Meta: More growth with a new AI architecture?

By Angus Luten
The Wall Street Journal
Translation: Laura Marcus

Facebook is moving its AI development from its research centers to its product groups to accelerate technology adoption and growth. Most companies, including Facebook, have traditionally focused their AI research in central hubs. Mita wants to go a different path.

Meta, the parent company of Facebook, announced last week that it would decentralize the development of modern artificial intelligence and machine learning tools.

Andrew Bosworth, chief technology officer of Meta, explained in an online post that the previous focus on a few independent R&D centers makes it difficult to integrate new AI capabilities across the company.

“With the new model, we will put these AI systems back into the hands of the Meta product groups,” Bosworth said. “We believe this will speed up the adoption of important new technologies across the organization. At the same time, we can continue to set new standards.”

After a weak first quarter and the slowest quarterly revenue growth since its IPO a decade ago, Meta is looking to get more of everything. Javier Olivan is set to become the new chief operating officer, succeeding Sheryl Sandberg. Known for his focus on growth, he intends to continue to focus on growth and efficiency.

Jerome Bisenti has led AI research at Meta AI for the past four years. He is expected to leave the company at the end of the month as part of the restructuring. Pesenti led important projects, such as the development of an artificial intelligence-powered supercomputer.

Like Meta, many companies with core AI research centers are finding it difficult to turn their newly developed technologies into commercial assets. Hubs were hardly integrated into the other departments and worked in a similar way to the AI ​​service being outsourced.

“It’s a clear indication that the company wants to focus on integrating AI into its products,” said Bill Group, director of the National Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign. The reorganization illustrates the company’s focus on finding commercial applications of AI. It’s not about researching new technologies in general.

According to John Carvell, a spokesperson for Meta, the main purpose of AI research work will not change under the new architecture. This also applies to the majority of managers. Joel Beniot and Antoine Bordis continue to lead Facebook AI Research (FAIR), the company’s core artificial intelligence research center.

However, the research center will operate in the future as part of Metas Reality Labs Research. This department oversees work on virtual reality, augmented reality, and other important parts of the Metaverse.

At its annual developer conference in October, CEO Mark Zuckerberg said that Facebook’s rebranding to Meta Platforms reflects the growth opportunities associated with the Metaverse. In the Metaverse, users interact through avatars in video game-like worlds, using digital headphones and other devices to work, shop and play. But according to most tech experts, the Metaverse won’t be around for a few years.

Reality Labs lost nearly $3 billion in the first quarter. This is due to the additional expense of developing headsets, software, and other virtual reality tools. This shows how committed the company is to this. However, this area was one of the few bright spots in Meta’s poor first-quarter results, as revenue grew from $534 million to $695 million. In contrast, profits fell from the previous year by 21 percent to 7.46 billion dollars.

According to Nick van der Meulen, a researcher at MIT’s Sloan Center for Information Systems Research, Meta’s decision to decentralize AI development shows “how important AI is to the future of the Metaverse.”

For a tech company like Meta, it no longer makes economic sense to make product development dependent on a single line of business in the future. “They have enough experience with AI to make it an integral part of all their products,” says van der Meulen.

When a product group like Reality Labs can develop their own technology, they don’t have to constantly compete for the resources of a central team or justify their research focus. Instead, they can develop their own solutions based on direct customer feedback.

“Think of it as research first, and then at some point you need good technical management and results,” said Eric Schmidt, former Google CEO and chair of the National Security Committee on Artificial Intelligence. “This is a good example of the spread of artificial intelligence.”