The global race to shape the future of artificial intelligence has taken a fresh turn as renowned computer scientist Yann LeCun unveiled the first major milestone of his new venture. The former chief AI scientist at Meta Platforms has secured about $1.03 billion in funding for his startup, Advanced Machine Intelligence, widely known as AMI, positioning the firm as a serious contender in the rapidly expanding AI industry.
The funding round values the company at roughly $3.5 billion before the investment, an impressive figure for a young firm still developing its core technology. The financing was co-led by prominent venture firms including Cathay Innovation, Greycroft, Hiro Capital and HV Capital, along with Bezos Expeditions, the investment arm linked to Amazon founder Jeff Bezos.
For LeCun, who spent more than a decade shaping Meta’s artificial intelligence strategy, the new company represents more than just another startup. It reflects his long-held belief that the current direction of AI development is fundamentally limited. While most companies are racing to build ever larger language models, AMI is trying to build a different kind of intelligent system altogether.
The company’s approach centres on what researchers call “world models”, systems designed to understand and reason about the physical world rather than simply predict the next word in a sentence. If successful, the technology could reshape the next generation of intelligent machines, from industrial robots to everyday consumer devices.
Table of Contents
A Break from the Dominant AI Playbook
In recent years, the artificial intelligence boom has been largely defined by large language models, the technology behind popular chatbots and generative AI tools. These systems are trained on enormous amounts of text data and learn to predict the next word or token in a sequence.
While this approach has produced impressive results, LeCun has repeatedly argued that it is not enough to build truly intelligent systems.
According to him, models that rely mainly on predicting words or pixels lack a deeper understanding of how the real world works. They may appear intelligent when answering questions or generating text, but they struggle with reasoning, planning and interacting with complex environments.
AMI’s mission is therefore to move beyond language-based AI by developing systems that can learn how the physical world behaves. These models would absorb information from videos, spatial data and real world interactions rather than relying primarily on text.
The goal is to create machines capable of planning actions, predicting outcomes and developing a form of practical understanding of their surroundings.
Experts say this shift could be crucial for applications that require more than conversation skills. For instance, a robot working in a factory or assisting people at home needs to understand objects, movement and cause and effect. Language prediction alone cannot deliver that level of awareness.
LeCun believes that combining reasoning, memory and perception will ultimately produce more capable and reliable AI systems than the current generation of models.
Global Investors Bet on a Different AI Future
The scale of funding behind AMI reflects growing investor interest in alternative approaches to artificial intelligence.
Over the past two years, billions of dollars have flowed into AI startups worldwide as companies and governments compete for technological leadership. Industry analysts estimate that global investment in artificial intelligence reached more than $200 billion in 2025 alone.
Against this backdrop, AMI’s $1 billion plus funding round stands out as one of the largest early-stage investments in AI research.
The backing of international investors from Europe, the United States and Asia signals confidence that LeCun’s ideas could produce a breakthrough. Some investors see the startup as an opportunity to challenge the dominance of the current AI ecosystem led by American tech giants and Chinese research groups.
The company itself has an international footprint. Its headquarters are in Paris, but it also operates offices in cities including Montreal, Singapore and New York, drawing talent from across the global AI research community.
AMI’s leadership team includes several former researchers from Meta and other major technology companies. The startup is led by Alexandre LeBrun as chief executive, while LeCun serves as executive chair and scientific guide for the project.
This mix of academic research and industry experience reflects the company’s ambition to develop both fundamental AI breakthroughs and practical products.
Targeting Industry Before Consumers
Although many people associate artificial intelligence with chatbots and digital assistants, AMI’s early focus is likely to be on industrial applications.
According to LeCun, the company intends to work with organisations operating complex systems such as manufacturers, automotive firms, aerospace companies, biomedical researchers and pharmaceutical groups.
These industries often require machines that can reason about physical processes, predict outcomes and make decisions in real environments. For example, AI systems might help control advanced manufacturing equipment, manage autonomous vehicles or assist scientists in analysing complex experiments.
Such applications demand a deeper understanding of the world than today’s language-based models can provide.
Over time, however, the technology could reach consumers as well. One possible future application is domestic robotics. For a robot to function effectively in a home, it would need a degree of common sense about objects, movement and everyday situations.
That kind of capability is precisely what AMI’s “world model” research aims to achieve.
Another potential avenue involves wearable devices. LeCun has suggested that the technology could eventually be integrated into smart glasses developed by Meta, enabling them to interpret and respond to real world environments more intelligently.
While such applications remain speculative, they illustrate how the technology might evolve beyond industrial use.
The Larger Battle Over AI’s Future
AMI’s launch comes at a moment when the global artificial intelligence industry is undergoing rapid transformation.
Major technology companies are pouring resources into building ever larger models, often requiring enormous computing power and data resources. This strategy has produced impressive breakthroughs, but it has also raised concerns about cost, sustainability and the concentration of AI power among a handful of companies.
LeCun has long been one of the most prominent voices questioning whether scaling language models alone will deliver human-level intelligence.
His departure from Meta at the end of 2025 marked the end of an era. During his tenure, he founded the company’s influential AI research division known as Facebook AI Research, later called FAIR, which produced many important advances in machine learning.
Now he is attempting to chart a different path for the field.
Some researchers see AMI as part of a broader shift in AI thinking. Instead of focusing purely on language generation, scientists are increasingly exploring systems that combine perception, reasoning and interaction with the physical world.
These ideas draw inspiration from how humans learn. People do not understand the world only through words. They learn through observation, experience and interaction with their environment.
By building AI systems that learn in a similar way, researchers hope to overcome many of the limitations that plague today’s models, including hallucinations, lack of common sense and difficulty handling complex real-world tasks.
Still, the path forward is uncertain.
Building world models that truly understand reality is an extremely challenging scientific problem. Even with substantial funding, it could take years before the approach produces tangible results.
Yet for LeCun and his supporters, the potential reward is worth the effort.
If AMI succeeds, it could help usher in a new generation of artificial intelligence systems that do more than talk. They could understand, reason and act in the world around them.
In the increasingly crowded AI race, that vision offers a compelling alternative to the dominant strategies shaping the industry today.
Join Our Social Media Channels:
WhatsApp: NaijaEyes
Facebook: NaijaEyes
Twitter: NaijaEyes
Instagram: NaijaEyes
TikTok: NaijaEyes
