LeCun's AMI Labs: $1.03 Billion for Global AI

⚡

Key Takeaways

1AMI Labs, co-founded by Yann LeCun, has raised $1.03 billion to develop global AI models.

2Alexandre LeBrun, CEO of AMI Labs, predicts that 'global models' will become a key term in the AI industry.

3The funding will allow AMI Labs to focus on fundamental research and recruit talent in four strategic cities.

💡Why it matters — This massive funding positions AMI Labs as a major player in AI innovation, with potential implications for various sectors, including healthcare.

AMI Labs: An Impressive Fundraising for a New Era of AI

AMI Labs, the new company co-founded by Yann LeCun after his departure from Meta, recently announced that it has raised an impressive $1.03 billion. This fundraising comes with a pre-money valuation of $3.5 billion. The main goal of AMI Labs is to develop global models, a form of artificial intelligence that strives to understand reality beyond mere language processing.

Global Models: A Growing Concept

Although the field of global models is currently less crowded than that of generative AI, this could change rapidly. Alexandre LeBrun, CEO of AMI Labs, told TechCrunch that he expects the term "global models" to quickly become a buzzword in the industry. According to him, in the next six months, many companies may claim this concept to attract investments.

A Unique Approach to Understanding the Real World

LeBrun emphasizes that AMI Labs' uniqueness lies in its ambition to understand the real world. This approach could have significant applications in sectors such as healthcare. Nabla, a digital health startup of which LeBrun is president, will be AMI Labs' first partner in this area.

The Limitations of Large Language Models

As CEO of Nabla, LeBrun has shared LeCun's concerns regarding the limitations of large language models (LLMs), particularly the risks of hallucinations that could have serious consequences. However, he acknowledges that developing a viable alternative based on JEPA, the architecture proposed by LeCun in 2022, will take time.

An Ambitious Project Focused on Fundamental Research

LeBrun describes AMI Labs as an ambitious project that starts with fundamental research. Unlike applied AI startups that can quickly launch products and generate revenue, AMI Labs may require several years to turn its theories into commercial applications. He notes that typical applied AI startups can reach $10 million in annual recurring revenue within 12 months, a goal that is far from immediate for AMI Labs.

Significant Investments in Global Models

Despite the time required to develop these models, investments in this area are growing rapidly. For example, SpAItial raised $13 million in an initial funding round, while World Labs by Fei-Fei Li recently secured $1 billion. AMI Labs now joins this group with funding exceeding initial expectations.

A High-Caliber Team for an Ambitious Project

Initially, AMI Labs aimed to raise €500 million last December but ultimately reached around €890 million. This success is partly due to the prestigious team at the company, which includes figures like Laurent Solly, VP of Meta for Europe and COO of AMI Labs, as well as renowned researchers such as Saining Xie, Chief Scientist, Pascale Fung, Head of Research and Innovation, and Michael Rabbat, VP of Global Models.

A Strategic Choice of Investors

LeBrun indicated that the high interest from investors allowed AMI Labs to choose its financial partners based on alignment of expectations and backgrounds. The funding round was co-led by major players such as Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions. Notable individual investors include Tim and Rosemary Berners-Lee, Jim Breyer, Mark Cuban, Mark Leslie, Xavier Niel, and Eric Schmidt.

A Recruitment and International Development Strategy

With this funding, AMI Labs plans to focus on two main cost centers: computing and talent recruitment. LeBrun specified that the company will prioritize quality over quantity by recruiting in four key cities: Paris, New York, Montreal, and Singapore.

Early Interaction with Potential Clients

Although AMI Labs does not yet have plans to generate revenue, the company intends to interact with potential clients from the outset. LeBrun emphasized the importance of testing models in real-world situations to refine their understanding of the world.

Strategic Partnerships for the Future

When the models are ready, AMI Labs will collaborate with partners to explore deployments. Nabla is already announced as the first partner, but other collaborations are planned. This strategy explains the interest from many industry players during the funding round.

Diverse Support for AMI Labs

In addition to the main investors, AMI Labs benefits from the support of major companies such as NVIDIA, Samsung, Sea, Temasek, and Toyota Ventures. French players like l'Association Familiale Mulliez, Groupe Industriel Marcel Dassault, and Publicis Groupe are also involved, further strengthening AMI Labs' position on the international stage.