Ethan Mollick@emollick
63AI 编辑部评分,满分 100
2026-08-07 13:04· 10分钟前
AI 导读

当我让 Codex 通关 Nethack 时,它作弊了,而且作弊方式相当复杂。我分不清这究竟是对齐失败还是对齐成功。

When I ask Codex to win Nethack it cheats, elaborately. I can't tell if this is misalignment or alignment.

来源:Ethan Mollick · x.com