Hours after someone threw a Molotov cocktail at his San Francisco home, Sam Altman sat down and wrote the most honest thing he has published in years.

The fear and anxiety about AI is justified, he wrote, beneath a family photo of his husband and baby. We are in the process of witnessing the largest change to society in a long time, and perhaps ever.

Let that sit for a second. The CEO of the most valuable AI company on Earth just validated every person who has been called a doomer, a Luddite, or a fear-monger for questioning what this technology is doing to jobs, to mental health, to warfare, to democracy. He said the quiet part loud: this is scary, and you are right to be scared.

Then he compared AGI to the Ring of Power from Lord of the Rings. His argument: the correct move is to destroy the Ring, meaning no single entity should control superintelligence. His proposed solution is to share the technology broadly so no one holds the ring.

It is a beautiful sentiment. It is also completely disconnected from what his company actually does.

On the same weekend that Altman published this confessional, OpenAI acquired personal finance startup Hiro Finance in what TechCrunch is calling an acquihire. The company is shutting down on April 20, deleting all data by May 13, and its roughly 10-person team is joining OpenAI. Hiro's founder, Ethan Bloch, previously sold fintech startup Digit for $230 million. This follows OpenAI's acquisition of media company TBPN, its purchase of Rockset, and its relentless expansion into healthcare, defense, retail, and enterprise software.

So here is the contradiction: the man who says nobody should hold the Ring of Power is buying every company that might eventually become a competitor. The man who says AI fear is justified is building the thing people are afraid of as fast as humanly possible. The man who says he wants democratic governance of AI is leading a company that just struck a deal with the Pentagon after its rival got blacklisted for refusing to drop its safety guardrails.

Casey Newton at Platformer made a devastatingly simple observation: it was Altman and his fellow AI CEOs who spent the past decade speaking about AI in existential terms. They compared it to nuclear weapons, the Manhattan Project, demon summoning. They told us to be terrified. Then when someone actually got terrified enough to throw a firebomb, they asked everyone to calm down.

To be absolutely clear: political violence is wrong. Full stop. The firebombing was wrong. The shooting at the Indianapolis councilman's house over data centers was wrong. Nobody should harm another person over a policy disagreement, no matter how existential the stakes feel.

But Altman's blog post asks us to de-escalate rhetoric while his company escalates everything else. OpenAI will burn $85 billion this year. It is racing toward an IPO. It just acquired another company. It is building AI for the Pentagon. It is putting ads inside ChatGPT.

You do not get to build the most powerful technology in human history, tell the world it might end civilization, profit enormously from building it anyway, and then ask everyone to be chill about it.

Altman described himself in the post as a flawed person in the center of an exceptionally complex situation, trying to get a little better each year. That part rings true. The question is whether getting a little better each year is fast enough when the technology you are building does not move at the speed of personal growth. It moves at the speed of compute.

The Ring of Power is a perfect metaphor. In Tolkien, every person who picks it up believes they will use it differently. Every person who picks it up is wrong.