Skip to content
Stanford CS153 Frontier Systems | Teaching AI to Touch Atoms
This video isn’t embeddableWatch on YouTube →
Stanford OnlineStanford Online

Stanford CS153 Frontier Systems | Teaching AI to Touch Atoms

For more information about Stanford's online Artificial Intelligence programs, visit: https://stanford.io/ai Follow along with the course schedule and syllabus; visit: https://cs153.stanford.edu/ In this CS 153 Frontier Systems lecture, Periodic Labs cofounders Liam Ferriss and Dorje Chubak — veterans of OpenAI's post-training team and DeepMind's Genome project, respectively — walked students through their eleven-month-old mission to apply AI to the physical world, specifically using autonomous labs to accelerate the discovery of new materials including high-temperature superconductors. Operating out of a 40,000-square-foot Menlo Park facility where machine learning researchers work alongside physicists and chemists, their AI system (named Onnes, after the scientist who discovered superconductivity in 1908) runs a continuous loop of computational prediction, robotic synthesis, and experimental verification — closing the feedback cycle between digital intelligence and physical reality in a way no purely in-silico approach can. A key early lesson was abandoning their original plan of spending the first year purely computational: building smaller, semi-manual labs first allowed them to direct the research program and understand what to scale, faster. Dorje emphasized that the results of applying LLMs to actual atoms have exceeded even their own expectations, while Liam stressed that sample efficiency in reinforcement learning — not benchmark climbing — is the core technical frontier when physical experiments can't be arbitrarily scaled up the way digital rollouts can. They closed by pushing back on student anxiety about AGI displacing their careers, arguing that most scientific domains remain largely untouched by LLMs, that the bar to make meaningful AI-physical world progress is still surprisingly low, and that the history of civilization is essentially a materials story — making their work, in their view, among the highest-leverage bets anyone can make right now.

Dorje Chubakguest
Sep 29, 202642mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Periodic Labs builds AI-robot loops to discover superconductors and semiconductors.

  1. Periodic Labs is building an AI-driven, robot-assisted materials R&D stack that iteratively proposes, synthesizes, characterizes, and learns from real-world experiments to accelerate discovery in superconductors and semiconductors.
  2. A major lesson from their first year is that early, smaller semi-autonomous lab loops beat a phased ‘compute first, robots later’ plan because they rapidly expose practical bottlenecks and produce grounding data.
  3. They argue that progress hinges less on flashy grand predictions and more on automating mundane scientific operations—sample tracking, impurity control, and interpretation of instruments like X-ray diffraction—where errors commonly derail results.
  4. Their methodology emphasizes active learning and sample-efficient model-based RL, while expressing skepticism that Bayesian optimization’s uncertainty estimates remain reliable under scientific distribution shift.
  5. They frame the broader mission as steering powerful AI toward positive, physically grounded impact—improving humanity’s ability to engineer matter—while acknowledging that full end-to-end autonomous discovery remains an unsolved, incremental systems challenge.

IDEAS WORTH REMEMBERING

5 ideas

Start with small, messy lab loops—not a ‘big lab switch-on’ plan.

They expected to spend the first year doing purely computational “in silico” work, then later turn on a high-throughput autonomous lab. Instead, they found that building smaller semi-manual/semi-autonomous loops early was essential for steering the research program, selecting the right equipment to scale, and closing the experiment–model feedback loop fast.

Scientific progress is dominated by error-correction and workflow details that AI can automate.

Beyond predicting promising superconductors/semiconductors, the AI is valuable for mundane but consequential tasks: detecting sample mix-ups, controlling impurities, choosing mixing procedures, and interpreting characterization outputs (e.g., XRD patterns). These ‘local error corrections’ compound into real scientific throughput.

Reality contact (experiments) is the differentiator for AI that ‘touches atoms.’

They argue that models won’t ‘think’ their way to new materials from textbooks alone; they must be grounded in experimental interaction. New data from synthesis and characterization is treated as core training signal to deepen a model’s physical understanding and improve future decisions.

Prefer active learning over Bayesian optimization for messy, shifting scientific domains.

Dorje claims Bayesian optimization often disappoints in practice because estimating uncertainty under distribution shift is even harder than point prediction. By contrast, active learning is critical in real-world settings where the model is competent in some regions and clueless in others; they run this style of iterative frontier-expansion daily via experiments.

Pick a high-impact physical direction (atoms/electrons) rather than chasing abstract ‘AGI’ generality.

They choose superconductors/semiconductors because both hinge on electron–atom interactions and interfaces, are central to modern technology, and face urgent materials bottlenecks (e.g., energy loss/heat in chips, scaling limits). The bet is that pushing AI hard on this ‘important direction’ beats hoping for magical generality.

WORDS WORTH SAVING

5 quotes

Scientific progress is like a bunch of mundane things attached to each other.

— Liam Ferriss

There’s something amazing happens when you have automated intelligence affecting not just Python, but the actual atoms that surround you.

— Dorje Chubak

We don’t think that you can just start thinking your way to a new room temp superconductor by reading a textbook, shutting it, and thinking super hard. You have to make contact with reality. You have to carry out experiments.

— Liam Ferriss

Onnes was one of the first people who said scientific research should be done at industrial scale.

— Dorje Chubak

If you can automate science, there’s no end to it.

— Dorje Chubak

Autonomous high-throughput labsAI-to-robot experimental feedback loopsSuperconductors vs semiconductors motivationActive learning in real scientific domainsSynthesis planning and synthesizabilityXRD/characterization automationSample-efficient model-based reinforcement learning

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.