04-03-2026Price:

The Frontier

Your signal. Your price.

AI & TECH

AI needs hands to revolutionize material science

Friday, April 3, 2026 · from 1 podcast
  • AI cannot engineer new materials without controlling physical lab experiments.
  • Scientific literature is too unreliable to train effective models.
  • High-energy physicists now lead AI, applying physical rigor to neural networks.

AI has mastered language, but atoms remain out of reach. Liam Fedus, co-founder of ChatGPT and now of Periodic Labs, argues that AI's next leap requires literal hands - robots to conduct real-world chemistry and physics experiments.

Fedus spent years scaling large language models, but he sees a fundamental bottleneck. Scientific papers, the presumed training data for material science, are often wrong. Values for the same material can vary by orders of magnitude. Training a model on this data, Fedus says, only reproduces human error and confusion.

The solution is a closed loop. At Periodic Labs, AI directs robotic labs to synthesize and test materials, capturing ground-truth data to refine its own models. This moves AI from pattern recognition in text to active intervention in the physical world, aiming to design new semiconductors and aerospace alloys.

This shift is attracting a specific kind of talent. Fedus notes that many leading AI researchers, from Anthropic's Dario Amodei to himself, are former high-energy physicists. After the Higgs boson discovery, that field hit a hardware wall. AI offered a new frontier where first-principles thinking could yield immediate, measurable results.

These physicists treat neural networks as physical systems to be measured and optimized, bringing a hard-nosed experimental rigor to model training. They apply scaling laws to predict performance, treating software with the discipline of a lab experiment.

Liam Fedus, No Priors:

- Science ultimately isn't sitting in a room thinking really hard.

- You have to conduct experiments to interface with reality.

The ambition is a productivity spike for the physical world, akin to an agricultural revolution for materials. But the feedback loop will be slower, constrained by the stubborn pace of atoms, not bits.

Source Intelligence

What each podcast actually said

AI for Atoms: How Periodic Labs is Revolutionizing Materials Engineering with Co-Founder Liam FedusApr 3

  • Many AI researchers like Dario Amodei and Adam D'Angelo have physics backgrounds, a trend Liam Fettis attributes to physicists' principled thinking and high leverage in AI.
  • Periodic Labs is building an AI foundation lab for atoms, focusing on applying AI to material science, chemistry, and the physical world.
  • Periodic's AI system acts as an orchestration layer, using large language models to direct experiments and specialized neural nets designed for atomic systems.
  • Fettis says the acceleration of digital software engineering creates an imperative to connect AI systems to the physical world for scientific and technological progress.
  • Current AI technology, including improved reasoning and reliable tool use, is now sufficiently advanced to connect AI to the physical world, unlike in 2022.
  • Periodic leverages existing models for coding and language, spending zero effort on improving them, to focus its machine learning efforts on physical world frontiers.
  • A key data challenge in materials science is that reported property values from literature often span orders of magnitude, making ground truth difficult to establish without experiments.
  • Periodic's approach relies on an interactive closed-loop system where experimental data feeds back to identify aberrations and patterns to drive the next experiments.
  • Data generalization for physical systems is often domain-specific; a model trained on quantum mechanical objects doesn't help much with fluid dynamics.
  • Fettis sees the most internal advances where there is an abundance of data in a specific chemical or material space.
  • The biggest capital cost for Periodic's work is GPU compute, not physical infrastructure, though lab setup has long lead times and calibration difficulties.
  • Fettis believes AI systems exhibit odd spikiness in intelligence, being world-class in one domain but potentially poor in adjacent ones, challenging the idea of intelligence as a scalar.
  • Software engineering self-improvement by AI is happening now due to cheap, verifiable environments like unit tests, but this doesn't automatically translate to other domains like biology.
  • Fettis says AI research self-improvement is a slower outer loop than software engineering because experiments require GPUs and hours to evaluate model convergence and scaling properties.
  • While reliable robotics would be a huge accelerator, Periodic currently uses hybrid human-automation systems and off-the-shelf robotics to generate sufficient high-throughput data.
  • Fettis views the future as one where AI generates matter, profoundly impacting semiconductors, aerospace, and energy by increasing the pace of physical world development.
  • The multidisciplinary collaboration at Periodic, with physicists, chemists, AI researchers, and engineers, is allowing veteran scientists to see their fields fundamentally change.

Also from this episode:

Robotics (1)
  • Fettis sees the interface of AI with the physical world via robotics as a transformative opportunity, given labor shortages and the vast number of people who work with the physical world.