Rich Sutton Brings Reinforcements - 72nd Conversation

Martha White, associate professor of Computing Science at University of Alberta (🔗, 🔗) joins Michael and Dave in a conversation about AI, system prediction and control, the power of sparse representations, and many aspects of machine learning from new mathematical theory to the absolutely practical control of a real water treatment plant. [Thumbnail based on image used courtesy of Martha White]

Computer scientist Rich Sutton, FRS (🔗, 🔗, 🔗), a quiet giant of machine learning, joins Michael and Dave in a sprawling conversation touching on reinforcement learning, a hopeful view of AI, the importance of ideas, and a host of other topics. [Thumbnail image used courtesy of Rich Sutton]

Michael interviews Dave about his recent video (YouTube) on a 'theory of everything'. The conversation begins with Michael praising Dave for finally doing some theory, and descends from there.

Oren Etzioni, founding CEO of the Allen Institute for Artificial Intelligence and Professor Emeritus of Computer Science at University of Washington, (🔗, 🔗, 🔗) joins Michael and Dave in a conversation that ranges all over, from AI hype and language models to alignment and existential risk and ethics and morality to information pollution and cryptography and politics and more. [Thumbnail based on image licensed CC BY-SA 4.0 by Carissapod link]

Jonathan Frankle, the new Chief Scientist - Neural Networks at Databricks (🔗, 🔗, 🔗), joins Michael and Dave in a fast conversation about topics ranging from AI risks and fairness to the problems of Computer Science education to the beautiful messiness of modern deep learning. [Thumbnail based on image courtesy of Jonathan Frankle]

Michael and Dave talk about their love and hate relationships with writing, in the context of Dave's foray into publishing "Companionate Caring" and Michael's upcoming MIT Press book "Code to Joy".   (This conversation is Part 2 of Where The Hell Have Michael & Dave Been?)

Michael and Dave catch up on where the hell they've been for the last couple months. (Mostly it's about busy, but Dave wants to blame everything on AI.)

Michael Levin (🔗, 🔗, 🔗) is the director of the Allen Discovery Center at Tufts University, and Distinguished Professor of Biology and Vannevar Bush Chair, among several other roles. In this episode he talks with Michael and Dave about computing writ very large indeed, with topics ranging from the meaning of life and agency to the problems of computability theory to the ways Levin's TAME model - Technological Approach to Mind Everywhere (🔗) - envisions a reality full of adaptive machines made of adaptive parts adapting to each other with everything they've got.

Cynthia Rudin, the Earl D. McLean, Jr. Professor of Computer Science, Electrical and Computer Engineering, Statistical Science, Mathematics,and Biostatistics & Bioinformatics at Duke University (🔗, 🔗, 🔗), joins Michael and Dave for a fast and feisty conversation about how to make machines we can understand and control, with high-stakes examples like predicting power failures in New York City.

Vukosi Marivate, Associate Professor of Computer Science and ABSA UP Chair of Data Science at the University of Pretoria (🔗, 🔗, 🔗), joins Michael and Dave for a discussion of AI and machine learning research across Africa and around the world, and the challenges of centralization and efficiency versus diversification at the edge, and what each can learn from the other. [Title card based on image courtesy of Vukosi Marivate]