Reform AI Alignment
Scott discusses the potential dangers of AI going off a cliff and the possibility of deceptive alignment, shedding light on Eliezer's recent comments and the reactions to GPT-4.In this clip
From this podcast
The Gradient
Scott Aaronson: Against AI Doomerism
Related Questions
Can AI become superintelligent as discussed in the episode 159 - We’re All Gonna Die with Eliezer Yudkowsky, the episode Eliezer Yudkowsky: Dangers of AI and the End of Human Civilization | Lex Fridman Podcast #368, and the clip Alignment and AI Risks, as well as in the episode Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality and the clip Human-Level Intelligence and AI Alignment?
Can AI become superintelligent as discussed in the episode Eliezer Yudkowsky: Dangers of AI and the End of Human Civilization | Lex Fridman Podcast #368 and the clip Alignment and AI Risks, as well as in the episode Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality and the clip Human-Level Intelligence and AI Alignment?
Can AI become superintelligent as discussed in the episode Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality and the clip Confirming Solutions in Alignment?