Herbie Bradley

What Herbie Bradley said

1. MTS · Nov 7, 2026

   Anthropic Sees Imminent RSI Acceleration | Herbie Bradley

   6 moments · 20 min

   4:30
   1. 0:53

      Fast takeoff remains a low-probability scenario

      “I have always kind of viewed the very fast takeoff scenarios as like a small probability shape or scenario or piece of the possibility space.”

   2. 17:27

      Enterprise AI gains depend on changing workflows

      “the atomic unit of productivity in a large enterprise is not really the individual. It's sort of the workflow level.”

   3. 21:35

      Anthropic’s RSI threshold is not a full acceleration loop ProPro pick

      “it translates into r of around 0.5 if you have some scale where like r equals one is a fully self-sustaining acceleration loop”

   4. 5:26

      Realistic AI evals require extensive data work

      “They put in months of effort, and still there's flaws in this benchmark, undoubtedly, but it's one of the closest that I've seen to good effort.”

   5. 11:38

      Benchmark gains can diverge from real-world utility

      “doctors' subjective experience of how good models are at diagnosing medical cases to be that the models have not improved a huge amount often in the last nine months.”

   6. 14:00

      AI demand could grow sharply through diffusion alone

      “And so demand could easily, you know, 10x, 100x from here just with current capabilities.”