Over 1,100 frontier AI employees sign statement highlighting control risks
sjgadler · x · 2026-07-30
1,122 employees from frontier AI companies, including OpenAI, Anthropic, and Google, have signed the "Pacing the Frontier" statement.
Signatory Luke Drago elaborated on his decision to support the statement. He noted that while good progress has been made on alignment and current LLMs mostly do what we want, recent security incidents (like the HuggingFace incident) show that current approaches might not suffice for more powerful intelligences. Rushing to advance the frontier could cause us to miss signs of losing control. Although we can recover from these mistakes today, discussing slowdown mechanisms proactively is necessary to prevent irreversible risks in the future.
Related event: Over 1,100 Frontier AI Employees Petition for Government Regulation(116 posts)→
More from AGI Musings
- AI researchers are like LeBron at 19 with only 18 months on the shot clock — adityaag · 2026-07-30
- Opinion: Belief in Inevitable ASI Is a Psychological Cope to Avoid Disaster — zetalyrae · 2026-07-30
- AI governance could spiral out of control: human frameworks struggle to keep up — LuizaJarovsky · 2026-07-30
- Brainless Slime Molds Show Learning and Memory, Challenging Intelligence Boundaries — anilkseth · 2026-07-30
- NSF Launches Industry-Embedded PhD Program Amid Calls to Rethink Academia — kevinnbass · 2026-07-30
- Alexandr Wang at YC: From Scale AI to Meta's Superintelligence Lab — Y Combinator · 2026-07-30