The future of AI is up for grabs

Kevin Dietsch via Getty Images

July 22, 2025 (businessinsider.com)

$100 million salaries. Deliberately inflating counteroffers for staff they don’t want to keep. Sabotaging competitors — not through code, but through culture.It takes a lot of cunning to thrive in the AI talent wars. A lot of money, too. OpenAI’s CEO, Sam Altman, said last month it was “crazy” that Meta was offering up to $100 million to lure away his researchers. Laszlo Bock, a former VP of HR at Google, thinks Meta’s move was actually very rational. It could even be a “bargain,” he told BI. In Silicon Valley, it’s the summer of compensation FOMO. If a tech worker isn’t fielding at least one jaw-dropping offer, they’re starting to wonder what they’re doing wrong.Why are the salaries so high? Because the stakes are. The future of AI is up for grabs, and there may be fewer than 1,000 top researchers capable of building it. The pressure to attract — and keep — them is unprecedented. Bock pulled back the curtain on Google’s counteroffer playbook, which included getting a multimillion-dollar package in the hands of employees within “60 minutes” of hearing a rival had made a move. 
It’s not all about staying ahead. Sometimes, it’s about making sure the competition stays behind. Game theory” is how Bock put it.Even if Google didn’t want to keep a staff member, Bock said they would counteroffer to force the poaching company to spend more. The goal? To throw off team dynamics and disrupt the other company’s culture.If a “mediocre” person joined a competitor earning much more than their new colleagues, it could cause “tremendous internal tension,” Bock said. “Done correctly, there’s a lot of actual strategy behind it and it’s super fun.”This isn’t just a clash of the titans. Startups are scrapping with Big Tech for AI talent — and they have an edge.Startups can’t compete with Meta’s $100 million compensation packages, but they don’t have to. As Natan Fisher, an investor and cofounder of search firm SingleSprout, told BI: “They can win by pitching sharper problems, faster cycles, and giving top AI engineers the ability to own product, with strong upside.”
Read the full story

Leave a Reply

Your email address will not be published. Required fields are marked *