Loading activity…
@epochairesearch
Investigating the trajectory of AI for the benefit of society.
AI
A user mentions “harder things to RL on” and questions what would make Google “finally get serious.”
AI
A post says the improvement has come since its author and others began streaming the game.
AI
Epoch says Claude Fable 5.1 and GPT-6 Astra lead, but remain far from fully automating its work.
Science
OpenAI says many proofs have computer-checkable Lean formalizations, but warns that some unformalized results could have issues.
AI
Epoch AI fitted weekly figures through mid-August; Rohan Paul says the result reflects API list prices.
AI
EpochAIResearch says the histories span December 2022 to December 2025 and show the cohort using ChatGPT more intensively over time.
AI
Epoch AI Research says Opus 5.5 narrowly leads GPT-6 Astra, while Sonnet 5.5 roughly matched Claude Fable 5.1.
AI
Epoch AI says the top score on its Furniture Assembly Benchmark rose from 28% to 80% in just 10 months.
AI
Epoch AI labels nine of its initial 15 benchmarks “Flawed” and four “Verified,” with insufficient information to review two. A Terminal Bench contributor questions the audit’s rigor.
AI
Epoch AI report and expert reaction compare Huawei chip output to Nvidia through 2026.
AI
Pseudonymous commentator @scaling01 connects several scaling forecasts in one post.
AI
Executive at Hot Chips cites slides showing Groq 3 LPX hardware speed.
AI
Researchers argue official statistics miss Nvidia profits from fabless chip sales.
