AI Might Be Conscious, But Not As We Thought
Sabine Hossenfelder
Save 5 minutes
Original Video
Summiz
6 min
Original1 min
with Summiz5x
faster learning efficiencyBrief summary
AI Might Be Conscious, But Not As We Thought explores evidence of introspection in language models like Claude and Llama, arguing that consciousness may emerge during training while emphasizing the need for reliable ways to measure it.
Notion, Obsidian or Markdown.
