Cubuk: Frontier AI labs haven't trained LLMs on physics or chemistry
Doge Cubuk · Building an AI Physicist: ChatGPT Co-Creator’s Next Venture · Sep 30, 2025 · at 29:53
Doge Cubuk, co-founder of Periodic Labs and former head of materials science research at Google DeepMind, discusses current limitations in frontier LLM reasoning abilities.
0:00 / 0:05exact quote · 5.9s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“The frontier AI labs have figured out how to train them on math and logic, but not yet on physics chemistry.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Doge Cubuk
Cubuk: A 200K superconductor would reshape fundamental quantum physics
“For example, if we could find a 200 Kelvin superconductor, even before we make any product with it, to be able to see such quantum effects at such high temperatures, I think would be such an update to people's view of how they see the universe.”
Cubuk: Quantum mechanics foundation models are AI's next frontier
“And we feel like teaching these LLMs to be foundation models, but for quantum mechanics will be the next frontier for LLMs.”
Cubuk: Periodic Labs is building an automated powder synthesis lab
“We're going to have a powder synthesis lab, and turns out this is one of those methods where robots can do it, like very cheap, simple methods.”
Cubuk: Physical experiments provide unhackable training signals for LLMs
“As we measure it, the LLM will get very clear signal. It's hard to hack, you know, unless, unlike these other LLM training techniques, it's like really what you see in real life is the signal that's going to the LLM.”
Cubuk: Out-of-domain AI scaling power-law slopes can be practically useless
“We published a paper where we saw that as you increase the size of your training set, the IID performance, the in-domain performance improves the power law. Out-of-domain performance also improves the power law, but depending on what the outdoor domain is, lik…”