Reflection on @dwarkesh_sp
's reflections on his interview with Rich Sutton, and why LLMs are exciting because they are Skinnerian, Popperian and Gregorian creatures. Minds with finite capacity cannot adapt forever without having to forget previous knowledge. This is true of
MACHINE LEARNING
-
LLMs as Skinnerian, Popperian, and Gregorian Creatures: Cognitive Limitations
By
–
-
FSD works without headlights because cameras see better at night
By
–
FSD works just fine without headlights. Cameras can see way more than you can at night. Take your iPhone out at night. You will see far more in the screen than your eyes can perceive.
-
World Model turns your house into a Holodeck
By
–
This is how a World Model turns your house into a Holodeck. https://t.co/veLd456wIH
— Robert Scoble (@Scobleizer) 12 décembre 2025This is how a World Model turns your house into a Holodeck.
-
Lightweight Discriminative Models: Overlooked ML Approach
By
–
Somehow I entirely missed the discriminative models that are lightweight, easy to understand, train, and infer.
-
Odyssey-2 Pro: Intelligent Interactive Real-Time World Model Available Soon
By
–
https://t.co/IrMUUAtjfg https://t.co/8fazB7BxN6
— Robert Scoble (@Scobleizer) 12 décembre 2025Say hi to Odyssey-2 Pro—our most intelligent, interactive, and general-purpose world model yet. Every pixel is generated in real-time, and frame-by-frame the model simulates the world forward. Today for minutes, soon for hours. Available to developers and our API very soon.
-
Gradual Tech Improvement Expected Until Field Maturity
By
–
Ongoing gradual minor improvement in product capabilities is what you'd expect in pretty much any tech field at pretty much any time, until that field is extremely mature and stable.
-
GPT-5.2 aces AIME math without tools
By
–
GPT-5.2 Thinking scored 100% on AIME (math without tools) Pretty good progress
-
NeurIPS 2025 Breakdown: Papers, Datasets, Workshops, Environments
By
–
Dive into the full #NeurIPS2025 breakdown — papers, datasets, workshop moments, and why environments will define 2026.
-
2026: AI Environments and Agent Systems Become Core Focus
By
–
2026 is shaping up to be the year of environments in AI.
From MCP-Bench and SEC-bench to DefenderBench and TheAgentCompany, NeurIPS 2025 made it clear: diverse, scalable environments are becoming core to agent systems, evaluation, and real-world reliability. Our #NeurIPS2025 -
Neural Operators Impact Across Science and Engineering
By
–
My talk on AI+Science hosted by the @ChenInstitute is now online https://
youtube.com/watch?v=Jyymk4
hptKk
… I go into details of neural operators and how they are impacting broadly across science and engineering