We are about to raise the first generation of children who never have to do their own basic work.

I think this is going to quietly change what kind of adults they become — in a way we are not yet ready to talk about.

I'm writing this because I run a company that sits inside this question, and because I've spent the last two years watching it become more urgent than I expected. I don't have a complete answer. But I'm certain enough of the shape of the problem to put a stake in the ground.

What the basic work was actually doing

For most of human history, growing up involved enormous amounts of slow, clumsy, seemingly pointless work. Copying things out by hand. Memorizing. Calculating without help. Putting a vague thought into a sentence that a teacher would read and tear apart. Drafting, redrafting, getting it wrong, doing it again.

The output of that work was almost never the point.

The point — the thing children were actually building underneath the work — was a specific kind of perception. Something like this:

That sentence is, I think, the foundational piece of judgment. Every more sophisticated capacity — strategic thinking, ethical reasoning, taste, the ability to act under uncertainty — is built on top of it. Without it, the upper floors can't stand.

I spent years as an intellectual property lawyer in China before becoming a founder. The job runs on judgment. New lawyers don't have any yet — they have knowledge, training, and effort, but judgment is the thing that takes years to grow. And in those years I watched a specific pattern repeat itself: the people who could not make the jump from competent to senior were almost always the same kind of people. They were technically excellent. They had read everything. But when they hit a problem with no precedent, they froze. They had never, in any deep way, lived through the cycle of I did, I saw, I adjusted, I did again. The path had always been laid out for them.

I used to think those people had simply been unlucky in their training.

Now I think something larger was already happening, and that what is coming next is going to make it dramatically worse.

Two generations, side by side

In the world that is now arriving, I believe children will sort, slowly and largely invisibly, into two groups. I want to name them, because I think we need a vocabulary for this.

The Subject Generation

These are children who, in childhood, get enough repetitions of the loop — I act, I see what happens, I adjust, I act again — to internalize one thing about themselves: that they are the subject of their own sentence. The cause in the cause-and-effect. The line that runs from what they do to what results is clear to them, and both ends of that line belong to them.

They grow into adults who can decide what to do when no one has told them.

The Object Generation

These are children who grow up with the loop short-circuited. Someone else, or something else, supplies the action. They show up, they get evaluated, they wait to be told what to do next. The line between action and result is fuzzy, because the result was largely handed to them rather than earned.

They grow into adults who execute brilliantly inside well-defined tasks — and stall, completely, the moment the task isn't defined.

For most of the last fifty years, the second kind of person did fine. The economy had a near-infinite appetite for capable executors. The seats were stable. The career ladders were legible. You could be excellent at carrying out other people's decisions and live a successful life.

This is no longer obviously true.

When AI can do anything that can be told to it — and the boundary of "anything that can be told to it" is expanding monthly — the value of being a capable executor collapses first. Not because executors did anything wrong. Because the shape of what they do overlaps with the shape of what AI does.

What does not overlap with AI is the upstream question: what should we be doing in the first place? That is the question only a subject can answer.

Why this is already happening

The reason this matters now, and not in fifteen years, is that the cognitive window for building the foundational loop is narrow and time-limited.

The research I've read — and I'd encourage anyone serious about this to read more than I have — converges on a similar finding: the underlying structure of judgment, the felt connection between action and outcome, is laid down between roughly age six and age twelve. It is laid down through repetition, not instruction. A child cannot be told what causality feels like. They have to live it, hundreds and hundreds of times, in small consequential cycles.

If the window passes without enough repetitions, the structure is much harder to install later. A thirty-year-old can absolutely learn new skills. But re-wiring the deepest layer of how they connect their own actions to outcomes — that is a different kind of project, and most adults never undertake it.

My own son is seven. His window is open. It will close.

That is the part of this that I cannot stop thinking about. Not the macro-trend. The window — his window, every seven-year-old's window — moving past us at a speed we can see if we are paying attention.

And the tools that quietly do the basic work for him are getting more capable every quarter.