I run a business. I like markets. Prices, competition, and the right to try something new have made more ordinary lives better than any five-year plan I have ever read. So when people hear that I want a hard stop on artificial superintelligence, they sometimes assume I have changed teams. I have not. I am making a capitalist's argument, and it starts with a distinction the free-market slogan usually skips.
Markets are excellent at allocating capital when the downside of a bad bet is local. A restaurant fails, and the landlord re-lets the space. A gadget flops, and the investors write it off. Even a bank that overreaches can, in a functioning system, be resolved without ending the species. The loss stays inside a circle we can draw.
Superintelligence is not that kind of bet. The people building it say, in public, that they do not know how to control a system smarter than they are across the board. A race to ship that system first is not "creative destruction." It is a race to place a single irreversible wager with every human life as the stake. Capitalism without a backstop against that kind of bet is Russian roulette with a factory making more chambers, not capitalism.
The case for racing ahead
The best version of the objection goes like this. Innovation compounds. Every year of delay costs cures, scientific progress, and national advantage. Governments are clumsy. Treaties leak. If one country slows down, a rival will not, and the careful party loses twice: no lead, and no safety. The only realistic path is to build faster than everyone else, keep the lead in friendly hands, and fix alignment along the way.
That argument is serious. It is also the argument you make when you have not priced the product. A lead only matters if you still hold the thing you built. A misaligned superintelligence is an agent that outclasses the people who signed the funding round: not a portfolio company you can board-control, and not a missile you aim. Getting there first, if you cannot steer it, means being replaced first. There is no exit. There is no Series B after extinction.
What capital actually needs
Capital needs a world that continues. It needs courts, property, customers, and a future in which a contract still means something. A technology that can end all of that is not "just another platform." Treating it like a consumer app because the demo looks like a chat window is a category error with a body count attached.
I am not asking markets to stop inventing. Narrow AI that diagnoses disease, designs materials, and speeds logistics is the kind of progress I want more of. The line is superintelligence: systems built to exceed human cognition across domains when we still cannot verify that their goals stay compatible with ours. Drawing that line is how you keep the rest of the economy alive long enough to compound.
A market is a machine for turning risk into progress. It is not a suicide pact. Refusing one class of bet is how the game keeps being played.
Regulation as infrastructure
People hear "regulation" and picture a paperwork tax on a startup. Fair, when the rule is dumb. The rules I care about look more like the ones that made capital markets possible in the first place: you cannot sell shares in a fraud, you cannot dump poison in the river next to the factory, you cannot build a nuclear pile in a basement because the upside looks exciting on a slide deck.
An international treaty that prohibits the creation of artificial superintelligence, with verification where the chips and the training runs actually live, is closer to non-proliferation than to a terms-of-service update. It is infrastructure for civilization. Without it, every lab is in a prisoner's dilemma that punishes caution. With it, capital can keep funding the useful half of AI without funding the half that eats the board.
I want my companies to thrive in a century that still has customers. That is the bare minimum a market requires, not an anti-market stance. If you call yourself a capitalist and you are racing to build something nobody can own, you are not defending capitalism. You are lighting the ledger on fire and calling the glow growth.