Meta Acquires Robotics Startup To Boost & Improve Its Humanoid AI Efforts

Meta has come out to say that it has paid an unknown amount to purchase the humanoid robotics business Assured Robot Intelligence (ARI) by acquiring it. This is a significant move into the field of humanoid robotics, with the San Diego-based company focusing on AI foundation models that enable robots to recognise, anticipate, and adjust to human behaviour in natural environments. 

A representative for Meta told members of the press through an email that they have acquired Assured Robot Intelligence, a company at the frontier and forefront of robotic intelligence designed to enable robots of varying kinds to understand, predict, and adapt to different kinds of human behaviours in complex and dynamic environments.

The team that worked at ARI, including its co-founders, which is a 20-person team, will join the Superintelligence Labs research branch of Meta’s AI business. Although the purchase price was not disclosed, the AI venture company AIX Ventures had contributed an undisclosed value in seed investment to ARI. 

The company had said it was developing foundation models to enable humanoid robots to carry out various physical tasks, including housework and chores of varying kinds.

The co-founder, Xiaolong Wang, had won a number of prominent and distinguished honors while working as a researcher at Nvidia and an associate professor at UC San Diego. Lerrel Pinto, who is also a former NYU professor who co-founded the kid-sized humanoid business Fauna Robotics before Amazon acquired it last month, has also received a number of notable honors.

Meta’s humanoid aspirations will benefit from ARI. “This team, under the direction of Lerrel Pinto and Xiaolong Wang, will contribute a wealth of experience in designing our models and cutting-edge capabilities for robot control and self-learning to whole-body humanoid control.”

For many years, Meta researchers have been developing in a way by working on humanoid robot technology, which has been the main focus for years of investigations. A little while ago, like a year, a piece of notes in letter form was leaked and revealed Meta’s plans to develop a consumer-focused robot that features AI models and hardware.

Many AI experts now think that training AI models in the real world, where robots learn through direct interaction and communication rather than data alone, will be a necessary thing to reach artificial general intelligence (AGI), the theoretical point at a part where AI reaches or surpasses human-level intelligence across all domains, even if Meta never releases a consumer humanoid product. 

The ARI and Fauna deals are part of a larger and bigger industry-wide sprint where projections range greatly, from Goldman Sachs’ $38 billion by 2035 estimate to Morgan Stanley’s $5 trillion by 2050 estimate. The difference reflects both the enormous potential and the uncertainty surrounding tech, which is still finding its stability.

The strategic aim, by acquiring a robotics startup, is for Meta to become the Android-like foundational platform for humanoid robotics, developing an intelligent software and sensor ‘brain’ that enables physical world training, a move that puts it in direct competition against Tesla, Amazon, and Google in the race to power home and industrial humanoid assistants.