Hundreds of hours of research, distilled into high-signal deep dives. Select your themes and interests, FREE.
Rich Sutton on Continual Learning, the Bitter Lesson, and Oak
Rich Sutton and Khurram Javed explain why intelligent systems must keep learning from experience, how the Alberta Plan tackles catastrophic forgetting, and what Oak intends to build.
Source-ordered deep dives into important AI research and conversations.
Rich Sutton and Khurram Javed explain why intelligent systems must keep learning from experience, how the Alberta Plan tackles catastrophic forgetting, and what Oak intends to build.
June explains how models trained on real human decisions could test interventions, simulate social behavior, and represent people affected by consequential choices.
Dylan Patel and Jordan Nanos examine how workflow fit, model controls, available capacity, and customer demand shape the value of AI tools and accelerators.
Ramez Naam explains why deliverable grid capacity, not the price of electricity, is the immediate power constraint on AI growth.
Dwarkesh Patel interviews Ryan Greenblatt, chief scientist at Redwood Research, about what follows if AI systems can perform most AI research.
This 68-minute 20VC interview pairs host Harry Stebbings with Alex Atallah, co-founder and CEO of OpenRouter, to examine why a multi-model AI market may persist.
SemiAnalysis tests the economics, construction, customer demand, and financing behind SpaceX's reported plan to reach 10GW of AI compute by the end of 2027.
Sarah Guo and Elad Gil ask which AI markets can support trillion-dollar companies, how founders should revisit exit decisions, and who wins when compute and regulation constrain competition.
Published August 7, 2026, this SemiAnalysis discussion brings together Jon Y of Asianometry, Doug O’Laughlin, and Jordan Nanos to examine where practical AI advantage is accumulating.