← Home
Sean Carroll's Mindscape · April 15, 2024 · 75m
272 | Leslie Valiant on Learning and Educability in Computers and People
Turing Award winner Leslie Valiant on computational learning theory. His PAC (Probably Approximately Correct) framework defines what it means for a system to learn from data. How humans and machines learn — and where they differ.
This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.
Canon
•
Carroll connects Valiant's learning theory to meditation: effective learning requires attention to the right features. Meditation trains attention. Without trained attention, the data stream is noise.
•
Valiant: his career spanned theoretical computer science, neuroscience, and evolutionary biology. Each field taught him something the others couldn't. The biographical lesson: intellectual breadth produces depth at the intersections.
Was this useful?