Games Worth Playing

I was shaped by early Nintendo Japanese role-playing games. Final Fantasy, Dragon Warrior, the whole pixelated apprenticeship of being a small character in a large world with a real problem in front of you and a long grind between you and solving it. The shape was simple. You wake up. You are the hero. There is a thing that needs to be done. The thing is hard, by design. The grind is the point.

I never put that frame down.

I wake up every day and see a character who has to do something hard. I take the hardest road through whatever room I am in because the hardest road is, demonstrably, the one with the most experience points. I studied math in school for exactly this reason — it was the subject everyone complained was too hard, which told me directly that the experience curve was steepest there, and I wanted the multiplier. I have viewed life, since I was small, through the lens of a game. The lens has not aged out. The lens has aged in.

There’s a cost to the game-shaped mind. The cost is that I get distracted by new games. I have started many projects in my life. I have had technical successes inside novel ideas. And then I look up, two years later, and a new game has opened, and the cost of switching is approximately zero in my head because the game logic is play the most interesting one, and the most interesting one is always the new one. This is a real personality cost. It produces output. It does not produce one massive output the way some lives do; it produces a portfolio of medium-sized outputs that an outside observer reads as scattered and that I read, internally, as a well-stocked save file with many side quests cleared.

I love the specialist role. Specialists go deep. Specialists know the actual mechanics, the actual exception cases, the actual undocumented edges. I have read a lot of books across a lot of subjects, and the joy of those books is the depth — the moment where the simplified surface explanation falls away and the real complexity comes into view. That is the gameplay I was after.

The specialist has a problem, though, which I have written about elsewhere and want to name again here: the specialist is the easiest role to enslave. The deeper the expertise, the higher the salary, the larger the golden cage. Specialists who do not plan for an exit become indispensable, and indispensable means captured. The slave mindset has been waiting for them at the end of the corridor the whole time.

This is why I made financial discipline a top priority. Not because I am frugal by temperament — I am not, particularly — but because financial freedom is the only general-purpose escape mechanism the specialist has. The salary is a leash. The savings are the bolt cutters. The day you can walk is the day the game opens back up.

There was a fork in the road, years ago, where I considered playing the obvious next game — take an idea, raise venture capital, push it hard, win big or fail loudly. The standard founder script. The slot every ambitious technical person in my cohort eventually puts a coin into.

Fortunately my own pathologies saved me. The idea I built, instead of refining itself into something a venture firm could underwrite, kept growing in scope and depth in ways that made it less and less investable. I built an unfundable monster. It was novel, it was technically interesting, and it was completely impossible to pitch to a board because the value it produced did not factor cleanly into a five-year exit narrative. I look at this now and I am grateful. The unfundable monster saved me from the specific slavery I would have walked into — the slavery of a CEO with a board, a cap table, a quarterly burn rate, and a roadmap full of mandates I would not have written for myself. The board-of-directors mantle is one of the heaviest a free human can put on. I avoided it by accident. I am keeping the accident.

This is, in retrospect, an instance of a broader pattern: the things I was incapable of doing well were often the things I should not have done in the first place. The “failure” was the system working.

Now we are inside the AI moment, which is a strange place to be standing. AI is the most interesting new toy of my technical lifetime. The pace of change is, by any historical comparison, absurd. It is rewriting how every industry will work in a way that is going to be both better and crazier than almost anyone is currently pricing in.

I am glad I am not in any business right now. Whatever you are building, AI is about to change the floor and the ceiling under it at the same time. The next five years inside any company will be a continuous renegotiation of what is automatable, what is leverage-able, what is obsolete, and what is the new thing. I have run inside hard-rate-of-change environments before. They are exhilarating. They are also exhausting in a way that compounds. I am pleased to be observing this one from the outside, with my own time, on my own terms.

The interesting paradox is that the most powerful tool I have ever had access to has arrived inside the period of my life when I am the most free to use it for whatever I want, and the things I most want from it are things that do not exist yet. The general-purpose models are remarkable. The specific applications I would actually want — for my wife’s health, for research into MS, for a tighter loop on my own biology, for the dozens of small workflow ideas I keep filing in my head — none of them are products you can buy. They are products someone has to build. The financial discipline that bought me the freedom also bought me the lack of need to buy anything in particular. Most consumer luxury is a marginal gain in performance over the non-luxury version. I do not need a fancier car. I do not need a fancier watch. We gave a chunk of money to a zoo because we like animals, and that purchase generated more meaning than any consumer purchase I have made in years. Beyond that, the market does not have what I want, because what I want has to be built, and the question is whether I am the one who is going to build it.

This is the question of retirement nobody warns you about. You spend twenty or thirty years winning a specific game, and the day you finish you have to pick a new one, and the new one is not assigned. There is no game manual. There is no quest giver. There is no NPC to walk over to. The world is open. You have to choose what you are doing now, on purpose, with nothing externally telling you it matters.

Most retirees fail this test. They pick golf, which is fine, or television, which is not. They pick decline, often, because decline is the path of least resistance once nothing is asking anything of them.

I am not going to do that. The remaining question is what. Is it a business? Probably not. A business is a useful tool in service of a game, but a business as the game itself is another harness. The mantle is heavy. I just took it off. Putting it back on for any reason that is not extraordinary would be a failure of taste.

The current answer, day to day, is fitness and biking — both of which I have written about and both of which carry their own real game logic. The body has hard variables. They respond to inputs in measurable ways. You can compound for decades. There is a leaderboard you only ever compete with yourself on, which is the only leaderboard that matters anyway. This is real. It is not the whole answer. But it is part of it, and the part I have is more than most people ever pick on purpose.

The structural induction problem

When I was younger I asked a version of this question and could not answer it. I was depressed and cynical, and I was looking at the standard human script — school, job, marriage, kids, retirement at sixty-five, then a slow fade — and finding it lame. Not in a teenager-rebellious way; in a real engineering way. The script had no base case. It deferred meaning forever. The meaning of school was to get the job. The meaning of the job was to provide for the family. The meaning of the family was the kids. The meaning of the kids was their school, job, family, kids. The recursion never bottomed out. It was meaning by passing-the-baton, with nobody ever holding the baton long enough to look at it.

I wanted a base case. I wanted a definition of meaning that did not depend on the next iteration of the same loop providing the meaning the current iteration was missing. I went looking for it through technical excellence — if I could be world-class at one thing, surely the thing would mean something on its own terms — and I did become world-class at the thing, and the thing meant a lot, and then I finished it and the question came back unchanged: okay, now what.

The kids-as-answer never worked for me, and I have come to make peace with that. Kids are the most common cultural base case offered for this exact problem, and I do not begrudge anyone who picks them, but they do not resolve the structural induction. They defer it by one generation. The kids will ask the same question eventually, and if the previous generation’s only answer was well, I had you, the kids are going to find that answer load-bearing on nothing. The recursion has to terminate somewhere.

For me, looking at the whole thing from inside a fifteen-year marriage, an early-retired calendar, a body I have rebuilt, and a wife I am still in the trench with — the answer is, basically, that the whole thing is madness, and the only sane move inside the madness is to pick a variable and push it as hard as you can, because the pushing is itself the meaning. Not the result. The pushing. The grind is, recursively, the base case. Final Fantasy was right.

The variables I am still pushing

So here, plainly, is what I am pushing:

The body. The marriage. The wife’s disease, treated as a long technical problem with real variables. My own metabolic ceiling, treated as another technical problem. Writing — the essays on this site, which exist because writing forces a thought to either resolve or be exposed as confused, and I want to know which mine are. And the technology swings I have not given up on, even given the unfundable-monster history, because AI is finally a multiplier large enough that small teams can build large things, and the things I want for my wife and for myself are still not products I can buy.

If a swing connects, it connects. If it does not, I will take another. The portfolio-of-medium-sized-outputs problem is still my problem, and at this point I have made peace with it as a feature rather than a bug. Specialists with discipline build small empires. Generalists with discipline build interesting lives. I am the second one. The empires are someone else’s game.

The line I’ve been trying to avoid for the whole essay: the real game is living for yourself on your own terms. Any other answer is cope. The for the kids answer is cope, because the kids will eventually have to find their own answer. The for the company answer is cope, because the company will replace you within a week of your last day. The for the country answer is cope, because the country is a story we are agreeing to tell about a piece of land. The for the legacy answer is cope, because the legacy is what other people choose to remember about you and you do not get to vote.

You live for yourself. You set the variable. You push. You enjoy what gets built. You accept that nothing you build will last past the heat death, and you push anyway, because the pushing is the meaning and the pushing is available to you starting today.

The hero in the JRPG does not solve the recursion either. The hero saves the world, the credits roll, and a new game starts somewhere else with a new hero. The save file is the only thing that persists, and the save file is yours. Play accordingly.