This morning at Scaleway’s ai-PULSE conferenceFrench billionaire and Iliad CEO Xavier Niel has given some additional details about his plans for a Paris-based AI research lab.
This new lab, called Kutai, will be a private non-profit company dealing with artificial general intelligence. He will collaborate with PhD students, postdocs and researchers on research papers and open source projects. When Iliad first revealed this research lab, the company said that Niel was committing 100 million euros to this project ($109 million at today’s exchange rate).
“Thanks to some amazing friends who are there today, we are now close to €300 million in funding for this initiative,” Niel told the conference. Among these “friends” is another French billionaire, Rodolphe Saadé, the CEO of French shipping and logistics giant CMA CGM, who is also paying €100 million. There are other smaller contributors, such as Eric Schmidt’s foundation and some anonymous donors.
This is just a starting point as Kyutai is open to more donations. “What’s interesting with so many journalists in the room is that the project will potentially be of interest to other investors,” Saadé said at a news conference after the announcement.
As Kyutai will be working on basic models, they will also need some computing power. The good news is that Scaleway, the cloud division of Iliad, recently acquired a thousand Nvidia H100 GPUs. These top GPUs are essential for inference and modeling and will be available at a cost to Kyutai.
Kyutai has already started hiring for its core scientific team. Six men took the stage this morning to talk about their past work and what they have in mind for the research lab — Patrick Perez, Edouard Grave, Hervé Jegou, Laurent Mazaré, Neil Zeghidour and Alexandre Defossez. They previously worked for Meta FAIR’s AI research team, Google’s DeepMind division, Inria, etc.
Patrick Perez, who previously worked for Valeo, will be the director of the research laboratory. Kyutai also assembled a team of scientific advisors who are well-known artificial intelligence researchers — Yejin Choi, Yann LeCun and Bernhard Schölkopf. They’ll just check everyone’s work once or twice a year and give feedback.
One of the reasons Kyutai believes she can convince some researchers to join her lab is that researchers will be able to publish research papers.
“Unfortunately, the big tech companies tolerate scientific publications less and less. Apart from being an ego boost for researchers, it helps advance research and contributes to the common good,” Niel said during the press conference.
Of course, this isn’t the first open AI research lab. OpenAI, as the name still suggests, started as a non-profit organization. But things changed dramatically after Sam Altman started working full-time at OpenAI in 2019. OpenAI moved to a more traditional corporate structure and raised funding from Microsoft.
Other companies are also working on basic open source models, such as Meta with its Llama model and Mistral AI. Kyutai’s models will also be open source, but the researchers describe their work as open science. They plan to release open source models, but also the training source code and data that explain how these models were released.
“In terms of timing, I don’t think our goal is necessarily to go as fast as Mistral, because we aspire to provide a scientific purpose, an understanding and a code base to explain the results,” Defosset said in the report. Press conference. But they expect to have something to share within a year.
Mazaré, another researcher from Kyutai’s team, still described the first open-source model of Mistral AI as a success because many community members have refined it and are exploring use cases based on the Mistral 7B model.
It will also be interesting to see if a research lab is more efficient at producing fundamental models compared to private companies, and how private companies will leverage Kyutai’s work for commercial applications.
Image Credits: Romain Dillet / TechCrunch
“I also firmly believe in open source and we must turn it into a French advantage,” French President Emmanuel Macron said in a pre-recorded video message to the conference.
France position: Setting use cases, not models
Macron also used this opportunity to define and defend France’s position on European AI law, saying use cases should be regulated, not model makers. France is pushing for the weakening of the AI law in tripartite dialogues (a tripartite dialogue is a negotiation between Europe’s three main bodies, the Parliament, the Commission and the Council).
“Regulation is not the enemy of innovation, quite the opposite. It is not a matter of defining good models, but we must ensure that the services available to our citizens are safe for them, for other economic actors and for our democracy,” Macron said.
“With work on European AI regulation currently in ‘trilogues’, regulation needs to be controlled rather than punitive, to sustain innovation and regulate use rather than the technology itself,” he added.
Niel basically sided with France’s position on this issue during the press conference. According to him, Europe is lagging behind in AI innovation and regulation will slow down European entrants and reduce their chances of catching up.
“At the moment we are more in the innovation part than the regulation part. Creating regulations means creating barriers for competitors,” Niel said.
Maybe if French AI companies become massively successful, things will change. “I’d love it if one day we could talk about French imperialism in AI,” Neil added later in the conversation.
Image Credits: Romain Dillet / TechCrunch
