Hendrycks: Anthropic has almost captured the ecosystem meant to hold it accountable
zetalyrae · x · 2026-11-19
Dan Hendrycks says numerous AI safety groups privately agreed with his statement but felt unable to sign without risking relationships or future funding, and that Anthropic has "almost, but not entirely, captured the ecosystem meant to hold it accountable." The retweeter notes this is becoming common knowledge, sparking debate over the independence of the AI safety ecosystem.
More from AGI Musings
- Mustafa Suleyman interviews Yuval Noah Harari on AI abundance and its side-effects — mustafasuleyman · 2026-11-19
- Healthcare CEOs describe an 'AI cold war' as hospitals and insurers deploy competing AI — DrDatta_AIIMS · 2026-11-19
- Nathan Lambert: AI safety community shouldn't treat Anthropic as the only voice — natolambert · 2026-11-19
- Erik Hoel: engineering non-conscious AI is a civilization-level requirement — erikphoel · 2026-11-19
- Cornell Tech's Ayush Sekhari is recruiting PhD students in AI via Cornell CS — yoavartzi · 2026-11-19
- When should frontier AI labs pause? A decision framework says short and targeted — Manderljung · 2026-11-19