Jan Gunash

There is a parenting insight that does not get enough attention outside of child development circles.

When you raise a gifted child, the advice you eventually receive is counterintuitive: let them struggle. Let them sit in the frustration of not knowing. Because intelligence means things come easily early, and when things stop being easy, a child who has never learned to be frustrated will quit. Not because they lack ability. Because they never built the muscle that converts difficulty into persistence.

We understood this about children and pretend it does not apply to us.

AI tools are removing that tax.

Not maliciously. Not even incorrectly, in many cases. They are doing exactly what they are designed to do: reduce friction. Draft the email. Generate the outline. Write the code. Summarize the document. The frustration that used to live between you and the finished thing has been largely eliminated. And we are calling this progress.

But here is the question worth sitting with: what was living inside that frustration?

Frustration is not just an unpleasant sensation you push through on the way to an answer. It is the signal that your thinking is being tested. It is where critical thinking actually happens, in the gap between what you expect and what you encounter, between what you know and what you need to figure out. When you remove frustration, you do not just remove discomfort. You remove the cognitive load that was quietly building capacity.

And then there is the question of information itself.

When something is fed to you, how much of it do you actually question? A search engine used to return ten links and make you do the work of reading, comparing, and synthesizing. You had to hold conflicting information and decide what to trust. That friction was annoying. It was also the mechanism through which you developed judgment.

Now the synthesis comes pre-assembled. The answer arrives confident, fluent, and complete. And the very fluency of it makes it harder to interrogate. We are wired to trust things that sound certain. We are not wired to ask whether something that reads clearly was actually reasoned carefully. Bias, confirmation bias, and misinformation are rampant and getting worse. One response is to combat it with access to information at the same speed. The better response is to grow into people who seek truth and data before forming an opinion. The tools are not building that habit. They are quietly eroding it.

The parallel to the gifted child is not a metaphor. It is the same mechanism at a different scale.

A child who never struggles does not develop grit. An adult who outsources every hard cognitive task does not develop it either, they lose it. Slowly. Quietly. In the same way a muscle atrophies when it stops being used. You do not notice until something requires you to lift something heavy, and you cannot.

There is a second question underneath the first one, and it is harder to sit with.

When you create something you did not truly make, built from the patterns of a million things others have made, assembled by a system you prompted, does it feel like creation? Does the thing carry the satisfaction of something that cost you something? Or does it feel, on some level, like picking up a trophy for a race you did not run?

I am not asking this to be precious about craft. I am asking because the answer reveals something important about what we are actually optimizing for.

Eudaimonia, the deeper form of happiness rooted in meaning and life satisfaction, does not come from the output. It comes from the process of becoming someone who could produce the output. It is a side effect of doing something that genuinely costs you something. Hedonia is the shortcut. It feels good in the moment. But the more you chase it, the less it delivers, and the more easily disrupted you become when friction returns.

We are building lives, workflows, and cognitive habits around the elimination of friction. At exactly the moment when the problems worth solving are becoming more complex, more ambiguous, and more resistant to easy answers.

The gifted child who never learned to be frustrated does not become less capable in easy environments. They become capable only in easy environments. And the world has a way of eventually presenting problems that are not easy.

The question is not whether to use the tools. The question is whether you are using them in a way that preserves the frustration tax, or whether you are quietly exempting yourself from it, one frictionless output at a time.

Because the thing you are building when you sit in the discomfort of not being done yet is not just the work.

It is you.

And as the tools become limitless, what are we left with?

This is the part nobody wants to answer out loud.

Capital is moving in one direction. The best engineers, the sharpest researchers, the most ambitious founders are being pulled toward the companies building these systems. Billions in investment, the most talented people in a generation, and the bulk of technological urgency are concentrating inside a handful of organizations whose core product is the automation of cognitive work.

The services being built on top of that infrastructure are not free. Every query, every generated output, every API call runs on compute that costs real money. What you are paying for today is subsidized. The underlying bet is that compute costs will fall fast enough that the economics eventually work. The companies building the infrastructure are absorbing that gap right now, running at a loss or at razor margins, betting on a future where the cost to run these models drops dramatically. That bet may be right. It probably is.

But there is a cost that does not appear in any of those calculations.

The cost is on the humans using the tools while they wait for the economics to resolve. Not a financial cost. A developmental one. Every day that friction is removed from thinking, every synthesis that arrives pre-assembled, every creative act that gets outsourced to a system trained on the work of others, is a day where the capacity that friction was building does not get built. That cost does not show up on a balance sheet. It accumulates quietly, inside people, across years.

So what remains when the tools become truly limitless?

Here is the honest answer, with some optimism left in it.

The tools will keep improving. The economics will eventually stabilize, and access will widen. And there will be people on the other side of this transition who used the tools without losing themselves in them. Who stayed curious about whether the answer was right, not just whether it sounded right. Who kept doing hard things on purpose, not because they had to, but because they understood what was at stake. Who used the leverage to go further, not to go less deep.

Those people will be rare. They will also be the most valuable people in any room.

The talent and capital flowing toward these tools is real, and the tools will be extraordinary. But the person who can think clearly underneath them, who has the judgment to direct them and the grit to know when to set them aside, that person does not get automated. That person gets amplified.

The frustration tax is still being collected. The question is whether you are paying it, or deferring it.

Deferring it always costs more in the end.

---

Want an email when the next piece comes out randomly?