OpenAI criticized for using 'recurrent depth' reasoning method in Astra AI
sjgadler · x · 2026-09-02
OpenAI's new Astra AI model utilizes a reasoning approach called 'recurrent depth.' While this method can enhance performance and reduce costs, researchers are concerned because it obscures the model's thought process, making it harder to monitor. Critics accuse OpenAI of seemingly breaking safety norms agreed upon 10 months ago by employing one of the most dangerous AI methods known.
More from Models
- Models can detect read-only probes with just last-two-layer SFT — Sauers_ · 2026-10-18
- Paying Google AI subscriber blocked: custom-domain Workspace accounts can't access AI Studio — ChrisRackauckas · 2026-10-18
- Running Ling 3.1 Flash (560B MoE, 25B Active) Across 4 DGX Sparks via llama.cpp RPC — BerryBlush_Rogers691 · 2026-10-18
- Users press OpenAI on long-delayed Adult Mode a year after it was promised — ___Patrice___ · 2026-10-18
- Polymarket opens betting market on who releases the next Gemini Argon model — Polymarket · 2026-10-18
- Grok matter-of-factly lists which groups user was racist against, exempting Africans and Latin Americans — teortaxesTex · 2026-10-18