AMI Labs secures major funding round

AI research startup AMI Labs, co-founded by Turing Award winner Yann LeCun, has raised $1.03 billion in funding at a $3.5 billion pre-money valuation. The company focuses on developing “world models,” a type of artificial intelligence designed to learn from real-world environments rather than only from text data.

The funding round places AMI Labs among a small but growing group of companies working on this emerging AI approach.

A focus on AI systems that understand the real world

AMI Labs aims to build AI systems that learn from observation and interaction with the physical world. This approach differs from traditional large language models (LLMs), which mainly rely on text-based training.

“My prediction is that ‘world models’ will be the next buzzword,” AMI Labs CEO Alexandre LeBrun told TechCrunch. “In six months, every company will call itself a world model to raise funding.”

LeBrun believes AMI Labs’ research approach is distinct. The company plans to develop AI systems that can better interpret real-world situations. One potential use case is healthcare.

AMI Labs’ first disclosed partner will be Nabla, a digital health startup where LeBrun currently serves as chairman.

Limitations of large language models

LeBrun said his experience leading Nabla highlighted the limitations of current AI systems. In healthcare, inaccurate outputs from language models could create serious risks.

Both LeBrun and LeCun have discussed these challenges publicly. Their work now focuses on alternative architectures such as JEPA (Joint Embedding Predictive Architecture), which LeCun proposed in 2022.

However, LeBrun noted that building a viable alternative will take time.

“AMI Labs is a very ambitious project, because it starts with fundamental research. It’s not your typical applied AI startup that can release a product in three months, have revenue in six months, and make $10 million in [annual recurring revenue] in 12 months,” LeBrun said.

Because the work begins with basic research, commercial applications may take several years to emerge.

Growing interest in world model startups

Despite the longer timeline, investors have shown increasing interest in companies developing world models.

For example, SpAItial recently raised a $13 million seed round, which is relatively large for an early-stage European AI startup. In addition, Fei-Fei Li’s World Labs secured $1 billion in funding last month.

AMI Labs now joins this group with one of the largest funding rounds in the sector.

Experienced leadership team

The Paris-based startup initially aimed to raise €500 million late last year. However, strong investor demand helped the company secure about €890 million.

AMI Labs’ leadership team includes several well-known figures in AI research and technology.

Along with Yann LeCun serving as chairman and Alexandre LeBrun as CEO, the company also includes:

Laurent Solly, former Meta VP for Europe, as COO
Saining Xie, chief science officer
Pascale Fung, chief research and innovation officer
Michael Rabbat, VP of world models

LeBrun said strong investor interest allowed the company to choose partners who shared its long-term research goals.

Investors backing the funding round

The funding round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions.

Several other investors also participated. These include industry-linked companies and well-known individuals such as Tim and Rosemary Berners-Lee, Jim Breyer, Mark Cuban, Mark Leslie, Xavier Niel, and Eric Schmidt.

Corporate investors include Nvidia, Samsung, Sea, Temasek, and Toyota Ventures. French organizations such as Association Familiale Mulliez, Groupe Industriel Marcel Dassault, and Publicis Groupe also joined the round.

Investment to support research infrastructure

The new funding will mainly support two key areas: computing infrastructure and research talent.

AMI Labs plans to build teams across four locations:

Paris, the company’s headquarters
New York, where LeCun teaches at NYU
Montreal, where Rabbat is based
Singapore, which offers access to AI talent and Asian markets

LeBrun said the company will prioritize hiring experienced researchers rather than expanding rapidly.

Early collaboration with potential customers

Although AMI Labs does not plan to generate revenue immediately, the company intends to work with partners early in the research process.

“We are developing world models that seek to understand the world, and you can’t do that locked up in a lab. At some point, we need to put the model in a real-world situation with real data and real evaluations,” LeBrun said.

Nabla will be the first partner to test early versions of these models. However, LeBrun expects more partnerships in the future.

“This may explain the presence and strong interest of certain industrial players and potential partners in the investment round,” he said.