Toolmaker or Tool User: Finding Your True Nature in the AI Age

I’ve been wrestling with a question that won’t let go: who’s more valuable, me using the LLM or the people building it?

It surfaced during a conversation, and I haven’t been able to shake it since.

It’s the kind of question that feels important because it touches on something real—value, purpose, identity in the age of AI. But it’s structured like a competition. A zero-sum choice. As if the answer is either the violinmaker or the virtuoso, but never both.

After days of thinking, I realized the real question is different: What are you choosing to pay attention to?

The Fish Who Built the Ocean

There’s a David Foster Wallace parable that keeps surfacing in my head. Two young fish are swimming along, and an older fish swims by and asks, “Morning, boys. How’s the water?” The young fish keep swimming. Eventually one turns to the other and asks, “What the hell is water?”

Wallace’s point: the most obvious, important realities are often the hardest to see.

I’ve been coding with LLMs for over a year now. Every day. Building systems, debugging issues, prototyping features. At some point, the LLM stopped being a tool I picked up and became the environment I work in. I don’t think “should I use the LLM for this?” anymore. That’s like asking “should I use oxygen for this breath?”

This is water.

But here’s what wrestling with this question forced me to see:

I’ve been swimming in it without asking what it means to swim in it.

The Musician’s Paradox

The question came through an analogy: Who’s more important—Stradivari who built the violin, or the virtuoso who plays it?

The obvious answer: both. Without the luthier, no sound. Without the player, no music. The violin sitting in a case has potential but no kinetic energy. The virtuoso without an instrument has intent but no expression.

But LLMs break the analogy in a specific way.

A violin is passive. When you play it, it doesn’t suggest a melody back to you.

When you code with an LLM, it does.

It’s not a tool you pick up. It’s a collaborator that talks back, suggests alternatives, and sometimes catches things you missed.

So the question isn’t “tool or toolmaker” at all. The question is: who are you when the tool has agency?

The Default Setting

Wallace had another insight in that speech. He said humans are “hard-wired” with a default setting: everything in our experience supports the deep belief that we are the absolute center of the universe.

He wasn’t being pessimistic. He was being precise. The default mode of human consciousness is self-centered because that’s how we experience existence—from inside our own heads, looking out.

And that default setting shapes how we ask questions. Like “who’s more valuable—me or them?”

But what if that’s the wrong frame entirely?

I spent a few days on this. Talked to an LLM about Maslow’s hierarchy, self-actualization, ego versus true nature. The model came back with something I didn’t expect. It said Maslow himself revised the pyramid late in life. He added a sixth level above self-actualization: self-transcendence.

Self-Actualization: “I am fulfilling my potential.” (Still self-centered.) Self-Transcendence: “I am serving something higher than myself.”

The shift isn’t about better or worse. It’s about orientation.

Self-actualization asks “what can I become?”

Self-transcendence asks “what does this moment need from me?”

Two Kinds of Happiness

The LLM walked me through another distinction: hedonic versus eudaimonic well-being.

Hedonic happiness is biological feedback. Food, status, comfort, dopamine hits. It operates on a treadmill. You get the hit, it fades, you need a bigger hit next time.

Eudaimonic happiness is alignment. Flow states. Meaningful work. Deep connection. Acting in accordance with your values even when it’s hard.

The marker: eudaimonic happiness can coexist with suffering. You can be exhausted, uncertain, or struggling—yet still feel a deep, resonant sense of rightness.

I used to think I was happy because I was productive.

Now I think I’m content because I’m expressive—using the LLM not to replace my work but to amplify my capacity to make things that didn’t exist before.

That sounds like progress. And it is. But it also brings up a real question: if the tool is doing half the work, where does “my work” begin and end?

The Via Negativa Test

The LLM suggested something I didn’t expect. It told me to run a “via negativa” audit. Strip away one instinctive comfort for 24 hours—social media, favorite food, filling silence with noise—and watch what happens to my mind.

The idea: the anxiety you feel is your ego dying.

The calm awareness that sits behind that anxiety is your true nature.

I tried it. I didn’t code with the LLM for a day. Just me, a text editor, and the problem.

And you know what I noticed?

The first hour: anxiety. “Am I wasting time? I could solve this faster.” The second hour: stubbornness. “I can figure this out.” The third hour: flow. I was in it. No interruptions, no suggestions, just me and the problem.

But here’s the thing—by hour four, I missed the LLM. Not because I was stuck. Because I wanted to share what I’d figured out. I wanted to say “look at this approach—does this make sense?” I wanted the back-and-forth.

So what does that mean? Is the LLM a crutch? Or is it a collaborator I’ve integrated into how I think?

The Water You’re Swimming In

Wallace argued that real education isn’t about learning what to think—it’s about learning how to think. It’s about “exercising some control over how and what you think.”

He said everyone worships something. Money, power, status, intellect. The question isn’t whether you worship, but what you choose to worship. And whether you’re conscious of the choice.

I think the question—“who’s more valuable?”—was really asking: what am I worshipping?

If I’m worshipping productivity, then the toolmaker wins. They built the engine. I’m just the driver.

If I’m worshipping output, then I win. The LLM generates code, but it doesn’t ship products. I do.

But if I’m worshipping the work itself—the craft, the process, the problem-solving—then the question dissolves.

The value isn’t in the human or the machine. It’s in the interface. The collaboration. The thing we make together that neither of us could make alone.

What I Think Today

I don’t think I’m a tool user. I don’t think the people at OpenAI or Anthropic are just toolmakers.

I think we’re swimming in something new. And most of us haven’t noticed it yet because, like the fish, we’re inside it.

The LLM isn’t a hammer. It’s not even a violin. It’s more like language itself—a shared, generative medium that shapes how we think while we’re using it.

And the question isn’t “who’s more valuable” any more than it’s valuable to ask “who’s more important, the person who invented grammar or the person writing the sentence?”

The real question is: now that you see the water, what are you choosing to do in it?

For me, right now, I’m choosing to keep coding with the LLM. Not because it makes me faster (though it does). Not because it makes me better (though it might).

But because the back-and-forth—the dialogue, the iteration, the “what if we tried this instead”—feels like the work I want to be doing. It feels like expression. Like practicing a craft that didn’t exist five years ago and still doesn’t have a name.

Maybe that’s self-actualization. Maybe it’s self-transcendence. Maybe it’s just me figuring out how to swim in new water.

I don’t know yet. But I’m paying attention. And that, I think, is the only choice that matters.

---

Sources:

• David Foster Wallace’s “This Is Water” - Full transcript
• Kenyon College Archives - Original speech context