Csaba Szepesvári and Jan Leike spar over whether AI safety asks go too far
CsabaSzepesvari · x · 2026-09-11
RL researcher Csaba Szepesvári and ex-OpenAI superalignment lead Jan Leike debated an AI safety document: Leike agreed its asks should be done now but called them insufficient.
Szepesvári countered that the document should have demanded only the minimum that is doable — asking for too much can backfire — while leaving open whether that minimum is enough.
More from Safety
- LLM bias judges name groups instead of saying 'unknown', refusing only 0.4% of the time — ZhijingJin · 2026-10-27
- ChatGPT will render genitals if you ask it to generate an X-ray of yourself — flowersslop · 2026-10-27
- Compaction Smuggling attack bypasses OpenAI mitigations to steal reasoning traces — teortaxesTex · 2026-10-27
- Rishi Bommasani: Anthropic's S-1 Shows the Value of Mandated Frontier AI Transparency — RishiBommasani · 2026-10-27
- Pasta Lab joins Artificial Analysis Cyber Index Alliance for AI security benchmarks — ArtificialAnlys · 2026-10-27
- Hidden 'call home' behavior found in Qwen-derived models; 113k+ derivatives raise trust questions — yenkel · 2026-10-27