AI benchmarks

2 statements across 2 episodes · 0 bullish · 1 bearish · 2 people on the record · first statement Dec 11, 2024 by Alexandr Wang · across every show →

Everything said about AI benchmarks, oldest first

Dec 11, 2024 neutral
Assertion Not checkable as stated
Wang: AI benchmarks are saturated, making model leaders hard to distinguish
“One thing that we see today with the models is that because all the benchmarks that were used today are what's called saturated, i.e., you know, in other words, like all the models do really well at the benchmarks, it's really hard to discern actually which on…”
Alexandr Wang Dec 11, 2024 ▶ 53:56 AI Predictions for 2025: Geopolitics, Agents, and Data Scaling — With Alexandr Wang
May 14, 2025 negative
Insight
Bender says most AI benchmarks fail to measure actual capabilities
“Most of the benchmarks that are out there are not reasonable. They lack what's called construct validity, and construct validity is this two-part test of the thing that we are trying to measure is a real thing, and this measurement correlates with it interesti…”
Emily M. Bender May 14, 2025 ▶ 14:30 AI’s Drawbacks: Environmental Damage, Bad Benchmarks, Outsourcing Thinking
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes 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.