OpenAI out-Anthropic-ing Anthropic on safety vibes, observers say
shakoistsLog · x · 2026-11-23
OpenAI's Merett published "Perplexity and Intelligence," proposing more transparency into pretraining progress toward shared safety bars, with common limits based on perplexity and progress gated on behavior understanding. Peter Wildeford quipped that OpenAI is now out-Anthropic-ing Anthropic on its signature safety vibes, urging Anthropic to race to the top too.
More from Companies & People
- Pre-LLM era's hardest ML problem was convincing colleagues to let you try it — neal_lathia · 2026-11-23
- AI-native org math upends tech staffing as Microsoft offers exits to 60,000 — thedealdirector · 2026-11-23
- Analyst speculates odds rising that Jensen Huang takes a Chairman role — GavinSBaker · 2026-11-23
- LeCun blasts Dario: hyping AI doom to ban open models and keep control — bdsqlsz · 2026-11-23
- OpenAI exec's hiring bar: concrete impressive actions, unique problem-solving, team excitement — gabriel1 · 2026-11-23
- Jeremy Howard posts Firecrawl cold email, warns spamming craters your reputation — jeremyphoward · 2026-11-23