Former Meta AI Chief Secures $1 Billion to Build Systems That Understand the Physical World

Former Meta AI Chief Secures $1 Billion to Build Systems That Understand the Physical World

2026-03-10 digital

Paris, Tuesday 10 March 2026
Valued at $3.5 billion, Yann LeCun’s new startup AMI has raised $1 billion to develop artificial intelligence that reasons and understands the physical world, moving beyond traditional language models.

A Paradigm Shift Beyond Language Models

On 9 March 2026, Advanced Machine Intelligence (AMI), a Paris-based startup, announced a massive capital injection of $1.03 billion, securing a pre-investment valuation of $3.5 billion [1][2]. The venture is spearheaded by Yann LeCun, who departed his role as Meta’s chief artificial intelligence scientist in November 2025 [1][2]. LeCun, a 2018 Turing Award laureate, has been a vocal critic of the current industry fixation on large language models (LLMs) [1]. He has publicly dismissed the notion that simply scaling up LLM capabilities will lead to human-level intelligence, labelling the idea “complete nonsense” [1]. Instead, AMI is built on the premise that true artificial intelligence must master the physical world rather than merely predicting the next word or pixel [2].

Digitalising Legacy Industries

The immediate commercialisation strategy for AMI focuses heavily on the digitalisation of legacy industries, targeting organisations that manage highly complex physical systems [2]. By integrating advanced artificial intelligence into sectors such as automotive manufacturing, aerospace, biomedicine, and pharmaceuticals, AMI aims to become the foundational provider of intelligent systems across diverse applications [2]. The startup plans to release its initial models swiftly, with early strategic partnerships already slated to include industry giants like Toyota and Samsung [1]. This approach highlights a broader trend in the digital economy, where scalable software solutions are increasingly required to navigate and optimise real-world physical operations, moving beyond traditional Software-as-a-Service (SaaS) boundaries into deep tech integration [GPT].

Europe’s Resurgent Tech Ecosystem

AMI’s historic funding round is emblematic of a wider resurgence within the European technology sector. Following several years of sluggish venture capital activity, the European startup ecosystem experienced significant momentum in 2025, with 27 companies crossing the $1 billion valuation threshold to achieve ‘unicorn’ status [3]. The United Kingdom led this resurgence by producing nine new unicorns, representing 33.333 per cent of the total cohort, followed by Germany with five, and Sweden with four [3]. France, where AMI is headquartered, contributed two new billion-dollar startups to this group, alongside nations such as Finland, the Netherlands, and Ireland [3].

The Evolving Competitive Landscape

The emergence of AMI also highlights the intensifying competition among global technology behemoths to define the next era of artificial intelligence. While LeCun pioneers alternative approaches in Paris, his former employer continues to double down on traditional models. In June 2025, Meta restructured its artificial intelligence operations, consolidating them under a new division named Meta Superintelligence Labs [2]. This unit is now led by Alexandr Wang, the former CEO of Scale AI, signalling Meta’s continued aggressive investment in large language models [2].

Sources & Ecosystem Partners

  1. www.wired.com
  2. nl.marketscreener.com
  3. www.linkedin.com

Artificial intelligence Venture funding