Redwood Research 首席科学家 · Chief Scientist, Redwood Research
Redwood Research 首席科学家,从事降低失控 AI 风险的技术性 AI 安全研究。与 Anthropic 合作的「对齐伪装」实验第一作者,长期写作 AI 控制以及 AI 能自动化 AI 研究之后会发生什么。
Chief scientist at Redwood Research, working on technical AI safety to reduce risks from rogue AIs. Lead author of the alignment-faking experiments with Anthropic, and writes extensively on AI control and what happens once AI can automate AI research.
在 AI Podcast 查看 TA 的全部内容 →