OpenAI's Chief Scientist Says We're Not Ready for AI's Risks

OpenAI's chief scientist just did something unexpected: he publicly warned that AI is moving too fast for us to handle safely. In an essay endorsed by OpenAI CEO Sam Altman, Jakub Pachocki says society is not prepared for what happens if machine intelligence keeps growing rapidly. This is a big shift in tone from the company's usual confident, we-got-this attitude. Pachocki isn't quitting his job or halting projects, but he's admitting that the people building these systems don't know how to fully control them.

So why should you care? Because AI is already writing emails, answering medical questions, and making decisions that affect your money and time. If the people building it say it could get dangerous, that's a signal for the rest of us to pay attention. Pachocki suggests that AI development should be limited by how safe we're sure it is. He also calls for outside regulation — government agencies, international bodies, or third-party auditors — to check whether AI companies are really following safety rules. That's notable because AI companies have mostly resisted oversight so far.

The essay is not an official policy change. It's an opinion by a key leader, shared publicly by the boss. Still, when the second-highest technical person at the world's most famous AI company says "we may not be on track" and "we need slower progress," that matters. Even if you don't trust OpenAI, this kind of internal doubt doesn't help their business. It's the opposite of marketing hype.

The honest catch: Pachocki is talking about future risks, not a problem that will visibly affect your phone tomorrow. Critics might say this is just a public-relations move to look responsible while continuing aggressive development. But at minimum, it shows that the people closest to AI are genuinely worried — and they think you should be too.