首页 > 社区 > LessWrong LessWrong 共 53 条资讯 Who Got Breasts First and How We Got Them Trees are mostly made of air and a generalizable lesson for AI safety Guardian Angels: LLM Personalization for Productivity and Security Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models x-risk-themed llm assistant personas seem increasingly incoherent (some subjective observations) Bad Problems Don't Stop Being Bad Because Somebody's Wrong About Fault Analysis Empowerment, corrigibility, etc. are simple abstractions (of a messed-up ontology) Models finding software vulnerabilities is not the primary source of cybersecurity risk Mnemonic portraits for 19,023 human genes My hobby: running deranged surveys Irretrievability; or, Murphy's Curse of Oneshotness upon ASI What I did in the hedonium shockwave, by Emma, age six and a half How Go Players Disempower Themselves to AI Intelligence Dissolves Privacy The Terrarium Do not conquer what you cannot defend Annoyingly Principled People, and what befalls them Current AIs seem pretty misaligned to me Morale 123下一页 » 相关分类 36氪·新商业观察 9GAGAll Posts Announcements - Dynalist Forum CNode社区 CTA南美水族家园 Chiphell Chiphell品质家居 Chiphell品质生活 Chiphell影像器材 Chiphell掌设笔电 Chiphell摄影作品 Chiphell新闻发布 Chiphell汽车摩托 Chiphell电脑硬件 Chiphell电脑讨论 Chiphell美食饕餮 Chiphell视听发烧 Chiphell软件系统交流 EETOP 创芯网论坛最新热门 Emacs China Emacs China最新主题 JIKEPopular Topics JIKE活动优惠 JIKE账号密钥 KeebTalk MT论坛