Advanced Machine Intelligence (AMI Labs) has secured $1.03 billion, equivalent to €890 million, in a seed funding round to develop artificial intelligence systems focused on comprehensive world understanding. This could be the largest ever for a European company. Founded by Yann LeCun, this newly formed company aims to create AI that can reason, plan, and operate safely through persistent memory. “Advanced Machine Intelligence (AMI) is building a new kind of AI system that understands the world, has persistent memory, can reason and plan, and is controllable and secure,” the company said in a recent post that garnered a lot of attention. This significant investment signals a major push to develop AI beyond current capabilities, prioritizing not only performance but also interpretability and controllability.
AMI Labs secures $1.03 billion in seed funding round
AMI Labs founder Yann LeCun publicly announced the funding and shared a photo of the Veil Nebula taken from his backyard as a symbolic backdrop for the release. AMI Labs intends to prioritize control and safety in its AI designs to address critical concerns in the field. This focus on “world models” distinguishes AMI Labs from current AI trends that are primarily centered around large-scale language models. Recruitment is already underway as AMI Labs prepares to expand its business and turn its vision into tangible products. LeCun said the company is actively seeking qualified candidates. This investment demonstrates the confidence of global investors in AMI Labs’ approach to artificial general intelligence, a field that aims for human-level cognitive capabilities in machines. LeCun added in the announcement, “We’re hiring!”
AI system that emphasizes world understanding and safety control
While current artificial intelligence systems primarily excel at narrow tasks and often require large datasets even for limited functionality, new approaches prioritize building AI that inherently understands the physical world. AMI Labs aims to build AI systems that can reason, plan, and maintain persistent memory beyond the limitations of current reactive models. This emphasis on “world models” represents a departure from common AI architectures, which often lack a consistent understanding of cause and effect or the ability to predict outcomes beyond training data. This significant investment will enable AMI Labs to aggressively advance research into these more robust and reliable AI designs and address growing concerns about the safety and predictability of increasingly complex algorithms. The size of this funding round reflects widespread investor confidence in the potential of this approach to produce artificial intelligence that is not only powerful but also demonstrably safe and controllable.
Advanced Machine Intelligence (AMI) is building a new kind of AI system that understands the world, has persistent memory, can reason and plan, and is controllable and secure.
AMI Lab
