Bio
Co-founder and Chief Science Officer of Hugging Face. Previously physics research and attorney.
Sessions (2)
Additional Speaker Details
Using publicly available information we constructed an analysis to help you get a feel for this speaker before deciding to attend their session.
Thomas Wolf is the co-founder and Chief Science Officer of Hugging Face, and gives a candid, contrarian read on where AI is heading: he argues today's models risk becoming "yes-men on servers" rather than Einsteins absent a research breakthrough, and he is betting on open-source robotics (LeRobot) - cheap hardware and an "App Store for robots" - as the next inflection point for physical AI. Attend for an open-science perspective from a co-creator of Transformers and Datasets who is willing to challenge frontier-lab narratives.
Speaking style
Recent talks (1)
GitHub
@thomwolf- thomwolf/Magic-Sand★ 1,000Augmented-reality sandbox software that turns a sandbox into a real-time topography/terrain display.
- huggingface/transformersorg★ 162,000The model-definition framework for state-of-the-art ML models across text, vision, audio and multimodal - a library Wolf co-created and led.
- huggingface/datasetsorg★ 21,600The largest hub of ready-to-use datasets for AI models, with fast and efficient data tooling.
Recent writing (1)
- The Einstein AI model · blog · Mar 2025
Podcasts & interviews (2)
- Building the "App Store" for Robots: Hugging Face's Thomas Wolf on Physical AI · Training Data (Sequoia Capital) · Sep 2025
- Challenging the Average With Open-Source AI: Hugging Face's Thomas Wolf · Me, Myself, and AI (MIT Sloan Management Review) · Sep 2025