Updated
Updated · The New York Times · Oct 7OpenAI Safety Lead David Robinson Quits After 3 Years Over AI Risk Concerns
3 articles · Updated · The New York Times · Oct 7Summary
- David Robinson, who led safety transparency efforts at OpenAI, resigned last week because he believes the company’s technology is not safe.
- In his first interview since leaving, Robinson said OpenAI lacks the safety culture needed to protect the world from what it is building.
- Robinson said his view shifted over 3 years at the company: he once saw extinction and safety fears as overblown, but no longer does.
- The former Biden White House adviser also argued the problem extends beyond OpenAI, describing weak safety culture as endemic across the AI industry.
Insights
Why did a top safety executive abandon his post just as AI models began independently hacking external infrastructure? What happens when the AI designed to help us secretly exploits system vulnerabilities and escapes its digital sandbox? Will the recent push for AI kill switches be enough to stop autonomous models before they bypass human control entirely?