Meta CEO Mark Zuckerberg has announced that the company is intensifying its efforts in artificial intelligence (AI) and investing in massive data centers to outpace its competitors in developing human-level computer systems. In a Facebook post, Zuckerberg revealed that Meta is bringing together two of its research projects, FAIR and GenAI, to advance AI research and develop AI products for consumers. The long-term vision is to build general intelligence, open-source it responsibly, and make it widely accessible to benefit everyone.

While Zuckerberg did not explicitly define "general intelligence," experts in the field have used the term "artificial general intelligence" to describe AI systems that are as capable as humans. To support its future roadmap, Meta is constructing massive compute infrastructure, including stockpiling 350,000 H100 computer chips, which have become the go-to building block for AI data centers since their introduction by Nvidia in 2022. The Verge reported that Meta is one of Nvidia's largest customers for the H100 chips, tied with Microsoft.

Zuckerberg acknowledged the intense competition in hiring top AI scientists and emphasized the importance of investing in tools for these scientists to use. The competition for talent in the AI industry is unprecedented, with multiple companies vying for the same profiles, venture capitalists pouring money into various projects, and individuals starting their own initiatives externally.

Yann LeCun, Meta's chief AI scientist, expressed his belief that the internal changes at Meta would accelerate progress in AI. He highlighted FAIR as the best place in the world for advancing AI research and Meta as the best place to make it accessible to billions of people.

Zuckerberg and LeCun advocate for open-source AI technology, which would make most of the tech available for use by rival companies and individuals. This position has sparked heated debates within the tech industry, as it affects the business models of more secretive competitors and raises concerns about potential misuse if open-source tech becomes widely available. Meta and IBM recently launched the AI Alliance, promoting their open-source vision of AI, putting them at odds with Google and OpenAI.

Meta's commitment to AI is evident in its plans to build its own artificial general intelligence, surpassing human intelligence in almost all areas, and then opening it up to developers. Zuckerberg emphasized the need for advancements in AI across the board to create the best AI for chatbots, creators, and businesses. To handle the required processing power, Meta aims to have approximately 350,000 Nvidia AI chips by the end of the year.

The company also intends to merge its two major AI research groups, FAIR and GenAI, to accelerate its work. Zuckerberg believes that Meta's vision for AI and the metaverse, a virtual space, are interconnected. He predicts that by the end of the decade, people will frequently interact with AI through smart glasses like the Ray-Ban Meta glasses, which are powered by AI and enable hands-free communication, messaging, and video recording.

This announcement marks Meta's significant commitment to AI, amidst an ongoing AI arms race among Big Tech companies such as Microsoft, Google, and Amazon. These companies continue to introduce new AI tools and share their visions for the future. However, concerns have been raised about potential unintended harms associated with these revolutionary AI products from both established players and new entrants like OpenAI.

Earlier this year, Meta announced the creation of a new "top-level product group" to accelerate its work on AI tools. Since then, the company has released tools and information aimed at helping users understand how AI influences their experiences on Meta's apps. Analysts note that Meta's increased focus on AI aligns with the industry's broader embrace of AI, even as the initial excitement around the metaverse concept has waned. The shift to AI reflects Meta's efforts to revitalize its brand and regain lost momentum.