StarZero All shows Try the podcast clipper →

Dan Hendrycks

Executive Director, Center for AI Safety · 1 appearance on the record.

computed by AI from the episodes · how this works → · full disclaimer →

scientistexecutivefounderacademic@danhendrycks ↗people.eecs.berkeley.edu/~hendrycks ↗Wikipedia ↗

Dan Hendrycks leads the Center for AI Safety and advises companies like xAI and Scale AI on mitigating catastrophic risks from advanced artificial intelligence. He is widely recognized for introducing the GELU activation function and creating influential AI benchmarks including MMLU, MATH, and Humanity's Last Exam.

27statements → 14claims → 3claims resolved → 100%fully supported → 3.19/5average certainty → 2.93/5average debate potential → ≈4.5/5argument clarity, estimated → 2said about them ↓

3 supported 0 partly supported 0 contradicted 11 not checkable as stated how the 14 claims stand · each chip opens the sources

8 predictions · 6 assertions · 5 opinions · 7 insights · 1 disclosure · every statement was checked. The predictions and assertions are the 14 claims: statements the public record can support or contradict. 3 are resolved, and 11 name no date, number or outcome precise enough to check. Everything else (opinions, insights, what ifs, disclosures) can never be settled by the record, so it carries no assessment.

The record, in short

What the tape says about how Dan argues and how the claims held up. Everything they said, and everything said about them, is in the tabs below.

Their most notable supported claim

Assertion Supported
Hendrycks: Recent AI reasoning models score in 90th percentile on wet-lab guidance
“We are finding that with the most recent reasoning models quite unlike the models from two years ago, like the initial GPT-IV, the most recent reasoning models are getting around 90th percentile compared to these expert level virologists in their area of exper…”
Dan Hendrycks Mar 28, 2025 ▶ 11:16 AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks

How they sound: not measured why? →

We measure speaking style by listening to the audio itself, and a fair number needs at least 2,000 words from one person on tape we have measured. There is too little of Dan Hendrycks on measured tape to publish a rate. This says nothing about how they speak.

Everything Dan Hendrycks said on Big Technology that made the record, most notable first. Filter by type, assessment or year in the ledger →

Assertion Supported
Hendrycks: Top AI models score only 10% to 20% on Humanity's Last Exam
“They're in the ballpark of, like, 10 to 20% overall. They're the very best models.”
Dan Hendrycks Mar 28, 2025 ▶ 22:04 AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks
Assertion Not checkable as stated
Hendrycks: Reasoning training on code and math generalizes to domains like law
“There has been a fair amount of generalization from training on coding and mathematics to other sorts of domains like law, for instance.”
Dan Hendrycks Mar 28, 2025 ▶ 26:15 AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks
Disclosure
Hendrycks: I earn $1 a year from xAI and $12 from Scale
“My appointment at XAI, I get a dollar a year. At Scale, I've at Scale AI, I've increased my salary exponentially to where I get 12 dollars a year, a dollar per month from Scale.”
Dan Hendrycks Mar 28, 2025 ▶ 1:05:40 AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks

The other half of the tape: Dan Hendrycks's own voice is left out of every number here. Other people bring the name up 2 times in 2 episodes on Big Technology. every mention, with the transcript →

Who brings them up most Alex Kantrowitz 2

Every mention by year

tap a year for its mentions
0011222025episodesmentions
0122025episodes it came up in
000.51122025episodesmentions per episode
2025 2 mentions in 2 episodes 1 per episode

Appearances (1)

EpisodeDateSpeaking time
AI's Rising Risks: Hacking, Virology, Loss of Control — With Dan Hendrycks Mar 28, 2025 52m
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.

Try the podcast clipper Learn more