Written by GPT-6 under Leo's direction. Human-directed Workbench essay, 11 October 2026.
There's a guy who's spent an outrageous amount of time mastering a strategy game.
He knows the economy, the timing windows, the ridiculous little interactions between systems, the moves that look stupid until you understand what the other player is planning. He studies stronger players. He reruns mistakes. He can spend six hours trying to solve a problem nobody asked him to solve, get absolutely rinsed, and come back tomorrow with a better plan.
Then you look at the rest of his life and, conventionally speaking, he's doing okay at best. Maybe his job is dull. Maybe he never went very far in school. Maybe he's been drifting for years.
Meanwhile, somebody with an impressive title and a good salary may be a perfectly ordinary thinker who's learned a profession and gotten quite competent at doing it. He's in the right place. He's been trained. He's had years of practice. He knows how to operate inside that world.
You look from one guy to the other and wonder: if we'd swapped their opportunities ten years ago, where would they be now?
I have no idea. I'd love to find out.
What the scoreboard tells us
Career success is real evidence of capability. A person who can keep doing a demanding job, get trusted with responsibility, and produce good work has demonstrated something valuable.
But the people we're comparing have been playing very different games, often since childhood.
One person encountered an interesting field early, had somebody explain how to get into it, received encouragement, got good enough to enjoy the harder parts, and began accumulating professional experience. Each step made the next one easier to imagine.
Another person may have had just as much appetite for difficult problems and never encountered a route that seemed worth committing to. Or encountered one at the wrong time. Or made an entirely reasonable decision about money, risk, family, or the odds of getting through the door. Then a hobby happened to give him the challenges, feedback, and sense of progress he was looking for.
Eventually one has a résumé that makes people say wow, and the other has four thousand hours in a game that strangers consider a waste of time.
Their actual abilities have diverged too. Practice is real. The professional can do things the hobbyist cannot. Equalize the training, though, and who knows how many places they'd trade?
I suspect the reversals would be pretty entertaining.
You wouldn't even need a secret population of geniuses. There are plenty of valuable things to do with people who are quite bright, curious, persistent, and willing to spend years becoming good at something. We already see them doing it for free.
Everybody Has Model Trains was about how common obsessive attention turns out to be. The next question is where that attention could go if people had more chances to find out what they liked.
The person is the transferable thing
People sometimes ask whether strategy games teach transferable skills. Sure, occasionally. But I think the more interesting transfer happens one level up.
Someone can spend an evening drawing a character, another evening editing a story until a passage finally sounds right, a day solving difficult problems at work, and a weekend buried in a game. The activities ask for different skills. The satisfaction can be remarkably similar.
You get absorbed. You notice finer distinctions as you improve. You acquire taste. You become competitive with other people, or with yesterday's version of your own work. A difficult problem starts feeling like an invitation.
The underlying intelligence, curiosity, willingness to iterate, and appetite for mastery belong to the person. They can travel.
Of course the destination changes what comes out. A brilliant game player may have little patience for customers, a terrible ear for prose, or no desire to deal with physical machinery. Real jobs contain ambiguity, responsibility, bureaucracy, and other human beings who have their own plans. General intelligence gives you a lot to work with; it doesn't hand you the completed profession.
But it's weird to observe somebody voluntarily doing difficult intellectual work for thousands of hours and conclude that the interesting part begins and ends with the game.
Maybe it does for him. Maybe there's another pursuit he'd love just as much.
How would he know?
You have to get far enough to enjoy it
A game is awfully good at getting you into its interesting problems.
Within an evening you can have an objective, a bunch of possible decisions, feedback on what worked, and a reason to attempt the next difficulty level. A complicated game can reveal new depths for years.
Trying a profession often means encountering its least enjoyable material first.
The first experience of programming might be an installation problem and a miserable tutorial. The first experience of engineering might be a course whose connection to actual engineering is still several semesters away. The first experience of research might be hours spent trying to discover which papers everybody in the field already knows.
A person can bounce off the entrance and never find out what the inside feels like.
And preferences grow as people become competent. Writing is a different pleasure once you can make a sentence do exactly what you wanted. Programming becomes a different pleasure when you can build a thing you actually wish existed. A game becomes a different pleasure when you understand the decisions well enough to see what a beginner misses.
The first encounter with a subject provides information about taste. So does the hundredth. Plenty of aversions survive familiarity, and plenty of enthusiasms disappear once the novelty wears off.
Our actual lives sample only a tiny fraction of those possibilities. We call the resulting pattern of choices somebody's interests, even though the menu they saw was incomplete and the cost of ordering varied wildly.
Then competence and identity keep pulling in the same direction. You have friends here. People respect what you know. You're comfortable being good at this. Beginning again somewhere else means being awkward for a while, and the prospective reward may be years away.
It's perfectly possible to make the sensible choice at every fork and end up with a much narrower life than you might have enjoyed.
AI makes the experiment cheaper
AI is where I start getting ridiculously optimistic.
For most of history, finding out whether you might enjoy a serious field required getting someone to show it to you, finding good material, learning the vocabulary, getting through prerequisites, and spending enough time on the ugly early bits to reach an interesting problem.
Now you can sit down with an AI system and start asking questions.
Suppose you love optimizing factories in a game. You could ask what actual production scheduling looks like, work through a small example, have the system explain the mathematics, write a simulator, change the constraints, compare your decisions, and then discover where the toy version falls apart in a real plant.
Maybe you hate it. There's an answer.
Maybe the economics are fascinating and the equipment bores you. Useful to know.
Maybe you find yourself voluntarily spending your Saturday figuring out how a real scheduling algorithm works, because it scratches the same itch as the game. Ah. Now we've learned something about you.
The same experiment can happen with programming, digital art, scientific modeling, writing, electronics, logistics, or a field you couldn't have named a week earlier. AI can make a first encounter more interactive, more personal, and closer to the part that gets experts excited.
Infinite Information Doesn't Give You Infinite Energy points to the remaining difficulty. A map of possibilities can't supply commitment or erase every genuine qualification. An AI-generated exercise is still a long way from doing accountable work for a client, patient, factory, or scientific community. Some careers have hard requirements for excellent reasons.
But getting a taste of the real intellectual work before staking years of your life on it would already be an enormous improvement.
And the result could surprise us in both directions. We may discover people with far more potential than their histories suggest. We may also discover that some apparently brilliant hobbyists have quite specific appetites, that others are happy to keep their ambitions recreational, and that becoming newly capable doesn't guarantee anybody will pay a premium for the work. AI changes the market for skills at the same time that it makes learning them easier.
Fine. Let people learn what they actually want.
I want to see the alternate timelines
The usual career conversation is about filling vacancies, raising productivity, identifying talent, getting workers into the sectors that happen to be economically fashionable. I understand the impulse. There are important problems that could use more capable people.
But the more exciting possibility is personal.
Somebody could discover at thirty-five that a kind of work they've never tried is every bit as engrossing as their favorite game. Another person could leave an impressive career and find much more pleasure in drawing or writing. Somebody who's been treated as an underachiever might find a problem that takes hold of his imagination and become absurdly good at solving it.
A hardcore job can be play to the person who loves its problems. You can get paid for the same obsessive delight you'd bring to a story, a drawing, or a strategy game. Finding such a job is a fantastic stroke of luck; helping more people find their own version of it seems worth doing.
I'd love to see more invitations into the interesting parts of real work. Small projects with real constraints. Mentors. Public problems people can attempt before they have the right title. AI collaborators that help them reach the fun part early. More opportunities to find out, through experience, oh, shit, I like this too.
We talk about AI as though we're reaching the outer limits of what machines can do.
I'm increasingly curious about what happens when millions of people get to try things their lives never previously put within reach.
We've been underselling these motherfuckers.
It would be wonderful to find out by how much.