Big Tech companies have been in an all-out bidding war to capture top AI researchers and engineers — perhaps none more conspicuous than Meta. CEO Mark Zuckerberg is trying to leapfrog his way to superintelligence, reportedly offering compensation packages in the hundreds of millions of dollars. One researcher at the startup Thinking Machines recently told the Wall Street Journal he turned down a Meta deal worth $1.5 billion.

They’re the kind of eye-watering sums you usually only hear about in professional sports. So we wondered: Does this strategy of collecting expensive superstars usually pay off for teams?

When it comes to an all-star roster, there’s one example in basketball that always comes up: the Dream Team, the U.S. mens basketball team at the 1992 Barcelona Olympics.

“The Dream Team is essentially, probably the greatest team ever assembled,” said David Berri, a sports economist at Southern Utah University and author of “The Wages of Wins.”

“They go to the Olympics, where nobody else has professional players, and you have the greatest professional players ever,” he said. “It has Larry Bird, Magic Johnson, Michael Jordan, Charles Barkley, Karl Malone, David Robinson — these are all Hall of Famers.”

The Dream Team dominated at the Olympics, winning most games by dozens of points and easily taking home the gold medal.

“You assembled this phenomenal amount of talent, and it worked out exactly as you hoped it would,” said Berri.

But he notes that kind of outcome is the exception, not the rule. Dream-team-like dominance is rare in professional sports. Despite all the statistical analysis and big money contracts, he said it’s still hard to buy your way to wins, often because we misjudge what makes a player — or an AI hotshot — valuable.

“Guess what? A star may be a star, not just because they are fantastic in their work, but because of the team around them,” said Ashish Nanda, a senior lecturer at Harvard Business School.

He pointed out star players sometimes lose some of their luster when they change teams. And the same is true in business. He studied top-performing stock market analysts who lost their edge after moving to a new firm.

“If you think that you can solve your problems by bringing smart people who have done well outside, paying them a ton of money and bringing them in and saying, ‘OK, do your magic again,’ that's a very risky business,” he said.

Those massive paychecks could also be a risk, said economist Michael Gove at the University of North Georgia. The research is mixed but pay inequality between stars and other players has sometimes been found to undermine team cohesion.

“Where people think things are maybe a little unfair, a little skewed, too much, and so it ends up having worse team performance,” he said.

The Yankees don’t always win the World Series, and Meta might not win every AI battle.

It’s possible to have too much of a good thing, said Eric Anicich, a management professor at the University of Southern California who describes a too-much-talent effect, “kind of challenging, this common and quite reasonable assumption that people have that adding more and more talented people to a team is just necessarily going to always increase team performance.”

Anicich and his co-authors found teams with the highest concentration of stars underperformed compared to more average lineups in sports that require a high degree of cooperation, like basketball. 

“A player has to be comfortable giving the ball up, right, which is not often the case if you're thinking about superstars in sports contexts,” he said.

In the AI context breakthroughs can come from individual prodigies, said Anicich, but bringing those ideas to market requires a lot of collaboration.

However, drawing conclusions from the outcomes of American sports might not be instructive in the AI race, said Stefan Szymanski, a professor of sport management at the University of Michigan and author of “Soccernomics.” He points out most American leagues impose some form of salary cap to level the playing field; not so in European soccer.

“There are essentially no restraints on what a team can spend,” he said. “Soccer is characterized by extreme inequality between the top and the bottom. The same teams win year in year out.”

He said the teams that spend the most, generally win the most. But the ball can always bounce either way.