ForeCast podcast

[AI Narration] Two Kinds of Test-Time Scaling

09/11/2026
0:00
31:47
Reculer de 15 secondes
Avancer de 15 secondes

This is an AI narration of "Two Kinds of Test-Time Scaling" by James Tillman. The article was first released on 9th November 2026.

You can read more of our research at forethought.org/research.

Outline:

(01:44) 1. Compute Extrapolation

(01:48) 1.1. What is it, and how does it relate to test-time scaling?

(10:14) 1.2. What would its consequences be?

(18:25) 2. Recent History

(29:28) 3. Conclusion

D'autres épisodes de "ForeCast"