Meta spared no expense in building a new AI team last year, shelling out multimillion-dollar pay packets to poach the researchers at the heart of its superintelligence strategy. But those who joined weren’t driven by money alone, according to Alexandr Wang, Meta’s AI chief who joined from his startup, Scale AI.
“It is a misconception to think that researchers are only motivated by money,” Wang said in the newspaper article. The Core Memory Podcast published yesterday (May 13). “For many of them, in fact, the financial prospects of staying wherever they are are looking very good.”
Wang himself was one of Meta’s most expensive works. The 29-year-old joined the team at startup Meta Superintelligence Labs (MSL) last June, after Meta acquired a 49 percent stake in Scale AI—the data-labeling startup he previously led—in a deal valued at $14.3 billion. Before that, Wang had been in talks with Meta CEO Mark Zuckerberg, who was increasingly “AGI-pilled,” dissatisfied with the progress of Meta’s Llama 4 model and eager to reset the company’s AI trajectory, Wang said.
Wang now oversees the MSL, which is divided into four teams. The largest, still unnamed, focuses on advanced AI research and includes talent from Google DeepMind, Anthropic, and OpenAI. Meta reportedly offered some OpenAI researchers signing bonuses of up to $100 million, OpenAI CEO Sam Altman said last year.
Wang pushed back on the idea that Meta just scrapped its way into the AI race, calling it a “narrative violation, or maybe the difference between the outside view and what it’s like day-to-day on the inside.” The idea, he said, stems in part from how fast the Meta was. “When I came in, I knew that if we wanted to build good models, we needed to have a team yesterday, so we had to just go and brush it up and do it as quickly as possible.”
Instead, Wang argued, it is MSL’s culture that attracts talent. He pointed out the high computer per researcher, organized groups, and the willingness to support ambitious research bets as the key pulls.
Still, Meta’s hiring has rubbed industry figures the wrong way. Altman, who was a friend of Wang’s and even lived with him during the COVID-19 crisis, has he reportedly described Meta’s behavior as “somewhat offensive.” Ashlee Vance, manager of The Core Memory Podcasttold Wang that the OpenAI boss had “no flattering things” about him before the episode.
Controversies like this are nothing new in Silicon Valley. Altman also went aims at Anthropic CEO Dario Amodeiand yet within the California case from a lawsuit brought by Elon Musk, one of the founders of OpenAI.
“I think some of this is unfortunate,” Wang said of the industry’s competitors. “My real hope is that all this hatred will subside over time and people will come together and see that we are building this very important technology.”
Criticism from Yann LeCun
Wang also faced criticism from Yann LeCun, a prominent researcher who left Meta last year after more than a decade shaping its AI strategy. In a January interview with the Financial Times, LeCun described Wang as “young” and “naive.”
“Yann is a remarkable, outspoken person, and I think everyone always knows what Yann is thinking,” Wang said. The two seem to have gotten back together: they met in India a few weeks after the interview, where LeCun congratulated Wang on the release of MSL’s latest Muse Spark model.
Wang also defended his background, noting his early experience as a software engineer at Addepar and Quora before founding Scale AI at age 19. As for the criticism that comes with his age, he remains undaunted. “People have been saying this all my time in Silicon Valley.”
More broadly, Wang said, public scrutiny comes with a place for high-level AI leaders. “It may be distracting, but I choose to include it in the work we do and put it out there.”
