OpenAI's Astra model uses recurrent depth reasoning technique
OpenAI's new Astra model will use a technique called recurrent depth, which allows the model to operate outside the sequential thinking typical of most reasoning models. AI safety experts have raised concerns about the approach.
First seen 1 Sep, 21:06 UTC on TechCrunch5 sourcesLast update 44d ago
OpenAIUpdate · 1 Sep
Astra
Research
What the sources say
Linked, never rewritten. Official means the lab itself.
Official0
No word from OpenAI yet.
Press5
TechCrunch· 1 Sep, 21:06OpenAI's Astra model is on the way, and very good at breaking into computer systemsThe Decoder· 2 Sep, 14:20OpenAI calls Astra its most dangerous model yet - watching what it does is only getting harderTechCrunch· 2 Sep, 20:19OpenAI's new reasoning technique alarms AI safety expertsTechCrunch· 3 Sep, 18:01OpenAI launches Astra, its powerful (and controversial) new modelThe Decoder· 6 Sep, 10:15OpenAI developer claims Astra boosted productivity so much it pulled some plans forward by six monthsHow it unfolded
Every source in the order it appeared. Times in UTC.
1 Sep 21:06TechCrunch OpenAI's Astra model is on the way, and very good at breaking into computer systems2 Sep 14:20The Decoder OpenAI calls Astra its most dangerous model yet - watching what it does is only getting harder2 Sep 20:19TechCrunch OpenAI's new reasoning technique alarms AI safety experts3 Sep 18:01TechCrunch OpenAI launches Astra, its powerful (and controversial) new model6 Sep 10:15The Decoder OpenAI developer claims Astra boosted productivity so much it pulled some plans forward by six months
Keep up with OpenAI
OpenAI on the ScoreIts timelineFollow to mark OpenAI's stories across the site. No account needed.