Meta’s superintelligence dream team will be management challenge of the century

By Jo Constantz, Bloomberg

Meta Platforms Inc. is spending a fortune to assemble the brightest minds in artificial intelligence. Chief Executive Officer Mark Zuckerberg may want to note: Research suggests that packing a team with too much genius can backfire.

So far, more than a dozen engineers from OpenAI have defected to Meta, joined by notable experts from Anthropic and Google’s DeepMind. Zuckerberg’s wager is that by concentrating top talent and giving them unlimited resources, he can gain ground on rivals and fast-track the development of AI systems so advanced that they will approach “artificial general intelligence,” the hypothetical point at which the model surpasses human-level capacities.

As it turns out, loading up a team with multiple superstars is not always a winning strategy, as any frustrated sports fan could tell you. Without expert management, too much talent in one group can lead to diminishing returns — or even outright failure — if egos clash and the chemistry is poor enough.

“There’s this belief on Wall Street and in Silicon Valley that you just bring the most capable people, put them together and magic just happens,” said Boris Groysberg, a professor at Harvard Business School who has studied team dynamics for over two decades. “No, magic doesn’t happen. What you have in many cases is a lot of jealousy, backstabbing, sabotage.”

At Meta, keeping the superstars out of these traps is a task that falls to Alexandr Wang, the 28-year-old former CEO of Scale AI, and former GitHub Inc. CEO Nat Friedman, 48, the newly appointed leaders of Meta’s 50-person superintelligence unit.

Decades of academic studies indicate the challenges they’ll face.

In the 1970s, management scholar Meredith Belbin observed that teams composed entirely of high-IQ individuals were prone to prolonged arguments, showed little cohesion and struggled to reach decisions, with members more interested in debating than collaborating.

In 2011, Groysberg and others published a study concluding that on Wall Street, after a certain point, adding more “all-star” analysts to research teams at elite firms actually hurt performance. There appeared to be a tipping point, usually reached when all-stars with overlapping areas of expertise accounted for roughly half the research team, where the analysts’ egos took over and they started gatekeeping information instead of cooperating.

Other research, meanwhile, shows that a team’s performance depends a lot on how well its members communicate and collaborate. Teams that allow more turn-taking in discussions tend to have higher “collective intelligence,” regardless of the raw brainpower in the room.

Supergroup Management Tips

A couple of management principles become especially important when overseeing powerhouse teams. One is that each person’s lane must be clearly defined.

“If everyone has clear swimming lanes, they’re not going to see each other as a threat,” said Lindred Greer, a professor at the University of Michigan’s Ross School of Business. If there are people with similar backgrounds and talents, Greer said that kind of duplication is ok – as long as they’re kept separate.

Another trick is to openly determine at the outset who will hold decision-making rights  on key issues, otherwise competition for authority may destroy the group.

“Sometimes hierarchy has a bad name — you’ve got a leader telling everybody what to do,” said Anita Williams Woolley, a professor at Carnegie Mellon University’s Tepper School of Business. “But actually, there are certain ways in which having a hierarchy helps groups coordinate.” The hierarchy can change based on the problem at hand, Woolley said, but clarity is crucial in an environment where everyone wants to be on top.

And then there’s building group chemistry, which means developing trust between team members, communicating openly and establishing a sense of shared purpose. While there’s a lot of science to these aspects of making great teams — from Richard Hackman’s 2002 book Leading Teams to Google’s Project Aristotle — many leaders aren’t willing to put in the time it takes, Groysberg said.

“We just don’t have a lot of executives and CEOs who have the rigor and patience to implement it,” he said. “I always say, if you need a high performing team on Friday, Thursday is not the day to start building it.”

Though the warning applies universally, it’s especially relevant at Meta, which is assembling its superintelligence team in a hurry to catch up with Google and OpenAI.

And then there’s the money, which almost never fails to complicate matters.

With all of Zuckerberg’s widely publicized poaching efforts, which have included the dangling of pay packages topping $200 million, compensation details for many of the new recruits have become public knowledge. That could strain team-building efforts and influence the team’s dynamics.

“For many groups in that sort of setting, what people are paid is almost like wearing your rank on your sleeve in the military – you walk in and you have two stars and they have three stars,” Woolley said. “It absolutely establishes what the hierarchy is. It’ll be important for the leaders to be absolutely clear if that’s true here as well.”

Michael Dell, CEO of Dell Technologies Inc., said in a recent interview that the extravagant compensation packages for new AI hires could rankle veteran Meta employees. “People generally have a sense of fairness, right? They want to be treated fairly relative to others and relative to the opportunities that they have out there in the overall market,” he said.

If discontent isn’t apparent right away, Groysberg said, that doesn’t mean it won’t manifest in the next compensation cycle.

Stars Managing Stars

Asked to comment on the potential management challenges for the Meta Superintelligence Lab, a company spokesperson told Bloomberg, “We know there’s a lot of interest in MSL, and seemingly everyone has an opinion, but we’re just focused on doing the work to develop personal superintelligence.”

Zuckerberg disputes some of the press reports about the specific packages he’s offering to AI experts but defends his strategy of recruiting an all-star team, telling The Information in a recent interview that AI “is going to be something that is the most important technology in our lives. It’s going to underpin how we develop everything at the company, and it’s going to affect society very widely. So we just want to make sure that we get the best folks to work on this, from entrepreneurs to researchers to engineers working on the data and infrastructure.”

Zuckerberg has arguably placed his largest bet on Wang. Meta invested $14.3 billion in Scale AI, without taking full control of the company. Scale AI generated $870 million in revenue last year offering data services to train AI systems. Some of its most prominent customers, including Google and OpenAI, are reportedly cutting ties with the company in the wake of Meta’s investment, fueling debate over whether Meta’s true goal was to acquire Wang and not the 49% stake it now holds in the business.

Though Wang has prodigious talent for math and science and now a successful track record as a founder — he dropped out of MIT after a year to co-found Scale AI in 2016 — neither he nor his startup has produced breakthrough AI research. At Scale AI, Wang’s team worked with an army of data-labeling contractors who provided what he described as the “picks and shovels” for the AI gold rush. At Meta, he’ll now have to earn the respect of a team of world-class AI scientists.

Wang has a reputation for relentlessness and set a demanding pace at Scale AI, declaring that “too much is the right amount,” as he wrote in 2024. Wang also became known for his drive to keep the organization as “talent dense” as possible, looking for people who would be able to match his ambition.

According to Groysberg, who recently wrote a case study about Scale AI, Wang participated in weekly hiring meetings, personally reviewing each candidate’s materials and forcing managers to rigorously defend their choices. “Alex goes through every single person in every function with a fine-tooth comb,” one executive told Groysberg. “He very much cares about hiring impressive people. And the way we measure that a lot of the time is through that balance of IQ and EQ. And then you bundle that up with just an extreme amount of grit.”

Whether what worked at Scale will translate to Meta remains to be seen.

“I think Meta got the smartest person that I know,” Groysberg said. “In that area, they got the star. And then the question is: Can they build the star team?”

More stories like this are available on bloomberg.com

©2025 Bloomberg L.P.