The Mathematician Terrified of AI Who Joined OpenAI Anyway

Jacob Tsimerman is one of the most brilliant mathematicians alive. He won the Fields Medal — the Nobel of math — at age 34. He has spent his career understanding the deepest structures of numbers. And he is absolutely terrified of artificial intelligence.

So he did the one thing that seems to make no sense at all. He took a job at OpenAI.

Abstinence does not prevent the apocalypse. If you believe the bomb is being built, you have a moral obligation to be in the room.

That’s the cold, hard logic behind Tsimerman’s decision. And it might be the most important AI safety strategy nobody is talking about.

You’ve probably heard the standard arguments. We need to pause AI development. We need to slow down. We need to regulate. These are all fine sentiments. But they miss the point. The train has left the station. The technology is being built in labs around the world, and no amount of op-eds or petitions is going to stop it.

So what do you do if you’re a world-class mathematician, you see the existential risk clearly, and you want to actually do something about it? You don’t stand on the outside and shout. You get inside the room where the decisions are made.

That’s exactly what Tsimerman did. He joined OpenAI’s reasoning team. According to the Wall Street Journal, he was deeply conflicted. He knew the risks. He knew the potential for catastrophic outcomes. But he also knew that staying in academia would mean having zero influence on the direction of the technology.

You can’t steer a car from the sidewalk. If you’re afraid of the crash, you need to be in the driver’s seat.

This is the uncomfortable truth that the AI safety community often avoids. The most effective way to mitigate existential risk is not to abstain from the technology, but to embed the smartest, most cautious minds directly inside the organizations building it. It’s a trade-off between complicity and helplessness. Tsimerman chose complicity. And he’s right.

Think about it. The people who are most worried about AI are often the ones least equipped to do anything about it. They’re philosophers, ethicists, journalists. They write papers and give talks. Meanwhile, the engineers at OpenAI, Google, and Anthropic are building the future. The only way to shape that future is to be part of the building process.

Tsimerman is not naive. He knows that working at OpenAI means he becomes part of the machine that might create the very thing he fears. But he also knows that the alternative — staying pure, staying clean, staying on the outside — is a form of moral abdication. If you believe the apocalypse is possible, you have a duty to do everything in your power to prevent it. Even if that means getting your hands dirty.

The best way to survive a fire is to be the one holding the hose.

This is a contrarian take. Most people assume that the ethical response to a dangerous technology is to distance yourself from it. Tsimerman’s decision flips that assumption on its head. He’s not running away from the fire. He’s running toward it. Because he knows that the only way to control the fire is to be the one who sets the boundaries.

The implications are profound. If you’re a top AI researcher, or a mathematician, or a physicist, and you’re worried about the future of AI, don’t join the protests. Don’t write the think pieces. Apply to the labs. Get inside. The future is being built right now, and the people building it are the ones who will decide its shape. The rest of us are just watching.

So here’s the question that every smart person afraid of AI needs to ask themselves: Are you willing to be complicit in the risk in order to have a chance at preventing the catastrophe? Tsimerman answered yes. And his answer might just save us all.

If you’re scared of the future, build it yourself. Because the only thing worse than being complicit in a disaster is being a helpless bystander when it happens.