The next big breakthrough will be AIs learning on the job

A clear-eyed argument for why continual learning — getting model improvements back into the weights rather than just context — is the bet frontier labs are making, with concrete framing (RLVR generalization, verifiability vs grindability) that helps engineers reason about where agent capabilities are headed by 2027.

• context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
• AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.

• 00:00:00The big research bet the labs are making
• 00:02:12Grindability is just as important as verifiability
• 00:06:10Will RLVR alone generalize?
• 00:08:41Getting the learning back to the weights
• 00:15:22Dreaming
• 00:17:23What 2027 looks like

• podcastThe data black hole at the center of AI
• podcastThoughts on AI progress (Dec 2025)
• podcastSatya Nadella — How Microsoft is preparing for AGI
• podcastSatya Nadella — Microsoft’s AGI plan & quantum breakthrough

• podcastDario Amodei — "We are near the end of the exponential"Dwarkesh Patel
• videoLearning on the Job: The Future of Post-Training — Raymond Feng, Applied ComputeAI Engineer
• videoWhat does the next training paradigm look like?Dwarkesh Patel

Checking sign-in…

Loading comments…