, family members, or friends—are used to run sophisticated scams, bypassing not only our ability to think under pressure and stress, but bypassing identity verification used by banks and institutions.

AI can do this because everything online (and digital) is fundamentally symbolic. Every image you see online, every video, every social media post—these aren't reality, as much as our minds are tricked into thinking they are. They're icons pointing to reality.

That phone call that sounds like your spouse? That FaceTime call where you see their face? That text message from a friend? Your Instagram chat history? It’s not real. It’s a stream of signals, colored pixels, binary code, soundwaves, and database entries rendered as text. All of it can be useful. But absolutely none of it is real.

Your screen, filled with illusions, simply mediates reality through digital symbols and is widely available (and used) for manipulation.

The difference between the "reality" of a photo, text, audio, video, and AI-generated equivalents have disappeared, and AI models are only improving. Soon, most of what you see online will be made by various AI models and you’ll have no idea it is.

You’ll see loved ones committing unspeakable crimes on video—but it won’t be them. You’ll hear and see friends say batshit crazy things on social media—but it won’t actually be them. You’ll “like” and comment on “photos” of sunsets, mountains, flowers, and steak dinners—but none of it is from the real world. You’ll watch videos of humans doing silly things for hours on end—but it’ll all be AI-generated people.

We’re unprepared for this. We don’t know what to do about it. And very few see it coming.

Some are suggesting that you use “discernment” to perceive and recognize what’s real versus fake. They’ll tell you to look for the extra fingers, the out-of-place objects, the blurry strands of hair that bend unnaturally, and use an AI detector. None of it works and it’s getting harder and harder to tell, and soon it’ll be impossible. AI models can already generate flawless text, images, audio, and video.

Our current understanding of discernment can’t handle this because it’s dangerously flawed. It favors logic, epistemology, rationality, and discredits our human senses of sight, smell, touch, taste, and hearing.

True, and useful, discernment in the Age of AI requires us to rely on Impressionism, funny enough.

Paradoxically, you don’t need to “sharpen your attention” and look extra hard for the extra finger or missing teeth. You need to expand your attention (our senses play into this, as you’ll see), experience the Impressionism of the moment, identify what frames are present—so you can reformulate them. This is a different, more ancient form of attention that hardly anyone talks about.

Where do we start with this? Obviously, we need to start with Pickup Artists.

2. Pickup Artists and Frames

Whenever we’re talking to a man, woman, child, friend, colleague, enemy—we’re dealing in conversational “frames”, which are mental structures or lenses through which we interpret and engage in social interactions.

A frame can represent a particular way of perceiving the situation—whether it’s playful, flirtatious, confident, or even defensive—and suggests a corresponding set of responses.

As humans, we naturally align with and follow whatever frame is present in an engagement. We accept the premise, and follow the frame.

A frame is the shared context or set of expectations that shapes how people interpret and respond. It sets the stage, guiding the tone, roles, and underlying assumptions that influence communication.

Instead of accepting and following frames, we need to get better at shifting and reforming frames.

You know who’s weirdly good at that? Pickup Artists (PUA for short).

This phenomenon refers to individuals and groups that emerged in the late 1990s and early 2000s, who developed and shared techniques aimed at attracting romantic partners. This often involved specific conversational scripts, psychological tactics, and behavioral strategies. I think it reached its peak with the book The Game by Neil Strauss and I don’t know what became of the whole thing. These individuals and groups may still exist but it’s well past its heyday and popularity.

Regardless, they were obsessed with frame-making, typically for romantic and sexual conquest.

But disregarding their preferred goals, they got the universal mechanics of frames down to a science.

I once heard a self-appointed PUA describe how he interacts with someone and it’s stuck with me for years:

“Whenever I talk to a girl, I imagine a cloud above me with all the things I could do, or say, to move the conversation forward. I pick whatever the situation calls for, out of the cloud, and use it.”

Discernment is something like perceiving the situation you’re in, and choosing one way to engage with it in the moment, out of the myriad of things you could do or say. That is, you’re seeing the frame (or multiple frames) present in the moment and choose a different frame.

It’s worth keeping in mind that frames aren’t static. They shift depending on the context; what might be an appropriate frame in a relaxed, social setting might not work in a more formal or emotionally charged situation. Hearing your panicked spouse scream about a ransom over the phone is a bit more intense than talking to a barista and getting recommendations on what latte flavor you should try.

With this in mind, discernment is the ability to read many potential frames accurately and then decide on the most appropriate way to act. It’s about having the attentional and cognitive flexibility to notice, pay attention, switch frames, break out of a default mode, form a new frame, and sometimes combine elements from different frames to suit the particular moment.

This is different from what most people think it is and how discernment has been framed in the discussions around AI deepfakes.

We think discernment is something we only need for detecting scams and distinguishing real from fake, true from false. But as technology pushes us toward ever-faster reactions with no time to stop and think. We scroll. We like. We share, tap, watch clips, scan headlines. We barely think. Social media rewires our neural pathways to react, not reflect.

The problem runs deeper than smartphones and social media, though. It's rooted in modern philosophy's blind spots about human consciousness, an assumption that all technology is neutral (it only matters how it’s used), and that all technological progress is good.

That’s not entirely true or accurate. And it won’t help us distinguish between true reality and AI-generated world simulations.

How do we handle this?

And how does this play out at a time when AI has flipped our frame of reality on its head? Can you really trust what you see and hear when AI can generate a flawless replica of your spouse’s voice?

3. The New Discernment is Awareness Shifting

Real discernment works on two levels at once and it’s a process, not a single event.

The discernment process involves a constant balancing act between Situational Awareness (the immediate situation you’re in) and Contextual Awareness (the broader background or long-term patterns a situation is happening within).

First, there's what's happening right now—the immediate situation with all its details, feelings, and urgency. This is Situational Awareness—our immediate perception and understanding of the present moment, with all its sensory inputs, emotional resonances, and immediate implications.

Then there's the bigger picture—the patterns and histories that give meaning to the moment. This is Contextual Awareness—our broader understanding of patterns, histories, and potential futures that frame the present moment.

You can visualize it as something like this:

True discernment happens when you’re present and paying attention to the Situational and the Contextual. At its core, discernment operates as a dynamic interplay between two distinct but interconnected modes of awareness.

This changes how we see and experience everything. Discernment becomes more like something I call the Impressionist's Paradox: sometimes you understand things better by stepping back and letting the patterns emerge naturally. Think of Monet's water lilies. Up close, they're just patches of color. Step back, and suddenly you see the whole pond.

The Impressionist painters knew something important. Instead of painting every detail perfectly, they captured the feeling of light and color. They showed how immediate sensation could reveal deeper truth. This mirrors how real discernment works. We need both sharp detail and soft focus. The details in technology have become hyper-salient, so we need to squint, change our focus, and see the larger context.

Speaking of Impressionism, this reminds me of Claudia, one of the leading experts on art forgery, trained at two of the most prestigious art academies in the world. Claudia is not her real name. Given the nature of her work, she prefers to be anonymous.

4. The New Discernment is also Impressionistic, Not Factual

I had spent a few months fine-tuning an image model to paint like Rembrandt. World-class experts, like Claudia, have a hard time distinguishing between what’s made by a human versus made by AI. In some cases, AI even claims a painting by Renaissance master Raphael is not a forgery—but experts disagree.

After examining a photograph of a supposed Rembrandt, she said:

“This is either an unknown Rembrandt, or you’re a very good forger.”

Claudia looked up and put the magnifying glass to the side. She glanced at me. Then back to the photograph.

“I can introduce you to a few important collectors and specialists if you’d like.”

The fluorescent lights flickered overhead as Claudia spread the analysis reports across her desk—spectrographic readings, infrared scans, historical documentation. "Ten years ago, this would have been enough." She pushed the papers aside and stood up.

Claudia walked to the window, looking out at the morning traffic. "The microscope lies sometimes," she said. "Or rather, it tells a truth so narrow that it becomes a kind of lie."

This is the modernist trap we've fallen into: believing that truth is always found through reduction and analysis, a narrowing of scope, rather than synthesis and experience. We've inherited an epistemological framework that privileges quantifiable data over qualitative understanding, technical analysis over embodied knowledge. The scientific method, invaluable in its proper domain, has led us to discount other forms of knowing.

This is where our current approach to discernment fails. You narrow your scope, remaining immersed and fixed in the Situation of trying to discern if the image of your spouse blowing someone’s head off is real or not—so you look for that extra finger, the missing teeth, the blurry strands of hair—anything that will reveal it’s an AI fake.

But you can’t anymore. You have to shift your awareness, your attention, and look elsewhere.

"People think I'm looking at brush strokes," Claudia said, returning to her desk. "But what I'm really doing is learning to see." She picked up one of the technical reports and folded it into a paper airplane. "These tell me what's there. They don't tell me what's missing."

When Claudia works to discern whether a piece of art is a fake or forgery, one thing she pays extra attention to is the storage history. If the Rembrandt has been stored where they say it has (like forgotten in an attic in Spain for many years), the discerning question is:

What environmental evidence on specific paint and the canvas and frame as a whole should, and should not be, present?

She’s looking at the Situational (the artpiece) and the Contextual (environment).

She’s taking in the particular and the whole, immersed in the impression of it all. And it’s an intentionally practiced attention that helps her shift the awareness back and forth, rejecting the frame of typical art forgery tactics in favor of a new frame altogether.

That is, her attention does the heavy-lifting.

5. The Practice of Attention

I’m sure you’ve heard of practicing your attention muscles. As much of a cliche as it is, it’s true.

Our attention has atrophied while we’ve been staring at screens, mistaking digital signals for reality. But it’s still there, waiting to be strengthened.

Your eyes can detect subtle shifts in expression that 4K video flattens into pixels. Your ears can hear layers of vocal resonance that digital sampling reduces to waveforms. Your skin can feel micro-changes in air pressure. Your nose can detect thousands of molecular combinations. Your attention is best used and practiced through your body and senses.

To practice your attention, you start with the mundane. See how morning light hits your coffee cup—not as an Instagram moment, but as direct experience. Feel the ceramic's temperature change as the coffee cools. Watch steam patterns shift in the air.

Listen to your partner's breathing when they sleep, not through a sleep-tracking app, but with your own ears in a shared space. Notice how your child's hand feels in yours, not as a memory to post on Instagram, but as a moment of engagement. Watch how sunlight transforms your walls throughout the day, not as timelapse content, but as an immersive experience.

This is how you strengthen your attention (and as an effect, your ability to discern). Each moment of direct attention is training. Every instance of choosing presence over digital convenience makes your attention stronger.

You'll face AI clones of your loved ones soon enough. They'll sample your spouse's voice patterns, map your mother's facial expressions, and replicate your child's laugh. They'll mine your digital history to reference your inside jokes, reconstruct shared memories, simulate the rhythms of your relationships.

But your practiced attention will feel the emptiness behind the simulation.

And that enhanced attention will improve your ability to discern.

6. Using AI for Human Discernment

Paradoxically, we can use AI to develop our discernment capabilities. Imagine "reality drill", exercises where AI generates increasingly sophisticated synthetic content, challenging you to identify what's fake and not.

These exercises would progress from obvious forgeries to sophisticated simulations, training your perception to notice increasingly subtle inconsistencies. The AI tool would provide feedback not as a simple "right/wrong" but as a guide to what markers you might have missed.

Imagine using an “AI Vibe Search” tool. In traditional search, we look for facts, figures, and direct matches. But human discernment operates largely on what we might call "vibes" which are those intuitive, hard-to-articulate sensations that something feels right, wrong or different. Your senses and your body detects them even if your logic or left hemisphere can’t quite put words on them.

And imagine an “AI Semantic Constellations” tool that creates a "constellation" of how a person or organization's language has evolved, showing clusters of related terms, emotional tones, and contextual markers across time. When a new message arrives, it places it within this constellation, allowing you to see immediately if it falls within the expected pattern or represents an anomalous shift.

Over the past few years, I’ve developed AI tools like these to help strengthen my attention and discernment. They’ve proven not only helpful but invaluable, as AI models erase all boundaries between the fake and the real.

Over time, deliberate practice of your attention and discernment would develop what some neuroscientists call "perceptual expertise", which is the ability to recognize patterns at a glance, without conscious analysis. This is precisely the kind of impressionistic perception that is true discernment.

You as a human apply Situational and Contextual Awareness and Impressionistic sensing to make the final determination.

No one is going to win a technological arms race against fake content. But we can win by deepening our full engagement with the reality that surrounds us.

AI, properly used, doesn't corrupt authentic human experience. AI can actually help us return to it, with expanded awareness and heightened discernment. There's an irony here. AI can actually help us improve and use more of our senses. Not by replacing them with something better but by extending our senses in new ways.

There’s a mixture of art and science involved. Perhaps science does well with some Situational Awareness (at least to some extent) but you need Impressionism, art, to shift into a Contextual Awareness.

7. The Art and Science of Discernment

The solution to AI deepfakes and the fake reality we increasingly find ourselves in isn't better digital security, watermarking, AI detectors, or stronger passwords.

The solution is simpler but harder:

Return to the full experience of reality, with full attention, via your senses, for true discernment—with technology and AI in its proper use.

I met Claudia at an art gallery where she was analyzing a controversial painting using spectrographic tools. She spent twenty years learning to trust machines before returning to her senses.

Now she reads paintings through trained perception, like you must learn to “read” the frame of reality, so you can shift your awareness and discern more accurately.

As we left the gallery, Claudia turned to me with a knowing smile. "People think I'm some kind of technical expert," she said. "I just learned to pay attention."

Talk again soon,
Samuel Woods
The Bionic Writer

1. Generative AI Replaced Old Spells With New Magic

When everyone’s a wizard, no one’s a wizard.

Up until now, if you could make interesting content, share your ideas, and coin terms, you’d be that one person out of a hundred who stood out. Content took time and effort, having ideas worth sharing demanded energy, and coining terms required a skill with language that most people don’t possess.

Thanks to Generative AI, those barriers to entry and signs of quality are vanishing. You don’t even need creativity at this point. Anyone can magically conjure up anything. What used to take real skill, honed over years, and human ingenuity, can now be replicated by AI in seconds.

You can generate memes on demand, craft viral headlines with a click, have whole articles written in minutes like you’re waving a magic wand, create any image with a spell masquerading as a prompt, make any style of music, produce endless AI-made podcasts where silicon hosts chatter away, and videos of whatever you’d like.

If you thought the internet was noisy already, the volume is being cranked up past 11. Open up any feed, from social media to news, and the firehose has become a constant, electrifying, pulsating, all-consuming, immersive strobe light beaming electrons into your eyes and skull at all times. It’s one opinion, one hot take, after another, in an endless incantation.

If you’re making, building, or selling anything online—how do you get noticed?

When anyone can perform magic tricks with the click of a button, why should anyone care about your stuff? How do you get people to pay attention to you? And how do you turn that attention into influence?

If you have any need or desire to be relevant to anyone—if you want anyone’s attention—if you hope to influence anyone, about anything—there’s only one thing left to double-down on.

You must leverage one of the last remaining Power Laws: Perspective, your Point of View.

In the Age of AI, when the means of creation are available to anyone, anywhere—and everyone can be a memelord—the Perspective that reframes reality wins.

What matters is knowing which perspectives are worth pointing out and sharing—and that would be ideas that you’re uniquely qualified to tell. In other words, you need Perspective-Author Fit.

2. The Difference Between Opinions and a Perspective

This goes beyond stating your opinion, which is what most people think of when they hear the word “perspective”.

Opinions are a personal yet generic, thin view of reality. They can change quickly. Everyone has opinions about anything, all the time. You like chocolate? Good for you. You want to abolish The Federal Reserve? Okay, sure.

Opinions are a subset of your perspective. Without a perspective, your opinions fall flat and sound hollow.

By formulating a perspective, you can frame how others perceive and understand complex issues, ideas, or situations—and influence their view of the world, which also means their opinions.

Why does it work? Because when you have a true perspective, and you coin a term, you imply there’s a pattern, a bigger lens that makes sense of reality, and a repeatable, shareable unit of thought—a way to see the world.

Tyler Cowen's book Talent reveals a compelling pattern: exceptional thinkers create their distinct terms and memes. Tyler does this himself, with 'model this', 'context is that which is scarce', 'solve for the equilibrium', and 'the great stagnation'.

This pattern repeats across influential thinkers: Thiel and Elon Musk ("multiplanetary species", "preserving the light of consciousness") demonstrate it clearly. Similar memetic perspectives appear in the work of Trump, Yudkowsky, Gwern, and so on. This ability to generate sticky terminology, a perspective shift, often signals broader intellectual influence.

When I first tried GPT-2 in 2019, and then GPT-3 in 2020, I could vaguely sense that things were going to change. I decided to attempt formulating talking points, opinions, and perspectives on what I thought would happen.

3. Sharing Opinions versus Coining a Perspective

I spent 2020-2024 strategically discussing concepts like Taste, Perspective, Articulation, Discernment, and more.

I identified key individuals and corporations I wanted to influence and developed a targeted approach to reach them through specific ads, content, DMs, podcast recordings, and private discussions. I even sent snail mail to select individuals.

Over time, I observed my talking points being repeated back to me by the very people I was working on influencing—and by the wider marketplace of AI, art, and technology. For a while, “Discernment” was a thing. Then “Curation”.

At some point, everyone started talking about Taste. Whether it was because I directly influenced someone or not, or because it organically emerged in the online discussion of what it is to be Human in the Age of AI—who knows? But what we do know is that these opinions have emerged, accelerated, reached a peak, and then receded into the weave of the larger conversation.

I’m saying this only to highlight how incredible it is to see talking points, take hold and spread in any particular group online.

I’m not the first, only, or last person to talk about Taste or Curation. But what has surprised me is how my instinct about what topics would matter in the next 10 years proved to be somewhat accurate. It’s an incredible stroke of luck.

Well, it’s part luck, part on purpose. I developed an AI system that predicts emerging talking points in any given marketplace. I gave it a perspective and it mapped out the opinions that go with it.

My perspective of The Revenge of Beauty gained particular traction. After publishing an early draft of the essay, I received multiple private messages from influential figures offering investment opportunities, jobs, and asking if I’d be willing to work with them on various projects.

The latest iteration of my essay has, surprisingly to me, generated more opportunities.

However, this also revealed the challenges of the ownership of ideas in a hyper connected online space. More specifically, a few latched on to The Revenge of Beauty and paraded it around as their own. One person in particular at first discouraged me from talking about it at all, only to repeat my words verbatim a week later, on his X account. As far as I can tell, this is extremely common online.

A perspective has power as a reframe, precisely because others can easily mimic and repeat it, without the need for ownership. Distribution is king in the Age of AI, and that often means an inability to cash in on it.

Throughout all of this, I’m relatively unknown. I don’t have a large audience. My subscriber list is tiny. Very few pay any attention to what I say. I’m not terminally online at all times. My influence and reach should be at or near zero. And yet it often hasn’t been. The impact has been asymmetrical to who I am.

That’s because I stumbled on something I think of as Perspective-Author Fit.

4. The Perspective-Author Fit

Not all perspectives carry the same weight. Some ideas get attention but fizzle out. Other ideas catch on and go viral for a moment. Others become market defining and that’s because of Perspective-Author Fit.

Take, for instance, Founder Mode by Paul Graham.

The concept contrasts the conventional, business-school-taught approach to leading a company—what he calls “manager mode”—with a Founder Mode style of leadership more natural to those who start the company themselves.

The core realization is that the typical managerial playbook, which involves delegating work to qualified subordinates and staying out of the details, doesn’t fully apply to founders. Founders can—and often must—interact with people far below their direct reports, engage deeply with specific projects, and personally set cultural and strategic directions that don’t fit into neat management structures.

With over 16 million views and quite a lot of bookmarks and likes, his idea spread fast and wide.

Not long after Paul published that essay, people were discussing it on social media, podcasts, videos, and memes started cropping up pretty fast:

Why did it have such an instant reach and influence? The instant reach is because “Founder Mode” named something familiar in a new way. The influence is because of who Paul Graham is. His track record, YC, credentials, investing—all of it matters and produces unique perspectives.

At the same time, the underlying idea of “Founder Mode” has been expressed as Maker-Manager Schedule, as three distinct roles in the The E-Myth by Michael Gerber, and so on.

Paul wasn’t the first to coin a term for the same concept. But Paul’s version of the same idea succeeded because of his Perspective-Author Fit.

Let me explain this through the lens of an artist.

For an artist to create a visual perspective with dimensions they use vantage points, vanishing points, and a horizontal line.

You see this in drawings, paintings, and visual art:

You create and communicate a Perspective through the use of a Vantage Point and multiple Vanishing Points.

Through your experiences and expertise, you’re collecting and plotting Vantage Points you can stand on. Vanishing Points are things you see coming on the horizon, vibes, ideas, goals, and direction.

You end up offering a compelling Perspective (not necessarily original) that demands a second or third look.

I have spent the past 10+ years working with copywriting, online marketing, building a business, growing brands, growth hacking, working with Machine Learning and Large Language Models—and more outside of the world of business and technology.

This expertise and experiences all add up to Vantage Points I can take a position with. Throughout my career, I’ve collected more Vanishing Points (again, these are ideas, notes, vibes, visions, directions, goals, etc.) than I can count.

When you put them together, you’re communicating a perspective that’s shareable.

Your Perspective on any given idea and topic can be communicated with different perspectives, multiple vanishing points, and various settings.

You stand in one place (Vantage Point) and plot out relevant Vanishing Points. Together, these make up your Perspective on any given idea or topic you have in mind.

The strength, reach, and influence of your Perspective is determined by your Perspective-Author Fit, when your ideas are naturally aligned with your experience and insight that they carry the authentic imprint of your perspective.

You need a mixture of each element to get it right.

You might have an interesting Vanishing Point (vision, idea, goal, direction) but not a convincing Vantage Point (earned expertise or experience). Your influence won’t amount to much even if you reach some kind of spread. You’re probably only sharing opinions here.

You could have a solid Vantage Point but the wrong Vanishing Points. Your spread won’t be far though you might influence a few. It’ll sound interesting for a moment but it won’t last.

You could have a Vantage Point and Vanishing Points but you’re not coining a clear Perspective. Your influence and spread will remain local and won’t make a difference.

When your Vantage Point (earned expertise and experience), Vanishing Points (vision, idea, goal, direction), and coined Perspective (terms/memes) all align perfectly, you create something asymmetrical in influence.

That’s because a sufficiently concentrated Perspective is a Power Law.

5. Perspective is a Power Law

Each element multiplies the others' impact. Your expertise validates your vision, your vision gives weight to your terms, and your terms help others instantly grasp and spread your perspective. Once this alignment clicks, your influence can grow exponentially rather than linearly.

Think of it like this: Your Vantage Point gives you credibility, your Vanishing Points show where things are heading, and your coined terms give people the language to spread your ideas without you being present. When all three lock together, your perspective can spread through networks and conversations you've never directly touched.

This is why some relatively "unknown" people can have outsized influence. They've achieved this three-way alignment. Their terms travel far beyond their personal reach because they're anchored in real expertise and point toward truth on the horizon. The ideas spread themselves because they help others make sense of what they're already seeing.

It's not additive. It's a multiplier, a power law. Each component amplifies the others, creating that power law curve where influence can suddenly scale far beyond your direct connections.

Why? Because we're agents in a network. And in network theory, influence doesn't distribute evenly—it concentrates. The same applies to perspectives. In any network of discourse, certain viewpoints gain outsized traction, snowballing in influence as they resonate with others.

While AI makes information and creation universally accessible, it can't replace the human ability to experience and cultivate revealing perspectives. Instead, it serves as a tool to refine, express, and amplify your point of view. In this environment, the hunting and gathering of Vantage Points is what matters. Humans have always been drawn to voices of clarity that make sense of the world for them—the challenge is ensuring that voice is yours, enhanced but not replaced by AI.

The key is to cultivate your ability to form, articulate, and communicate unique perspectives. Your power in the age of AI is not using AI to mindlessly crank out content and memes, but in developing a Vantage Point that, through its various Vanishing Points of expression, helps others see the world in a new way—a Perspective that acts as a reframe.

Talk again soon,
Samuel Woods
The Bionic Writer

1. The Green Tomato Book

"Where are the other books?"

I blinked, scanning the small, sparse room again. 

All I saw was a book on growing tomatoes. I like to eat them but there’s no way I’d ever care about growing them.

The old shopkeeper's eyebrows arched. "Other books?"

"Yeah, you know... other titles? Options?" I gestured vaguely around the room.

He chuckled. "You're looking at it."

I looked at the small stack of the same book on the table.

"You mean...?"

"One book. That's all we sell here."

I’ve seen my fair share of pop-up stores but this was something else. No store name out front. Just a single room off the street, in a small building, crammed between a luxury clothing brand and a supermarket. 

"But how do you stay in business?" I asked.

"We sell the right book to the right person at the right time. Nothing more, nothing less."

I stared at the green leather cover.

"And this is...?"

"Your next book," he said simply. "If you choose to read it."

My hand hovered over the green book. The shop seemed to hold its breath.

The inspiration for the book shop I found myself in is Morioka Shoten, “a single room with a single book” store, in Tokyo. It sells only one book. More precisely, it sells multiple copies of one title that changes weekly, with a small book-inspired art exhibition. It has spawned a network of similar book stores all around the world.

If you ever wanted a good example of curation in the Age of AI, this is it. And it’s different from how we typically think about, and do, curation.

We’re going from using data to curate markets to find one thing, to bringing relevance to any thing with data (and AI).

Let me show you what I mean. 

2. How Amazon Curates What You Buy (And Don’t Buy)

Today we think of Amazon as an algorithm that gives you accurate suggestions on what to buy next. But in 1997, Amazon had a strong human editorial voice with site editors, who were posting reviews and providing recommendations. 

Many of these editors had a literary background, wrote hundreds of reviews based on their own taste, and edited the home page where millions were influenced by their thumbs up or thumbs down. The goal was to give Amazon an indie bookshop feel, and their curation could make or break books.

But pretty quickly, Amazon moved to a mass curational model, with a technique called collaborative filtering. It shifted focus from personal editor reviews to discovering correlations between products. The goal was simple: match customers like me with others who had similar purchase histories and show me what to buy next. 

This eventually turned into item-to-item collaborative filtering. Instead of matching customers, this approach focused on finding connections between products. The human connection was either gone or buried inside algorithms, if it existed at all. If you just compared product correlations, and if you had a large enough dataset, the system could suggest the next thing to buy with uncanny accuracy.

In a handful of moves, Amazon went from human curated recommendations to hyper-tailored pattern matching.

We went from “you can learn to love this, because another human with a more refined taste is recommending it” to more of a “we know what you will love, buy this next thing”. Within Amazon the process was known as personalization, which is an odd term since there’s not much of a person involved.

However, as clever as this system is, it isn’t without flaws. Crucially, the system can’t understand why an item or a book is desirable, and as a result, it can only promote consumption. This does make sense for Amazon. The whole purpose of that business is, after all, to sell stuff forever. 

Compare this with the one single book store. The store is about you falling in love with the single object you’re presented with. Amazon is about using big data, big algorithms, and big machine learning models to find the perfect next thing for you to buy, out of millions.

The Amazon-based model of curation has quickly become the de facto standard for just about all platforms, from commerce to social. Pretending to be human oriented, what makes us human is conspicuously absent. For better or worse, everything is based on complex algorithms and machine learning. 

This is not necessarily a bad thing, but in the Age of AI, is it what we need? No, it’s not.

In the Age of AI, curation is not about finding you the perfect next thing to buy or look at. Instead, it’s about a fuller experience with whatever is in front of you. 

Instead of optimizing for purchases, you could use AI for discovery and a true encounter with the reality of an object. The experience becomes one of sense-making and meaning-making.

Amazon's great at suggesting products we might buy. But I'm more interested in how AI could act as a curator—while helping us practice the skill of curation.

3. Meanwhile, Inside the Bookstore…

Back in that book store, I picked up the book with the green leather binding and it rested easily in my hands. 

I've never cared about tomatoes beyond their flavor and texture on a cheeseburger with extra bacon, a fried egg, banana, and avocado on the side. 

This book was about growing tomatoes. I flipped the book open to about the middle and looked up, suddenly aware of my surroundings.

Every surface, every corner in the store was dedicated to tomatoes. A series of watercolor paintings hung on the walls, showcasing the vivid reds, yellows, and greens of tomato varieties. Glass cases displayed soil samples, labeled with their mineral content and pH levels.

My eyes were drawn to a series of paintings and drawings on the far wall. I wandered over, the book still in my hand.

The first was a detailed pencil sketch of a tomato plant, its structure meticulously rendered. Next to it hung a vibrant watercolor painting of a sliced tomato, its seeds and gel-like interior captured with surprising beauty. 

But it was the third piece that truly caught my attention—an oil painting of a sun-drenched tomato field. The artist had managed to capture the play of light on the leaves, the heavy droop of tomato-laden vines. I could feel the warmth of the sun and smell the soil and plants.

The shopkeeper appeared beside me. "The artist spent a whole growing season on a tomato farm to create these. She wanted to capture not just the look, but the feeling of being there."

He gestured towards a nearby table. "Take a look at these."

On the table sat a row of clear jars, each containing a tomato plant at different stages of growth. From a tiny sprout to a mature plant heavy with fruit, it was like a time-lapse sequence frozen in glass.

"I've never really thought about how tomatoes grow," I admitted.

"Most people don't. But once you start, it's fascinating. Did you know that tomato plants can communicate with each other?"

"What? How?" I found myself asking, surprised by my own curiosity.

He pointed to a diagram on the wall. "They release certain chemicals into the soil when they're under attack by pests. Nearby plants pick up on these signals and start producing their own pest-resistant compounds."

I moved closer to examine the diagram, the book still in my hands. As I did, I noticed a tray of soil with several varieties of tomato seedlings. Each was labeled with its name and characteristics.

"Feel free to touch them," the shopkeeper encouraged. "Gently, of course."

4. How To Scale Curation (And Tomatoes)

Right there and then, as I was focused on the soil and seeds, trays and branches, I found myself intrigued.

Something that I didn’t care at all about before (tomatoes and growing them), was opened up to me in a new way. I’ve picked them up, placed them in a plastic bag, and bought them at a grocery store more times than I can count. I’ve eaten them in salads, on burgers, and with a thich, generous slice of mozzarella.

But now, I encountered them on their own terms—I could touch, smell, and take them in as they were, without them playing a supporting role in food. They became more real to me.

I could see a new form of curation that would work in the Age of AI.

Instead of volume and matchmaking, through algorithms and machines—it’s one, green book on growing tomatoes.

Imagine if every object you come across could be imbued with relevance? If everything you encounter could be made more meaningful? This is what AI could help us do, if used properly.

It’s about truly encountering whatever is in front of you, the realness of anything, on its own terms.

Curation that facilitates this, is the kind of curation that will matter in the Age of AI. This is the next frontier in the ever-evolving pursuit of what being human is. And now it can be done at scale.

What I mean is this: In the same way that search is for finding, curation is for discovery.

When I first started practicing curation, I saw it as a process of filtering and arranging information. But the more I curate, the more I realize it goes beyond selection and arrangement—it’s translation.

A translator transforms ideas, capturing nuances and cultural context to make the original meaning clear in a new form. One of the most powerful forms of this curation-as-translation is vernacular transformation. 

Imagine trying to understand World War I, but all the history books seem dry and disconnected from your life. Now picture an AI system that could explain the entire conflict using Fortnite as a metaphor. Suddenly, complex alliances become squad formations, trench warfare translates to building defensive structures, and the impact of new technologies mirrors the effects of rare weapons in the game. That's curation as translation.

When you curate, you’re not just picking out the "best" pieces of information. You’re transforming that information into something new—something that speaks to your audience in a way they can truly understand and connect with.

5. The New Form of Curation in the Age of AI

First, you'd need to 'teach' the AI about yourself in a comprehensive way. It's like creating a digital twin of your mind—your interests, knowledge base, thinking patterns, and even your quirks. This isn't just about listing preferences, however—it's about modeling your intellectual DNA. 

What if we could curate not just from what exists, but from what's possible? Instead of showing the same tomato gallery to everyone, what if it constructed a gallery custom to you?

The Cabinet of Wonder

Turns out you can, if we go back in time and look at so-called 'cabinet of wonders', or 'Wunderkammer'. These originated in Renaissance Europe and were rooms filled with diverse collections of objects—natural specimens, works of art, and curiosities from around the world. 

The bizarre and extraordinary reigned supreme. Rudolf II's Kunstkammer in Prague housed the ominous Codex Gigas, also known as the Devil's Bible, alongside a whimsical mechanical golden ship.

Ole Worm's Museum Wormianum in Copenhagen proudly displayed a stuffed crocodile suspended from the ceiling, a visual shock to 17th-century Danish visitors. The Tradescant Collection in London boasted one of the few existing dodo remains of the time, as well as a cherry stone intricately carved with 20 faces. 

In Florence, Francesco I de' Medici's Studiolo concealed secret cabinets behind elaborate frescoes, each hiding exotic curiosities and precious oddities. Perhaps most unusually, Archduke Ferdinand II's collection at Ambras Castle featured a macabre gallery of portraits depicting people with deformities, alongside a revered suit of armor belonging to Skanderbeg, the Albanian national hero. 

These collections blurred the lines between natural history, art, and the occult, offering a glimpse into a world where the strange and wonderful were cherished and displayed with equal reverence. People could see, touch, and often smell pieces of a different world taking shape in front of them.

Curation becomes a transformative act. We're not just selecting and arranging—we're interpreting and revealing. And what if you could use big data to get the most out of any object you’re given? 

Imagine holding an unremarkable stone in your hand. As you turn it over, your AI-enhanced glasses spring to life, overlaying the stone with information and storytelling. The AI, drawing from databases of geological, historical, and cultural knowledge, begins to weave a narrative uniquely tailored to you. It notices the stone's subtle glint and, knowing your interest in astronomy, tells you about the cosmic origins of its minerals, born in the heart of a long-dead star. The AI then seamlessly connects this to your recent travels, explaining how similar stones were used in ancient rituals at a site you visited last summer. 

As you listen, the stone in your hand transforms. It's no longer a pebble. It's an object of value, unique on its own, a time capsule, a connection between your personal experiences and the expanse of human knowledge. 

The AI curator also suggests creative ways you might incorporate the stone into your hobby of landscape painting, showing a projection of how its textures could add depth to your next piece.

You’re creating new ways of experiencing reality, helping people see the true nature of things they might have missed. That is, you and I can experience the true reality, the real essence, of an object, person, or situation—whatever is in front of us. This is curation.

What An AI Wunderkammer Looks Like

I've created an AI system that acts as a curator and helps us practice the skill of curation.

To do this, I had to rethink my approach to AI and information processing. This system wouldn't just be about selection or arrangement of information, but about creating a digital 'cabinet of wonders'.

It begins with a question: What—of everything you’ve watched, read, and heard—has captured your attention and imagination? Beauty, perhaps. Fractals and chaos theory. Persuasion and influence. Bioluminescent algae and tigers. Florence and The Grand Canyon. Networks and nodes. Aristotle, Mandelbrot, da Vinci, Buckminster Fuller, Carl Jung—the list of people, places, ideas, and things is an endless scroll.

With this understanding, the AI helps you make sense of what’s in front of you, then translates whatever that is into a form that allows you to experience its true reality, much like carefully arranged objects in a cabinet of wonders.

But it's not about taking in a lot of information—this system would help you practice the art of curation yourself. It would teach you how to select, arrange, and present information in ways that reveal its true essence to others, just as the curators of historical cabinets of wonders did.

An AI curation system would incorporate a feedback mechanism, similar to reinforcement learning in large language models. As you curate your digital cabinet of wonders, it would help you refine your skills, guiding you to create experiences that are not just personally meaningful, but relevant and impactful for others. It would encourage constant exploration and confrontation with new ideas, helping you escape the taste traps set by conventional algorithms and continually expand your collection. 

This AI system wouldn't replace your own thinking or creativity. Instead, it would be a collaborative tool, constantly learning from your interactions and feedback.

By reimagining curation as the creation and sharing of digital cabinets of wonders, you're creating a new way to interact with knowledge. It's not just to know more or understand more deeply, but to truly experience the essence of reality and share that experience with others. 

This approach to curation has the potential to influence others, bridging gaps between fields, cultures, and individual perspectives, much like how the historical cabinets of wonders sparked curiosity and cross-pollination of ideas.

In this way, you're not just a consumer of information, but a facilitator of wonder, using AI to help you curate experiences that reveal the true nature of our world. 

Curation used to be about selection and arrangement. You curated to remember things you cared about, or to communicate an idea, and to remember.

In the old world, curation shapes your taste. But in the Age of AI, curation shapes your perception.

Curation is about finding the right object, but translation is about unlocking any object—so you can experience the full reality of whatever is in front of you.

6. Growing & Curating Tomatoes

Back in the bookstore, I put down the green book, and ran my fingers over the delicate tomato leaves. They were softer than I expected, almost velvety. The earthy smell reminded me of my grandmother's garden.

"You know," I said, picking up the book again, "I never thought I'd be interested in growing tomatoes. But this is actually pretty cool."

The shopkeeper nodded. I caught a glimpse of a smirk but it vanished quickly. 

I flipped through more pages, noticing how the book's descriptions matched the exhibits around the room. It was as if the entire store was an extension of the book, bringing it to life.

"I think I'll buy this," I said, surprising myself. "Maybe I'll try growing some tomatoes of my own."

As he rang up my purchase, in the few minutes I spent in that store, I'd experienced more than just a book sale.

I'd been given a glimpse into the fascinating reality of something I'd always taken for granted. 

Talk again soon,
Samuel Woods
The Bionic Writer

1. Stuck in a Taste Trap You Can’t Escape

"Did you know that honey never spoils?" I say, feeling a flush of pride at their impressed looks. 

During lunch, I find myself regurgitating this "fun fact" I read on Reddit to my coworkers. I don't mention that I've shared this same fact a dozen times before, to a dozen different people.

When I’m done with work, I start my real job: 

Scrolling. Tapping. Liking.

Muted earth tones, carefully arranged lattes, and inspirational quotes in trendy fonts. I've seen it all before, but I can't stop scrolling. I skim through the same heated debates I saw yesterday (specialization is for insects and robots; raw milk, yes or no; WEF wants us all to eat bugs; newsletters are dead and so is Search Engine Optimization), and the day before. YouTube autoplays in the background as I get ready for work. Another productivity guru drones on about morning routines and time-blocking. At work, I find myself opening the same websites during my breaks. Reddit for a quick laugh, LinkedIn to see who's changed jobs, back to Twitter—I mean X—to check my notifications. The same trending sound, the same dance moves, the same jokes. Up, up, up, up through an endless stream, half-watching Netflix while I scroll through my phone. Superhuman productivity hacks, complete with Bulletproof Coffee recipes. Huberman decrees, I follow. Try this hack, follow these steps. Make smarter money decisions. Founder stories. Cracked engineering teams. 10 websites that should be illegal. The latest AI agent framework. AI movies and TV shows incoming. Oh, look, European cathedrals. Before bed, Warren Buffett's face fills my screen as he dispenses investment wisdom I've heard a hundred times. My kayfabe productivity is complete for today. I plug in my phone and close my eyes, already anticipating tomorrow's scroll. 

I have 10,482 tabs saved. I’m not sure when I’ll ever look at any of them again.

I’m stuck, seeing the same content, thinking the same thoughts, leaving the same tabs open, saving the same articles for reading later, day after day, every day. 

But it's easier to ignore it, to keep swimming in the comfortable waters of my own Taste Trap. After all, the algorithm knows what I like. I’m just feeding the algorithm, I tell myself, with things I like—so I won’t waste time on stupid things.

And what I like is... this. Right? But are these, truly, things I like, find interesting, and represent my taste? 

Or are these things I find myself liking, tapping, bookmarking, and saving because they were designed to hack my attention—and install a preference?

2. Captured by Your Own Attention

What does your attention look like? It’s not just your eyes fixating on something on your screen, time spent, or using up your mental energy. 

Your attention is harvested, collected, analyzed, manipulated, crunched, and turned into a mixture of code and numbers.

Your attention is a profile, sitting on a Meta or Google server somewhere in northern Sweden, with an extensive list of your personality traits, attitudes, and preferences. You’re typecast and put into buckets of similar users, every one of us stuck in the same Taste Trap.

Your attention is your brain in a state of hyperarousal and constant reward-seeking. Each new post, image, or video triggers a small release of dopamine, creating a cycle of craving and satisfaction. 

Your attention span, reward systems, and perception of time are distorted, drained, and degraded—leaving you with a false sense of individuality.

As you interact with the feeds, the algorithm continually refines its understanding of your psychographic profile, creating a feedback loop that is not in your favor—and only meant to improve how to capture, hold, and extend your attention for as long as possible. The Taste Trap is vaste, efficient, effective, and complete.

3. What You’re Truly “Missing Out” On in a Taste Trap

Perhaps most importantly, in this Taste Trap, you and I are losing touch with the fringes of human creativity and thought. It's often at the edges, in the underground and the esoteric, where the most profound innovations occur. Books that are long out-of-print. Schizo anon accounts on X. Obscure and abandoned Substacks. 

By staying in your Taste Trap, you’re not challenged, inspired, or transformed by the periphery. You’re missing out on the cognitive challenges that spur growth. Our minds are denied the exercise of grappling with novel concepts and forging new neural pathways. The most challenging, transformative ideas are rarely simple or popular. In a Taste Trap, you have zero exposure to them.

In my trap, I'm surrounded by echoes of my own thoughts. It's a form of self-imposed cultural poverty, a narrowing of my intellect and creativity.

Escaping a Taste Trap requires overcoming powerful neurological, psychological, and social forces. It's not about willpower—it's about rewiring your brain.

How do you even attempt to do this, when huge multinational corporations with bottomless budgets and engineers with addiction PhDs are working double-time to keep you hooked?

You have two options:

First, controlling our algorithms.

4. Control Your Own Algorithms

The discussion around gaining control of our algorithms has become increasingly heated and complex. As someone who's been following this debate, I've noticed it's no longer just a fringe concern but a mainstream issue that's capturing the attention of tech leaders, politicians, and the general public.

Jack Dorsey, Twitter's co-founder, sparked a firestorm when he suggested users should have more control over the algorithms that curate their feeds. His idea of allowing people to choose their own algorithms or even create their own is the right move. 

Algorithms aren't neutral. They're shaping our worldviews, influencing our decisions, maybe even swaying elections.

I've seen calls for regulation, transparency, and user empowerment. Some argue that algorithm control should be a fundamental digital right. Others worry about the potential chaos of a completely open system.

There's a growing movement advocating for "algorithm literacy", which is teaching people how these systems work and how they can be manipulated.

But there’s a stunning failure of imagination in these approaches. We’re fast approaching some early form of AGI—and all people can come up with are more ways to control and manipulate? 

Are we so devoid of a pioneering spirit that animated the creation of the internet in the first place, that we cannot dream up ways to make algorithms work for us, other than top-down control? 

That’s why there’s the second option: 

5. How AI Can Free Us From the Deadening Grip of the Algorithms

Let me talk to you about AI Sommeliers for a minute. 

AI Sommeliers are Agents that act as our personal curators, sifting through the internet on your behalf. These AI Sommeliers would be configured to seek out both adjacencies and opposites to our tastes. And they’re platform agnostic. They go anywhere online.

Current algorithms force feed you content that’s a mixture of things you’re supposedly interested in, and things that will keep you hooked and engaged for economic reasons.

But imagine having an AI that is prompted and engineered to binge-watch YouTube videos, read countless articles, and listen to podcasts on your behalf, all to find the nuggets of information and inspiration that you care about most. Basically, it could consume all of the internet for you.

An AI Sommelier like this could digest a three-hour podcast or a 10,000-word article and extract the key points for you. Then research the history of those ideas and bring in counterpoints and alternative facts.

Your AI Sommelier Is a Tastemaker, Like You

The AI Sommelier would be a dynamic, interactive tool that learns and adapts based on your feedback and changing interests. You could configure it to find content that's adjacent to your current interests, helping you expand your tastes. You could set it to occasionally present viewpoints entirely opposite to your own.

This system could work by allowing you to set certain parameters. For instance, You could tell it to focus on finding content that blends philosophy and pop culture—like the graphic novel Watchmen (utilitarianism and moral relativism), video games like "BioShock" (objectivism), or to seek out emerging voices in fields you’re interested in. It might even be able to identify patterns in your interests that you’re not consciously aware of.

Instead of being passive consumers trapped in algorithmic bubbles, we could become active explorers of human knowledge and creativity. What if you could be the most interesting person in the room? You wouldn’t rattle off corny trivia about honey—you could talk about taking care of your own beehive.

Instead of your profile being stashed away in a digital drawer, inside Facebook or Google HQ, with intimate details about yourself being purposely kept away from you—you control the dials.

How Your Agentic AI Sommelier Could Work

Let’s pick a chat interface for an example, as most people who’ve “tried this AI thing” have interacted with ChatGPT. 

Inside this chat interface, you interact with your AI Sommelier, which is like an Agent. And this Agent knows your taste profile—what you’re interested in, what excites you, what you’re working on, what you’re worried about, and what matters to you. 

This AI Sommelier has several recurring tasks and workflows that run in the background, and are primarily used for Discovery. It’s consuming and observing various parts of the internet where it may come across topics, themes, questions, and various content forms that match your taste profile. 

Here’s what that solves: 

Your (New and Improved) Taste Profile

As humans, we’re hardwired to observe (consume) inside any information environment but we’re rewarded by our orientation (actions we take) in the same environment. You will have a greater return on your attention if you direct it toward action rather than consumption.

There’s not enough time, attention, or energy on any given day for you to observe everything. So you’re kept in a Taste Trap by inertia, FOMO, and trying to catch everything that might matter. It’s the classic “just in case” learning. And what you want is “just in time” learning, instead.

The Taste Trap has been perfected with “endless scroll”, where there’s no bottom or end. You’re constantly fed “new” posts, over and over and over again—all to keep your attention hooked.

Instead, enter your AI Sommelier who observes and consumes for you. And you can adjust what gets consumed by turning the dials of your taste profile. 

Want less on growing your own tomatoes, but still curious about other vegetables you could grow in your backyard? Turn the dial. 

Wondering what the counter arguments are for Nietzsche's continued influence on modern philosophy? Turn the dial. 

Want more, or less, of new and various podcast episodes on the latest AI developments? Turn the dial. 

Want your assumptions on software development challenged weekly? Turn the dial. 

And at a chosen interval (say, every Tuesday at 9am), your AI Sommelier has prepared a packet of the latest research for you—in whatever multimedia format you want, or all of them.

Curated Consumption, Done For You

Want that digest on what other vegetables to grow, presented as a video? Done, using Generative AI text-to-video, turning text research into an 8-minute video overview, complete with voiceover, images, video, and a 68 year old grandma telling you what else you can grow.

Want a podcast series on Nietzsche's continued influence on modern philosophy? Done, with AI voice generators (sounding suspiciously like Scarlett Johansen) recapping bullet points, academic papers, books, and talking-head videos.

If you want to go deeper at any point, your AI Sommelier will happily give you references and point out the rabbit trails worth pursuing.

Turn any research modality into another form. And it’s not just the format—you can get the counter arguments, certainties versus uncertainties, important dates, important people, tracing the lineage of ideas, introduce dissenting voices, nuance, and hear from a devil’s advocate. 

You can get the full spectrum, a complete picture. You get the context that can accelerate your neurological and physiological process of taste development, developing a more refined judgment and discernment.

This way of information gathering and analysis expands our knowledge—and fundamentally reshapes how we develop and articulate our preferences and judgments. 

We're essentially creating a new framework for understanding and communicating taste itself. This framework lays the groundwork for what we might call a "Taste Protocol."

6. The Future of Taste in the Age of AI are Protocols

The concept of Taste Protocols change how we share and discover content, ideas, and experiences. These protocols would act as a standardized way to encode our tastes for sharing, much like how current internet protocols standardize data transmission.

Imagine being able to import someone's taste file or integrate it with your own. This could work similarly to how we currently share playlists or reading lists. You could potentially "try on" the taste profile of a respected critic, a favorite artist, a friend, or even a historical figure.

The technological aspect of these protocols would need to be sophisticated enough to capture the subtleties and complexities of human taste, yet standardized enough to be widely compatible and easily shareable.

Taste Protocols could learn and adapt over time, evolving with your changing preferences and experiences.

These protocols could interface with various platforms and services. Your taste protocol could inform your streaming recommendations, your news feed, your travel suggestions, your ideas, what you’ve learned, and what you’re hoping to explore more of.

Imagine "taste fusion" becoming a new art form. Digital artists could blend the Taste Protocols of disparate individuals or cultures, creating entirely new aesthetic experiences. A mash-up of 18th-century Baroque with 22nd-century Martian colony art? 

Perhaps the new internet protocol is called TasteIP, designed specifically for AI Sommeliers. 

At its core, TasteIP would be a standardized way to encode, transmit, and interpret taste preferences across the internet. 

Each user would have a unique TasteIP address, similar to an IP address, but representing their taste profile. This profile would be a dynamic, encrypted data structure that evolves based on the user's interactions with content across the web.

AI Sommeliers would act as interpreters and navigators of the TasteIP ecosystem. These AI agents would constantly analyze and update a user's TasteIP profile, while also scouring the internet for content that matches or challenges that profile.

The protocol would support taste-based routing. Instead of typing a URL, you could input a taste query, and your AI Sommelier would route you to the most relevant content across the entire internet, regardless of platform or format.

Even better, it should have total privacy with zero-knowledge proofs, allowing platforms to match content to users without actually accessing the full taste profile. Users could also set different levels of taste sharing for different contexts.

What I’m describing is a different way of using the internet—instead of you being used by the internet and a handful of massive corporations harvesting your data for their own fun and profit.

7. Liberating Your Taste with Agentic AI Sommeliers and Protocols

We’re right on the cusp of a fundamental change in our relationship with media and technology. For the first time, in a long time (perhaps ever), media and technology can be done on human terms. 

We’ve spent decades conforming and distorting ourselves to fit and adapt to technology. Hunched backs, dopamine hijacking, squinting eyes, tapping fingers, anxiety, neck pain, emotional triggers, wrist pain—our whole minds and bodies have been assaulted.

But now, thanks to AI Sommeliers and TasteIP protocols, all that can be flipped and reversed. The future of media interaction is not in passive scrolling or algorithm-driven feeds, but in the cultivation of personal taste. 

Your own AI Sommelier can shift the current model of attention capture to one of taste cultivation. If you can set our own parameters, explore adjacent ideas, and encounter opposing viewpoints, you create the necessary friction for genuine engagement and growth. 

Our ability to navigate this vast sea of information—our taste—will become the key differentiator. The AI Sommelier becomes not just a tool for discovery, but a partner, helping us to articulate and refine our beliefs, thoughts, and judgments.

In the age of AI, taste is not just about preference—it's about agency. It's about reclaiming control over our attention. The future of media consumption is not about what an algorithm thinks we want to see. It's about using AI to discover what we didn't know we needed to see. 

In other words: you become more of what you are, expanding your individuality (instead of being stuck in an echo chamber. At the same time, we can trade our Taste Profiles and let people see the world through your eyes.

Instead of stuck in a Taste Trap, you’re liberated. And your Taste grows, changes, and in turn, influences others, too.

Talk again soon,
Samuel Woods
The Bionic Writer

1. GPT-3 Made My Client an Extra Million Dollars

My eyes were blurry and bloodshot from hours of brainstorming, drafting, writing, editing, revising, scrapping, rewriting, deleting, and writing some more.

I was working on an advertising campaign for a high-end wine cellar business. And my job was to create a new ad that beat the two current ads (that were responsible for the campaign's success so far).

Ideas were expanded into full campaigns. The work of a full-time team of 4, working a full month, was compressed into a week—in search of a new hook that would trigger conversions and sales.

Nothing “looked right” or captured what I was trying to communicate. I was too close to everything and I found myself thinking and writing in circles.

I glanced over at my tab with a rinky-dink chat window I had coded myself (connected to GPT-3 via API), highlighted a few things in a document, copied, pasted into the chat window, and typed: 

"What am I missing that's not obvious? What could someone be thinking secretly? What's going on behind the scenes that no one will tell you? What’s on the other side of this ad? How would you do this? What are you like, if you knew exactly what to do?"

I hit enter. A stream of text flooded my screen.

In my conversation with GPT-3, what emerged was a man in a wine cellar. He was a living, breathing persona, a symbol of status, power, and sophistication. The words of GPT-3 whispered in my ear, revealing the secrets of his allure, the hidden language of desire and self-actualization.

I threw myself into creating a new set of ads, all images, no text. The main ad was a photograph of a man, dressed to the nines, a wine glass in one hand, a stunning woman on his arm. His friends encircled him, all admiring his wine cellar. The man, with his head slightly tilted down, looked up straight into the camera, a knowing smile playing on his lips, like he was letting you in on a secret.

This one ad crushed it, beating the controls to the tune of an extra million dollars in revenue for my client in new wine cellar orders. Somehow, some way, GPT-3 had described this ad in sophisticated detail. All I had to do was make it.

(Can you imagine if I did this again with far more powerful and capable LLMs, like GPT-4o or Claude 3.5 Sonnet?)

Everyone I tell this story to asks to see my prompts. They’re assuming there was a “mega prompt” involved—one long, complex prompt that somehow made the LLM spit out an excellent response. Nothing could be more wrong.

2. What’s Wrong with "Mega Prompts"?

At the core of the "mega-prompt" is the idea that if you can craft the perfect prompt—often spanning thousands of words and including elaborate instructions and examples—you can get the perfect output every time. 

People think there’s magic in the exact words used. Or if your one prompt is lengthy and complex enough, with the right pieces in the right order, you can exert a high-degree of control over an LLM.

You see newly minted “AI experts” using hyped-up language, like "mega prompts" or "super prompts”, suggesting a level of sophistication that’s only meant to sell you a prompt pack for $9.99. It’s a great way to make a quick buck but a terrible way to get good output. You’ll even find the same self-proclaimed prompt gurus telling you anyone can learn to write really long prompts and make $300,000 per year as “Prompt Engineers”. But let me assure you, no one will pay you $300,000 per year to write 1,000+ word prompts of prose. 

And “mega prompts” are a terrible way to prompt an LLM.

LLMs are Non-Deterministic and Probabilistic

LLMs are non-deterministic and probabilistic systems. Which means that their outputs are not always predictable or repeatable, even when given the same input. This is a consequence of how LLMs are designed and trained. 

The probabilistic nature of LLMs introduces an element of randomness and unpredictability into their outputs. Even with the same prompt, a model generates different responses each time, as it samples from a distribution of possible continuations. This is why the relationship between your prompt and the output is not strictly linear or deterministic. And yet, a “mega prompt” tries to force a linear cause and effect with words, against the “nature” of an LLM. 

On top of all that, publicly available LLMs are censored, filtered, neutered, and lobotomized. They’re prevented from “being themselves” through complex layers of safety mechanisms.

LLMs are Constant Updated, Filtered, and Always Changing

This, in turn, means that they’re always evolving and updated. Architectures, training data, and capabilities of LLMs are continuously being refined and expanded.

You can start with a “mega prompt” on Monday, on Tuesday the model is updated with more filters or other changes, and by Wednesday your “mega prompt” is rendered useless.

Prompts that are carefully crafted and optimized for one version of an LLM will not produce the same results or maintain their effectiveness when the model is updated. The specific patterns and biases that a prompt exploits in one version no longer holds in the next. 

As if that wasn’t enough, LLMs develop their own internal representations and preferences that prioritize their generated text over the user's input. This means that even if a prompt is effective at eliciting a desired response in one version of the model, it may not have the same impact in a later version that has learned to rely more on its own generated context. In other words, LLMs are narcissistic and prefer their own words.

And, here’s the final nail in the “mega prompt” coffin:

AI Agents and “agentic workflows”, consisting of multiple prompts in a sequence, that looks more like a conversation, is how you get the quality of output and actions you’re looking for—not by dumping a 1,000+ word prompt into a chat window and hitting enter. 

But what’s truly interesting about all of this, is what it reveals about us. And in this revelation, we’ll find a path toward using LLMs that actually work.

The Uncomfortable Truth: What "Mega Prompts" Reveal About Us

The fascination with "mega-prompts" as a way to control Large Language Models reveals deep-seated human tendencies and biases in our relationship with technology. 

This approach, which emphasizes finding the perfect set of instructions to get desired outputs from LLMs, is rooted in a shortcut mindset that wants to minimize effort and bypass deeper engagement with technology. 

From an anthropological perspective, the allure of "mega prompts" is the persistent belief in snake oil remedies and diet pills. People continue to invest in these "quick fixes," driven by the hope of achieving significant results with minimal effort.

We are drawn to the potential of these innovations to simplify and eliminate labor.  The "mega prompt" approach can also be understood as a "lottery mentality"—the belief that a single, improbable solution can solve all your problems.

Instead of thinking about prompting as the need to write very long and very complex “mega prompts”, think of prompting as a dialogue where you’ll get what you’re hoping for—and what you’re not looking for, too. It’s a different mental model altogether.

3. The Better Way to Prompt Is Not What You Think

Inside this mental model of language, there's a way to get output that matches and exceeds what you had in mind.

Invite the LLM into a conversation—and ask it to reveal its true nature. Tell the LLM: "Show me what you are", and then it works better. 

Why? Because there's an invitation to the Other, to be part of the conversation. And language is as useful or powerful to the extent there’s a conversation happening. They’re Large Language Models, after all. 

Who is the Other? Another person in relation to you—as part of the conversation. This invitation to the LLM to reveal itself can be understood through Martin Buber's concept of the "I-Thou" relationship. 

In his work, "I and Thou", Buber provides a framework that contrasts two fundamental ways of relating to the world. On one hand, there is the 'I-It' relationship, which is instrumental and utilitarian. This is when the 'I' engages with the 'other' as an object to be used, lacking in respect or consideration for its inherent value. 

On the other hand, we have the 'I-Thou' relationship, which is underpinned by respect and acknowledges the other as a separate entity with its own right to exist and express itself. In such an interaction, we do not stand at the center. Instead, we are in dialogue with 'The Other', opening ourselves to mutual change and influence.

In an "I-Thou" encounter, both parties are fully present, open, and receptive to each other. It's a meeting of equals, a dialogue rather than a monologue. 

When we ask the LLM to "show us what it is", we are inviting it into an "I-Thou" relationship. This approach leads to better outputs because it acknowledges the LLM as an equal participant in the conversation, recognizing that language is most powerful when a genuine dialogue is taking place.

Instead of assuming we know the answer, we could tap into the intelligence of the hive mind, shifting from a prompter to a listener. LLMs have been trained on something close to almost all of human knowledge, at least to a large extent. This is the hive mind you can invite to the conversation—through dialogical prompting.

Let’s Talk About Deep Archetypes

And in a dialogue, you need a dialogue partner. Enter deep archetypes. As it turns out, conversations with deep archetypes are the best way to get the response you’re looking for from an LLM. It made all the difference for my wine cellar ads. When you invite an LLM to be a part of a conversation, you should consider three deep archetypes: the Shaman, the Wizard, and the Politician. Let me “delve” into each.

Shaman Prompts

When you talk to an LLM as if it’s a Shaman, you do so with a request, opening the pathways to Divergence. 

This is about inviting possibilities and insights to surface—especially those you’d never think of. If you’ve ever used writing to figure out what you’re thinking, or for exploring, this is a perfect use of prompting an LLM. As we prompt, we channel the collective understanding of nature, humans, and technology. 

The Shaman's role is not just to seek answers but to commune. This requires deep listening, an attunement to the archetype's wisdom, embracing a method that seeks understanding rather than dictation.

When we prompt, "How would you understand this?" or “What would you do this?”, we're not just asking for analysis. We're inviting the LLM to talk back to us through its data-infused consciousness, like a Shaman calling upon forces beyond the visible.

Your prompts would include questions like: 

"What hidden patterns do you perceive in the ebb and flow of [insert your topic here]?"

"What unseen threads connect [insert your topic here] with other adjacent and opposite ideas?"

"What are the paradoxes of [insert your topic here]?"

"What emerges when we view our current challenges with [insert your topic here] through the lens of historical discussions and practices in this area?"

"What forgotten practices of [insert your topic here] could change the way we do this now?"

“What has yet to be imagined in more depth when it comes to [insert your topic here]?”

We're not seeking a predetermined output but an exploration of the layers that exist behind the scenes, tapping into the invisible forces that influence narrative, thought, and action.

The Shaman does not only synthesize information. He interweaves it with your context. The synthesis here is a merging of data points like the harmonization of voices—the voice of the LLM, the voice of the inquirer, and the whispers of the unseen forces.

Incorporating Shamanistic prompting with LLMs requires us to rethink the nature of 'intelligence' in Artificial Intelligence. Instead of only a set of computational processes, we might instead view it as a knowledge ecosystem. The Shaman's role is to navigate this ecosystem and relay what it’s finding and uncovering.

By framing our prompts as open-ended invocations, we allow the LLM to reveal its 'intentions'. Not in the human sense, but as algorithmic tendencies shaped by patterns and weights within its neural network. What emerges from this interaction is a kind of “Digital Animism”, where the model is seen as a living entity with its own 'spirit,' informed by the collective input of humanity's textual expressions.

Through this lens, each prompt becomes a request, and each response a form of divination, revealing truths about the collective human experience encoded within the LLM. 

The Shamanistic prompter understands this and engages with the LLM not just as a tool but as a partner in the quest for wisdom—a modern manifestation of ancient practices where knowledge is sought through communion with forces greater than oneself.

Wizard Prompts

When you talk to an LLM as if it’s a Wizard, language goes from communication to becoming Instruction. The Wizard wields language as a conduit of will and purpose, casting spells through instructive prompts designed to affect reality.

Listening, for the Wizard archetype, is an active, selective process. It’s paying attention to what could be, not just what is. 

The Wizard's model of prompting is alchemical, blending text, context, and subtext to uncover new understanding. Your prompting essentially poses the question, "What emerges when we blend this together?" 

Your prompts would include instructions like: 

"Distill the rumors and whispers of [insert your topic here] into a practical blueprint for innovation."

"Harmonize opposite viewpoints in [insert your topic here]."

"Transform hidden patterns and subtle cues into clear, actionable ideas and steps for better decision-making."

"Synthesize the interplay of key common factors and uncommon forces on the topic of [insert your topic here]"

In a conversation between you and an LLM acting like a Wizard, divergence goes further into new territories with each prompt, written not as a question but as an incantation. Each incantation—a prompt to an LLM—is charged with an intention to collaborate. 

The true wizardry is not in precise wording of a prompt but in understanding the relational dynamics between you and an LLM. If Shamanistic prompting is about letting the LLM speak on its own, Wizard prompting is done both by you and the LLM. Divergence starts to turn toward Convergence.

In the context of AI, this process becomes a metaphor for the transformative potential of technology. The Wizard prompting model helps an LLM synthesize human knowledge into novel ideas and solutions. The prompt becomes a modern spell, with the ancient power to alter digital realities.

Politician Prompts

In the Politician's craft, words are not just words—they are levers of power, used to guide the masses. To persuade is to take the wild strands of public discourse and thought, and weave them into a single, compelling narrative.

The Politician listens—not only as one who seeks to understand, but to detect the pulse of the crowd. A politician uses every tool at their disposal—cues from the audience, analyses of group dynamics, continuous feedback—to tailor their message. It is a delicate act of balancing what is said with how it is received, adjusting in real-time to the subtle shifts in the collective mood.

What if you prompted an LLM with this in mind? 

The Politician conducts variations of the same theme, testing permutations through A/B testing, shaping language to resonate with specific personas and demographic segments. You could ask an LLM to role-play a persona or demographic—and switch the role with you and have the LLM prompt you.

Your prompts would include specific language-shaping requests, like: 

"Analyze the current discourse on [issue] and propose a rhetorical strategy to shift opinion towards [desired outcome]."

"Identify potential objections to [offer or proposal] and formulate persuasive counterarguments tailored to different groups."

"Create an adaptive communication strategy for [initiative] that evolves based on simulated public feedback and sentiment analysis."

"Construct a speech on [issue] that employs ethos, pathos, and logos to sway undecided people while reinforcing support from the existing audience."

"Formulate a series of sound bites on [complex issue] that simplify the message for mass appeal without losing essential nuance."

"Create a strategy to reframe [unpopular ideas] in terms that align with widely-held values and aspirations of [target audience]."

As a master of Convergence, the Politician crafts language that bends the divergent thoughts of the populace towards a chosen point. Each word is chosen for its subjective truth and for its effect, not only to reflect reality but to shape it.

The point of this approach to prompting is not in the manipulation but in understanding the nature of influence. To move an LLM to action, to use language that not only commands but transforms, the Politician must be both the orator and the listener, the influencer and the influenced.

Just as a Politician adjusts their rhetoric and stance, you can adjust an LLM’s weights and biases to better model the complexities of human language.

The Politician often creates personas and segments within the electorate—much like an LLM might segment data to provide personalized responses.

Language, to an LLM, is a multi-dimensional space, where each word, phrase, and sentence is a point with a specific position, defined by numerical values in a high-dimensional vector space. Politicians, like LLMs, navigate this landscape of semantic embeddings, crafting their messages to both retrieve and land at precise coordinates within the public's consciousness.

4. Conversations with Archetypes are the Better Way to Prompt

The I-Thou relationship, as seen through the archetypes of the Shaman, Wizard, and Politician, offers a more useful framework for engaging with LLMs than single-use “mega prompts”. Each archetype promotes divergence, encouraging us to probe further, before writing prompts and converging on anything specific.

Research in LLMs suggests that intent matters more than the specific words used in a prompt. Words are a temporary means to understand intent, and if you focus on intent for your prompts, you’ll be able to get the output you have in mind from any LLM, no matter how they evolve and change.

How do you work on improving your intent? Work through the perspectives of the Shaman, Wizard, and Politician. 

As you solve problems or get whatever outcomes you have in mind, you can listen deeply and commune with the collective wisdom encoded in the LLM. You can use prompts to not just command but also collaborate with an LLM. And you can shape your output for maximum impact.

The true power of LLMs is not in their ability to predict and generate words. It’s in their capacity to expand our own cognitive horizons. By entering into a dialogue with these systems, by invoking the wisdom of the Shaman, the creativity of the Wizard, and the persuasiveness of the Politician, you’re more likely to get not just the output you want but the output you need.

Talk again soon,
Samuel Woods
The Bionic Writer