Want to dive into Scott Alexander's work and his thousands of blog posts? This fan website lets you sort and do semantic search through the whole codex. Enjoy!
| Date | ||
| Length | ||
| Likes | ||
| Comments |
Scott uses a thought experiment about a robot that shoots blue things to argue that we mistakenly interpret simple programmed behaviors as goal-directed utility maximization, when really the robot is just executing code, not minimizing blue.
Scott proposes a LessWrong community survey to gather demographics and opinions, shares a draft, and asks readers to suggest additional questions and weigh in on privacy concerns about releasing raw data.
Scott argues that "extreme rationality" (x-rationality) - the formal study of rationality beyond normal intelligent thinking - provides much less practical benefit than the rationalist community believes, estimating a correlation of only 0.1 with success.