test-time compute

also referred to as: test time compute

4 statements across 3 episodes · 1 bullish · 1 bearish · 3 people on the record · first statement May 15, 2025 by Jeremy Howard · across every show →

Everything said about test-time compute, oldest first

May 15, 2025 negative
Prediction Not checkable as stated
Howard: Test-time compute scaling will hit diminishing returns within two years
“It's a thing where you get most of the juice out of it in the first year or two, so we're still in that. Phase at the moment, and we'll start to hit the curve off point pretty soon. Just like we did for training.”
Jeremy Howard May 15, 2025 ▶ 9:56 Jeremy Howard on Building 5,000 AI Products with 14 People (Answer AI Deep-Dive)
Oct 2, 2025
Insight
Douglas: Test-time compute solves harder tasks before RL distills them into models
“Test time compute lets you do harder problems than you can currently do, than you can like currently do off the cuff, and RL Then allows you to sort of distill that back into the model.”
Sholto Douglas Oct 2, 2025 ▶ 51:52 Sonnet 4.5 & the AI Plateau Myth — Sholto Douglas (Anthropic)
Oct 2, 2025
Insight
Douglas: Test-time compute executes reasoning while RL provides feedback on correctness
“One way of thinking about this is test time compute is doing a lot of reasoning, and then RL is the feedback signal on whether or not that reasoning was right or wrong.”
Sholto Douglas Oct 2, 2025 ▶ 50:58 Sonnet 4.5 & the AI Plateau Myth — Sholto Douglas (Anthropic)
Jun 4, 2026 positive
Insight
Roberts: Powerful pre-trained models are necessary for effective RL and reasoning
“If you have a powerful enough pre-trained model, then it can start to do well at RL. It can start to like think at use test time compute to for instance, solve, solve math problems that it wouldn't otherwise be able to do.”
Dan Roberts Jun 4, 2026 ▶ 27:15 OpenAI's Dan Roberts: Why AI Can Now Make Discoveries
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 conversations transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.