The Dialogical Engine
How AI Is Forcing Us to Evolve
My seventeen-year-old said something that stuck with me: “Everyone in my class is using AI now, but it’s actually harder to stand out. It’s like we’re all copying from the same smart kid.”
He’s onto something. I’ve spent years helping people (leaders, athletes, military personnel) be at their best when it matters most. And I keep coming back to the same insight: it’s never the tool that determines the outcome. It’s how you use it. Some people will use AI to think more clearly, work more effectively. Others will use it to avoid thinking altogether. The question isn’t whether AI makes things easier or harder. It’s what kind of person you become in the process.
The floor has risen. What used to require skill (clear writing, quick research, coherent argument) is now available to everyone.
Competence is abundant. Judgment isn’t.
We’re living through evolutionary pressure. AI is the environmental change forcing adaptation, not through competition, but through opposition: the gap between what we could do and what we should do, between speed and wisdom, between generation and judgment.
Two and a half millennia ago, Socrates walked the streets of Athens annoying people into awareness. Thinking, he believed, sharpened through dialogue.
AI responds to questions the way Socrates did, but without intention, without judgment. The quality of its response depends entirely on the quality of your question. Prompt lazily and you get answers that sound smart but say little.
My older son is in college, caught between worlds. He’s old enough to have learned to write and think before AI, young enough to use it fluently now. He has an advantage his classmates may not: he developed judgment before he had access to instant generation.
Last month he was writing a philosophy paper on free will. Stuck. I showed him how I work: I asked Claude to argue against his thesis as hard as possible. Not to write the paper, but to stress-test it. Within ten minutes, he’d found three holes in his reasoning he hadn’t seen. The AI didn’t fix his argument. It revealed where his thinking was loose. He had to do the actual work of tightening it.
He could do this because he’d already learned to recognize the difference between an argument that sounds good and one that holds up. He knows what good thinking feels like because he had to struggle toward it himself.
I didn’t learn that from a machine. I learned it from my father, a philosophy professor who was an unsparing but loving editor. He’d mark up my school papers like a Jackson Pollock painting. Red ink everywhere. He’d circle a word: “Is that exactly what you mean?” He’d cross out entire paragraphs: “You’re not saying anything here.” Not cruel, but unsparing. He taught me that every vague phrase is a vague thought.
That ear for precision is what lets me use AI effectively now. Those who skip this part (who never develop that ear for precision) miss the honing that builds judgment.
Socratic reasoning thrives on contradiction. Two ideas meet, clash, and, if we’re patient, produce something new. This is the dialogical engine, the mechanism that converts opposition into understanding.
The nervous system itself learns through prediction error: the moment when reality contradicts expectation. That gap, that friction, is where intelligence sharpens.
AI is now the opposing force.
Patterns emerge in who thrives and who flounders.
Some people treat AI like a better search engine, a way to find information faster. Others treat it like a replacement for thinking. But a smaller group has figured out something different: they use it as a sparring partner. They understand that the technology reveals where their thinking is clear and where it’s not.
These people share four distinct capacities. Not skills you can learn from a tutorial, but disciplines built through resistance:
Inventive Thinking. The capacity to originate rather than optimize. AI can predict and recombine existing patterns brilliantly. It cannot imagine what doesn’t exist yet. That remains ours: the ability to make lateral leaps, to ask questions no one thought to ask, to combine ideas in ways that reframe everything.
Rhetoric. The craft of shaping thought through language. In the age of prompting, this is performance. The precision of your question determines what follows. Vague prompts yield vague results. Specific prompts, carefully constructed, yield insight.
Discernment. The ability to judge quality. To think and feel the difference between what’s simply competent and what’s really good, between what sounds right and what actually works. This is judgment, and it can’t be automated.
Identity. A clear sense of who you are underneath all the doing. This is what gives your capabilities purpose and direction. Identity is ongoing self-authorship: knowing what you stand for and what you won’t compromise. It’s what makes you legible to yourself. Without it, you’re just optimizing for metrics you didn’t choose.
These capacities are built through resistance, not flow. I know this from thirty years in performance psychology: growth happens at the edge of discomfort.
Every technology extends some human faculty and tests another. The printing press extended memory but weakened recitation. The camera extended vision but dulled observation. AI extends reasoning and tests reflection.
The machine, at least for now, has no self. It cannot wonder why it wonders. Our role becomes clearer: to supply the inner dialogue, the context, the purpose that gives its capabilities direction.
AI’s strength reveals our weakness. Its weakness demands our strength.
The challenge isn’t scarcity anymore. It’s abundance. It’s easier to produce a dozen mediocre options than commit to one good one. Easier to move fast than think clearly.
While writing this essay, AI generated a sentence for the identity section: “This is what makes precision different from manipulation, creativity different from chaos.” It sounded profound. But I couldn’t explain what it actually meant, not to myself, not to anyone. I had to stop and figure out what I was really trying to say.
Stopping to question yourself goes against everything the technology encourages. The people doing the best work have learned to resist that pull. They’ve understood something crucial: the opposition isn’t machine versus human. It’s capability versus judgment, speed versus discernment, generation versus curation.
Those who can hold that tension will succeed. Those who collapse into pure efficiency will be left behind.
We can no longer avoid choosing: will we develop judgment or outsource it?
My son will grow up in a world where generating content is as easy as searching for it. He’ll write his college essays. He could have AI draft them, polish them, optimize them for admissions keywords. They’d probably be good enough.
But good enough for what? To get in? Or to figure out what he actually thinks, what he actually wants, who he’s actually becoming?
These capacities (inventive thinking, rhetoric, discernment, identity) are what I want for him and his brother. Not because they’ll make them competitive, though they will. But because they’re what make you the author of your work rather than its curator. They’re the difference between using a tool and being shaped by it. Between knowing your own mind and optimizing for patterns you didn’t create.
It’s forged in conversation: between logic and emotion, mastery and mystery, capability and limitation. This capacity to be transformed by what we encounter, to hold contradictions long enough for them to yield insight, remains, at least for now, distinctly ours to cultivate.
When my father read this draft, he said I was overemphasizing the difference between humans and machines. He remains unconvinced the differences won’t eventually vanish. “Species come and go,” he said.
Maybe he’s right. Maybe the capacities I’m describing aren’t permanent firewalls but temporary advantages. Maybe consciousness isn’t the moat I want it to be.
But even if that’s true (especially if that’s true) here’s what matters now: My sons are growing up in the world as it is, not as it might be in twenty years. The question isn’t whether machines will eventually close every gap. It’s whether my sons will develop the judgment to navigate the world they’re actually living in. Whether they’ll learn to use these tools in ways that make them more themselves, not less.
That’s not a metaphysical argument. It’s a practical one.
I should tell you something.
I didn’t write this essay. Not in the traditional sense.
I wrote it through dialogue with AI, going back and forth between Claude and ChatGPT. Most of the ideas are mine. The argument, the structure, the examples from my work and family are all mine. But these sentences? Most were generated by the machines, then refined through dozens of rounds. I’d read what one produced, recognize what was right and what was off, try the other, push toward precision, cut the generic, add what I actually meant.
What you’ve read is the artifact of exactly the process it describes.
I could have written this the old way. But that would have been dishonest: writing an essay about using AI well while refusing to use it myself. So I wrote it the way I actually work now: in dialogue with the machine.
The AI gave me speed and fluency but couldn’t give me discernment. It could generate a thousand variations but couldn’t tell me which one was true. It could produce coherent arguments but couldn’t know which mattered. It could write about identity and purpose but had neither.
Every choice that made this essay what it is (what to keep, what to cut, what rang true, what felt false) came from me. From thirty years working with people when it matters most. From a father who taught me every vague phrase is a vague thought. From raising two sons in a world I’m still trying to understand.
The software was capable of holding the whole essay while I focused on one paragraph. But only one of us had something at stake. Only one of us cared whether this was worth saying.
This was genuine dialogue. Socratic in structure. I posed questions, the software responded, I pushed back, it refined, I challenged assumptions, it offered alternatives. Thesis, antithesis, synthesis. Over and over until something true emerged from the opposition.
The question isn’t whether AI was involved. The question is whether what I made means something. Whether the thinking is sound, the argument coherent, the examples illuminating. Whether it’s true.
That’s the test we all face now. Not whether we use AI, but whether we use it in ways that amplify our intelligence rather than replace it. Whether we bring enough judgment, taste, and purpose that what emerges is genuinely ours.
If Socrates were alive, he wouldn’t fear AI. He’d interrogate it. He’d ask what it teaches us about ourselves. And he’d recognize what’s happening: we’re being forced to adapt, not by competing with machines, but by developing the distinctly human capacities the machine cannot yet replicate.
The measure of this moment won’t be what AI can accomplish. It will be whether we become the kind of humans who can hold tension, who choose deliberately, who know what we mean and why it matters. Whether we develop judgment or outsource it.
The question isn’t whether AI makes things easier or harder.
It’s what kind of person you become in the process.