OpenAI leaks mysterious model codename, Altman slams doomsday marketing
新智元 · wechat · 2026-08-24
OpenAI employee accidentally leaked model codename 'gpt-nathree' on GitHub, likely a checkpoint of GPT-6 Astra; Anthropic's new models 'claude-marshmallow-eap' and 'claude-melon-eap' exposed. Meanwhile, Altman admitted misjudging AI disruption timeline, criticized doomsday marketing, and emphasized human-centric values.
More from Companies & People
- Breaking waterfalls into short Sprints forces automation and leads to continuous delivery — JoeJustice · 2026-08-27
- Google runs 1M+ automated tests per second; human validation can't keep up with AI — JoeJustice · 2026-08-27
- The Future of Enterprise Apps: UI Is Overrated, Context Is King — matt_slotnick · 2026-08-27
- Gates should invest in farming robots instead of AI fear — mitchdeg · 2026-08-27
- Scrum Tip: Break down large items, replace docs with conversation — JoeJustice · 2026-08-27
- Countering NVIDIA Investment Doubts: Six Years of Dominance and Revenue Growth — cis_female · 2026-08-27