Interconnects

109 Episodes

1. The White House's plan for open models & AI research in the U.S.

   Published: 7/23/2025
2. Kimi K2 and when "DeepSeek Moments" become normal

   Published: 7/14/2025
3. The American DeepSeek Project

   Published: 7/4/2025
4. Some ideas for what comes next

   Published: 6/23/2025
5. Crafting a good (reasoning) model

   Published: 6/18/2025
6. The rise of reasoning machines

   Published: 6/12/2025
7. What comes next with reinforcement learning

   Published: 6/9/2025
8. How I Write

   Published: 6/6/2025
9. A taxonomy for next-generation reasoning models

   Published: 6/4/2025
10. Claude 4 and Anthropic's bet on code

    Published: 5/27/2025
11. People use AI more than you think

    Published: 5/21/2025
12. My path into AI

    Published: 5/14/2025
13. What people get wrong about the leading Chinese open models: Adoption and censorship

    Published: 5/6/2025
14. State of play of AI progress (and related brakes on an intelligence explosion)

    Published: 4/30/2025
15. Transparency and (shifting) priority stacks

    Published: 4/28/2025
16. OpenAI's o3: Over-optimization is back and weirder than ever

    Published: 4/19/2025
17. OpenAI's GPT-4.1 and separating the API from ChatGPT

    Published: 4/14/2025
18. Llama 4: Did Meta just push the panic button?

    Published: 4/7/2025
19. RL backlog: OpenAI's many RLs, clarifying distillation, and latent reasoning

    Published: 4/5/2025
20. Gemini 2.5 Pro and Google's second chance with AI

    Published: 3/26/2025