Sam Altman(@sama)
lisan say more mean things about us you're being too nice
2.0Score
TL;DR · AI 摘要
Sam Altman在X平台上的互动提及某用户应更直接批评，附带Lisan al Gaib关于GPT-5.5与Claude Mythos性能对比的推文。
核心要点
- Sam Altman鼓励用户lisan更直接地提出批评意见。
- GPT-5.5在特定任务上与Claude Mythos性能相近，解决专家级任务成本低、速度快。
- Lisan al Gaib通过LisanBench网站分享AI模型性能数据。
#Sam Altman#X平台#GPT-5.5#Claude Mythos#人工智能
打开原文Sam Altman on X: "lisan say more mean things about us you're being too nice" / X
Don’t miss what’s happening
Sam Altman 
lisan say more mean things about us you're being too nice
Quote
@scaling01
·
9h
GPT-5.5 is on par with Claude Mythos - GPT-5.5 average pass rate of 71.4% (±8.0%) - Mythos Preview 68.6% (±8.7%) - GPT-5.5 solved a task that takes a human expert ~12 hours in under 11 minutes at a cost of $1.73 x.com/AISecurityInst…
249
74
2.2K
188
Read 249 replies