Kyutai has begun hiring for its core scientific team, with Patrick Perez set to be the director of the research lab. The lab will allow researchers to publish their work, a practice Niel claims is becoming less tolerated by big tech companies. The lab's models will be open-source, and they plan to release the training source code and data explaining how these models were created. The team expects to have something to share within a year. French President Emmanuel Macron has expressed support for the project, emphasizing the need to regulate AI use cases rather than model makers.
Key takeaways:
- French billionaire Xavier Niel has revealed more details about his plans for a Paris-based AI research lab called Kyutai, which will be a privately-funded nonprofit working on artificial general intelligence. The project has secured close to €300 million in funding, with Niel and Rodolphe Saadé, CEO of CMA CGM, each contributing €100 million.
- Kyutai will work with PhD students, postdocs, and researchers on research papers and open-source projects. The lab will also have access to a thousand Nvidia H100 GPUs acquired by Scaleway, the cloud division of Iliad, for model training and inference.
- The lab has already begun hiring for its core scientific team, which includes researchers from Meta’s AI research team FAIR, Google’s DeepMind division, and Inria. The team will be led by Patrick Perez, former Valeo employee, and will be advised by renowned AI researchers Yein Choi, Yann LeCun, and Bernhard Schölkopf.
- French President Emmanuel Macron has expressed support for the project, emphasizing the need for open-source AI to become a French asset. He also defended France's position on Europe's AI Act, arguing that use cases should be regulated, not model makers.