#64 Prof. Gary Marcus 3.0
Machine Learning Street Talk (MLST)
00:00
Abstract Models and AI Reasoning Challenges
This chapter discusses abstract world models and compositional structures in artificial intelligence, focusing on how human cognitive processes influence these concepts. It highlights the difficulties AI faces with stable symbolic representations and the significant distinction between memorization and genuine reasoning capabilities. The conversation also addresses practical implications in fields like self-driving cars, stressing the need for robust representation systems to navigate complex scenarios.
Transcript
Play full episode
