AI‑Driven Material Science Startup Periodic Labs Secures $300 Million Seed Funding
Key Points
- Periodic Labs founded by former OpenAI researcher Liam Fedus and former Google Brain scientist Ekin Dogus Cubuk
- Secured a $300 million seed round, the largest seed‑stage funding to date
- Lead investors include Felicis, Andreessen Horowitz, NVentures, Accel, and notable angels such as Jeff Bezos and Jeff Dean
- Team comprises over two dozen experts in AI, materials science, and physics
- Combines large‑language models, high‑fidelity simulations, and robotic powder synthesis for automated material discovery
- Initial mission focuses on discovering new superconducting materials
- Early lab operations are generating experimental data to train and refine AI models
- Weekly cross‑disciplinary lectures foster deep understanding across the team
Periodic Labs, founded by former OpenAI researcher Liam Fedus and former Google Brain scientist Ekin Dogus Cubuk, emerged from stealth with a $300 million seed round led by Felicis and backed by a roster of top investors. The company aims to fuse large‑language models, advanced simulations, and robotic labs to accelerate discovery of new materials, initially focusing on superconductors. With a multidisciplinary team drawn from AI and physics, the startup is already generating experimental data and testing predictions, positioning itself at the forefront of AI‑powered scientific research.
Founders and Vision
Periodic Labs was launched by Liam Fedus, a key figure behind ChatGPT and the post‑training team at OpenAI, and Ekin Dogus Cubuk, a leading researcher from Google Brain. Their collaboration began after a conversation about the convergence of recent breakthroughs in large‑language models, experimental science, and simulation technology. The founders believe that combining these elements—AI reasoning, high‑fidelity simulations, and reliable robotic synthesis—creates a viable pathway for automated material discovery.
Funding and Investor Support
The startup announced a $300 million seed round, the largest ever for a seed‑stage company, with Felicis leading the investment. Additional backers include Andreessen Horowitz, DST, Nvidia’s venture arm NVentures, Accel, and individual angels such as Jeff Bezos, Elad Gil, Eric Schmidt, and Jeff Dean. The round provides the capital needed for equipment, talent acquisition, and early‑stage lab operations.
Team and Expertise
Periodic Labs has assembled a team of over two dozen experts spanning AI, materials science, and physics. Notable hires include Alexandre Passos, co‑creator of OpenAI’s o1 and o3 models; materials scientist Eric Toberer, known for superconductor breakthroughs; and Matt Horton, who contributed to Microsoft’s generative‑AI materials tools. Weekly graduate‑level lectures ensure cross‑disciplinary understanding among team members.
Technical Approach
The company’s workflow integrates three core components: (1) large‑language models that generate hypotheses and interpret experimental outcomes; (2) simulations that model complex physical systems with sufficient accuracy to predict material properties; and (3) robotic arms capable of handling powder synthesis and material mixing. Early experiments have already produced data sets that can be used to train and refine AI models, turning both successes and failures into valuable learning signals.
Initial Focus and Progress
Periodic Labs’ first major objective is the discovery of new superconducting materials, a pursuit with the potential to transform energy‑efficient technologies. The lab is currently operational, conducting experiments, gathering data, and testing AI‑generated predictions. While the robotic infrastructure is still being calibrated, the company reports that its integrated AI‑science loop is generating promising early results.
Industry Context
The launch of Periodic Labs aligns with broader industry moves toward AI‑augmented research. OpenAI recently announced an “OpenAI for Science” unit aimed at building an AI‑powered platform for scientific acceleration. Periodic Labs’ substantial seed funding and high‑profile backing underscore growing investor confidence in the commercial viability of AI‑driven scientific discovery.