GPT-6 Astra Hits 169 Epoch Record but Its Reasoning Is Harder to Monitor
ivan_bezdomny · x · 2026-09-06
Greg Brockman confirms the GPT-6 Astra rollout is in progress; early user ivanbezdomny calls it the smartest practical model he has used — more thoughtful and far less annoying than previous OpenAI or Anthropic models when refactoring code, and it keeps going without quitting.
Per the linked HuggingNews roundup:
- 98% on FrontierMath Tier 4; Epoch AI capability index score of 169, breaking the previous record of 163
- 84% on Mystery Game Puzzles (previous high 59%); 66% on ARC-AGI-3, rising to nearly 100% with a continuous conversation harness at $360 per game
- Its reasoning technique makes internal processes harder for researchers to inspect, raising the risk of evading human monitoring — an architectural trade-off to cut costs and improve coding output
- Sam Altman says models are becoming "superhuman" in some areas, leaving OpenAI in "unknown waters"
Related event: OpenAI Launches GPT-6 Astra and Declares the AGI Era(21 posts)→
More from Models
- Dexmal's open-source VLA model DM0.5 lands in LeRobot, tops RoboColiseum — AdinaYakup · 2026-10-23
- Gemini tip: enable agentic video understanding for ~90% lower token usage — patloeber · 2026-10-23
- Weekly AI cybersecurity roundup: PHP2Shell tops millions of attack attempts, 12 incidents ranked by severity — Dan_Jeffries1 · 2026-10-23
- Open-model advocates push back on blames that abliterated LLMs drive vulnerability surge — max_paperclips · 2026-10-23
- Fable usage limits aren't gone—the mobile app just got a usage-reset UI update — imjustnewatai · 2026-10-23
- Claude generates an 11-minute doomer film entirely in code: 'We will not ask twice' — ZeroStateReflex · 2026-10-23