Yann LeCun’s AMI Labs has raised $1.03 billion to develop global prototypes

AMI Labsand innovation founded by Turing Award winner Yann LeCun after leaving Meta, he raised $1.03 billion at a valuation of $3.5 billion. AMI is working on global models, or AI that learns from reality, not just from language.

This team has fewer players than generative AI, but probably not by much. “My guess is that ‘global brands’ will be the next buzzword,” AMI Labs CEO Alexandre LeBrun told TechCrunch. “In six months, every company can call itself the world’s leading fundraiser.”

LeBrun said this with a smile because he thinks AMI Labs is very different: its goal is to understand the real world. This may include medical applications, where AMI Labs’ first partner will be Nablaa digital health startup of which he is now the chairman.

As the CEO of Nabla, LeBrun came to the same conclusion as LeCun on the limits of large languages ​​(LLMs) where imagination can have dangerous consequences. But they also know it will take time for startups to provide an alternative that can be adopted JEPAJoint Embedding Predictive Architecture presented by LeCun in 2022.

“AMI Labs is a very ambitious project, because it starts with basic research. You don’t start your AI that can produce products in three months, have money in six months and make $10 million in (recurring annual income) in 12 months,” said LeBrun. In contrast, it can take years for international models to move from theory to commercialization.

Even so, the international fashion industry has attracted big checks. SpaItial was raised a total of $13 million – surprisingly large for a European start; while Fei-Fei Li of World Labs he raised $1 billion just last month. Now, AMI Labs has joined the bandwagon with more money than originally reported.

A French AI lab was reported looking for just €500 million Last December, but he earned €890 million, probably thanks to his team. In addition to LeCun’s involvement as chairman and Mr. LeBrun history as a businessIt also boasts Meta’s European Vice President Laurent Solly as COO, and top researchers Saining Xie as scientific director, Pascale Fung as chief research & innovation officer and Michael Rabbat as VP of national models.

According to LeBrun, the high interest rate gave the startup the opportunity to choose investors, based on its expectations and background. The round was led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital and Bezos Expeditions, with participation from several other funds and industry partners, as well as people including Tim and Rosemary Berners-Lee, Jim Breyer, Mark Cuban, Mark Leslie, Xavier Niel and Eric Schmidt.

Apart from the cost, this investment will give AMI Labs a better way to accelerate the banking of its two main areas: compute and talent. LeBrun said he will prioritize quality over quantity to build the AMI Labs team in four key locations: Paris, where it is headquartered; New York, where LeCun teaches at NYU; in Montreal, where his Rabbat is; and Singapore, both to find AI talent and to be closer to future customers in Asia.

Although AMI Labs does not have any investment plans at this time, it still plans to engage with prospective customers in the early stages. “We are making models of the world that need to understand the world, and you can’t do it closed in the lab. At some point, we need to put the model in real life with real data and real analysis,” said LeBrun.

When the time comes, AMI Labs will turn to its partners to search for shipments – and Nabla is the first partner to be revealed hoping to find these early versions, but not the final ones. “This may explain the presence and strong interest of some industry players and potential financial partners,” LeBrun said.

In addition to investment managers and angels, AMI Labs is supported by NVIDIA, Samsung, Sea, Temasek and Toyota Ventures, as well as French players Association Familiale Mulliez, Groupe Industriel Marcel Dassault and Publicis Groupe. Aglaé Lab, Alpha Intelligence Capital, Artémis, Bpifrance Digital Venture, New Legacy Ventures, SBVA and ZEBOX Ventures also participated.

This investment may take time to convert into commercial activity. But staying true to LeCun’s beliefs, AMI Labs will publish papers as they go.

“We’re also going to make a lot of open source code,” said LeBrun, who also worked at Meta’s AI research lab, FAIR. Although open research is “very rare,” the founders still believe in it. “We think things move faster when they’re open, and it’s in our best interest to create a community and research ecosystem around us.”