Unsupervised Learning Concerns
Dylan expresses concerns about unsupervised learning, highlighting the importance of understanding what the system should do and how to verify its actions. The discussion delves into the need for normative considerations in AI development and the critical questions surrounding data curation for unsupervised learning.In this clip
From this podcast
Generally Intelligent
Episode 10: Dylan Hadfield-Menell, UC Berkeley/MIT, on the value alignment problem in AI
Related Questions