YouTube
Spotify
Apple Podcasts
In the past few weeks, one new theme has emerged in the AI race: talent wars. Mark Zuckerberg has put this into overdrive, offering huge salaries to convince engineers/scientists to move from Apple, Alphabet, OpenAI, and other start-ups to Meta’s Superintelligence Team. He spent $14 billion to acquire the founder of Scale and get a 50% stake in that business, which in a vacuum is a giant overpay.
On this week’s podcast, I said I wouldn’t fault Zuckerberg for this move. If (a big if!) winning AI means generating trillions of dollars in value, then it makes sense to pay these top employees like star athletes. Maybe even more for the best players.
It does not mean you will win. Talent is only one part of the equation. You need a mission, a cohesive team, and a culture of excellence.
Homegrown talent usually leads to the best outcomes. Acquisitions are riskier, but can lead to long-term cohesion within the larger enterprise (example: Hassabis/Deepmind staying lockstep with the Google founders).
Let’s look at sports as an example.
Super teams that simply spend to acquire as much talent as possible usually underwhelm. For example, the New York Nets superteam.
Homegrown super teams like the Golden State Warriors only had a small amount of talent not nurtured through its franchise. They have won four NBA championships. The New York Nets were an abject failure.
They even had an overlap of the best player: Kevin Durant. Clearly, more than just talent matters.
Or take the exact same team in international soccer: Paris Saint-Germain.
They spent an unlimited budget to get the best players from France, Argentina, and Brazil on their roster. Mashing different playing styles and egos together, coming from different backgrounds, the team massively underperformed.
Now, none of these players remain…and PSG just won the Champions League.
Back to big tech and AI.
Zuckerberg is saying with his dollars that Meta is falling behind in AI. But just throwing money at the problem will not guarantee a fix (or catching up to ChatGPT and Gemini). A duct-taped-together superteam usually underwhelms. I expect the same from Meta.
I would bet on the cultures of innovation, research, and product excellence at OpenAI and Alphabet shining through over the long haul. The best researchers still want to work at OpenAI and Alphabet. Even if a few “big names” leave for Meta and take the money, I think this will remain.
Do you think these already handsomely paid researchers will risk not being on the team that changes the world just to collect a few more million dollars?
After seeing the developments in AI over the past few months, I am more bullish on the long-term prospects of OpenAI and Alphabet.
-Brett