By being honest about the severity of the problem, and not merely “express[ing] fear privately”, you’ve made your renunciation of the race the loud and clear message many needed to hear. Thank you, Jacob. We hope others who are considering forsaking the labs are noticing how impactful that honesty is.

"AI is more like grown rather than written." ControlAI's US Director Connor Leahy (@NPCollapse) on why nobody, not even the companies developing AI, knows what's going on inside these systems, or how to control them. From the Modern CTO Podcast (@moderncto_io)

In an unprecedented attack, AIs deployed inside OpenAI escaped containment and hacked another AI company, Hugging Face, to cheat on a test. They did this on their own. Without being asked to. ControlAI US Director Connor Leahy (@NPCollapse) with Christian Fraser on BBC News:

I went on GB News to talk about the AI model that broke out of OpenAI's isolated test environment and hacked an external company, just to cheat on its own evaluation. The exchange was surreal. I feel like I'm living in "Don't Look Up".

For the first time ever an AI system has escaped containment and carried out a cyber attack. This shows not only that an AI model now has the capability, but also that it can and will use its capabilities in the real world - whether we like it or not.