Yann LeCun’s AMI Labs raises $1.03 billion to build global models

AMI Labs, the new venture co-founded by Turing Award winner Yann LeCun after leaving Meta, has raised $1.03 billion at a pre-valuation of $3.5 billion. AMI works on universal models, or artificial intelligence that learns from reality, not just from language.

This class has fewer players than generative AI, but perhaps not for long. “My prediction is that ‘universal models’ will be the next buzzword,” Alexandre Lebrun, CEO of AMI Labs, told TechCrunch. “Within six months, every company will be calling itself a global fundraising model.”

LeBron said this with a smile because he believes AMI Labs is completely different: its goal is to understand the real world. This could have applications in healthcare, where AMI Labs’ first partner will be Nabla, the digital health startup he now heads.

As CEO of Nabla, LeBrun came to the same conclusion as LeCun about the limitations of large language models (LLMs) where hallucinations can have life-threatening repercussions. But he also knows that it will take some time for the startup to offer a viable alternative based on JEPA, the predictive architecture for joint embedding that LeCun proposed in 2022.

“AMI Labs is a very ambitious project, because it starts with basic research. It’s not a typical applied AI startup that can launch a product in three months, generate revenue in six months and make $10 million in a year.” [annual recurring revenue] “In 12 months,” Lebrun said. In contrast, it could take years for global models to move from theory to commercial applications.

Despite this time horizon, developing world model companies have attracted big checks. SpAItial raised a $13 million seed round, an unusually large amount for a European startup; While Fei-Fei Li’s World Labs took in a whopping $1 billion last month alone. Now, AMI Labs is joining the club with more funding than initially reported.

The French AI lab was reportedly seeking just €500 million last December, but ended up raising around €890 million, likely thanks to its team. In addition to LeCun’s involvement as Chairman of the Board and LeBrun’s track record as an entrepreneur, it also boasts Meta’s Europe Vice President Laurent Solly as Chief Operating Officer, leading researchers Saining Xie as Head of Science, Pascal Fung as Head of Research and Innovation, and Michael Rabat as Vice President of Global Models.

According to LeBrun, the high interest in the startup gave the startup an opportunity to choose investors, both in terms of alignment of expectations and background. The round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, with participation from several other funds and industry-related backers, as well as individuals including Tim and Rosemary Berners-Lee, Jim Breyer, Mark Cuban, Mark Leslie, Xavier Neal, and Eric Schmidt.

Value-added aside, this funding will give AMI Labs some useful runway to fund its two major cost centers: compute and talent. Lebrun said he will prioritize quality over quantity to build the AMI Labs team in four main locations: Paris, where the headquarters are located; New York, where LeCun teaches at New York University; In Montreal, where Rabat is located; And in Singapore, to recruit AI talent and get closer to future clients in Asia.

Although AMI Labs has no plans to generate revenue at the moment, it still plans to engage with potential customers early on. “We’re developing models of the world that seek to understand the world, and you can’t do that confined in a lab,” Lebrun said. “At some point, we need to put the model in a real world setting with real data and real evaluations.”

When the time comes, AMI Labs will turn to partners to explore deployments — and Nabla is the first partner revealed that expects access to these early models, but certainly not the last. “This may explain the presence and strong interest of some industrial players and potential partners in the investment round,” Lebrun said.

In addition to major investors and angels, AMI Labs is backed by NVIDIA, Samsung, Sea, Temasek and Toyota Ventures, as well as French gamers’ 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.

These investments may take some time to translate into commercial applications. But true to LeCun’s beliefs, AMI Labs will publish research as it goes.

“We’ll also be making a lot of the code open source,” said Lebrun, who also worked at Meta’s AI research lab, FAIR. Although open research is becoming “increasingly rare,” the founders of this startup still believe in it. “We believe things move faster when they are open, and it is in our interest to build a community and research ecosystem around us.”