How much value can one person bring to a company? In Silicon Valley the answer is hundreds of millions — as companies offer nine-figure pay packages for top talent in an artificial intelligence arms race.
Mark Zuckerberg’s Meta Platforms is said to have agreed a multi-year pay deal of $200 million to poach Ruoming Pang, a former Apple engineer, for his “superintelligence” team.
Google, meanwhile, signed a $2.4 billion deal this month to bring in Varun Mohan, co-founder and chief executive of Windsurf, the AI coding start-up, and other senior Windsurf research and development employees.
“I’m focused on building the most elite and talent-dense team in the industry,” Zuckerberg wrote on Threads, one of his social media platforms, last Monday. “We’re also going to invest hundreds of billions of dollars … to build superintelligence.”
Apple, Microsoft and Amazon are also fighting over tech engineers as though they are professional athletes. Sarah Guo, a Silicon Valley venture capitalist, claimed on X that some top candidates were using agents to negotiate their compensation packages like international footballers.
The AI investor Deedy Das said three friends working as machine-learning engineers or researchers at OpenAI or Anthropic told him they had been offered between $8 million and $20 million in total annual compensation to join Meta. The sum included cash sign-on bonuses and stock options, said Deedy, who works at Menlo Ventures, a Palo Alto-based investor in Anthropic.
“There’s a lot of Fomo [fear of missing out],” he said. “I think if you break it down psychologically, it is such a massive jolt to your notion of how the world works, right?
“Some of the people who’ve been offered this money are like, ‘I planned out a career and I would be doing this for 30 years, and I would maybe make this amount of money in total, and now I’m making it in one year.’
“Similarly, everybody who has that skill, or believes that they have that skill and hasn’t been approached, is asking the question, ‘What am I doing? Why am I not there? Or, should I be targeting that job?’”
‘Crazy talent bubble’
The talent wars kicked off in November 2022 when OpenAI launched ChatGPT, forcing tech companies to dive head-first into artificial intelligence. But Silicon Valley insiders say the competition for the most exceptional industry talent has intensified in recent weeks.
In a blog post last Sunday, John Luttig, a partner at the San Francisco-based Founders Fund, said the tech industry was in a “crazy AI talent bubble” that represented “a new normal for the foreseeable future … If the top 1 per cent of companies drive the majority of venture capital returns, why shouldn’t the same apply to talent?”
Analysis of filings for the H-1B visa, which is offered only to specialist employees of exceptional ability, show that Meta has offered one software engineer a top base salary of $480,000 before stock options and bonuses. Salaries for machine learning engineers at Meta’s headquarters in Menlo Park, California, also exceed $400,000.
At Thinking Machines Lab, the AI start-up launched by Mira Murati, former chief technology officer of OpenAI, a base salary of $450,000 was listed for a machine learning scientist. Meanwhile, at OpenAI itself, led by Sam Altman, research engineers have been offered base salaries of $440,000.
Engineers at tech firms typically receive a large portion of their compensation in stock. OpenAI paid a total of $4.4 billion in stock-based compensation last year, according to The Information, a technology news media site. OpenAI presently has about 5,300 employees, equating to an average of $830,200 in stock compensation per person.
Graduates falling out of the market
A sharp divide in the talent wars is emerging as graduate tech roles are cut while experienced engineers are being showered with huge offers. A report published in May by SignalFire, the San Francisco venture capital firm, found that graduate hiring in tech had fallen 50 per cent compared with pre-pandemic levels.
“As AI tools take over more routine, entry-level tasks, companies are prioritising roles that deliver high-leverage technical output,” the report found. “So far, the fallout has hit new grads hardest, while demand for experienced engineers is still rising. As AI capabilities continue to grow, the real differentiator will be talent, and no one’s playing the game harder than the top AI labs.”
• Where — and who — are America’s foreign students? In charts
Das said the most sought-after talent either has “the ability to train models at scale”, or “the ability to develop infrastructure to run models at scale”.
Some tech companies are paying premiums of $200,000 on top of a base salary for some roles with machine learning skills, according to a report by J Thelander Consulting. The consulting firm analysed pay across 153 companies and found that data scientists and analysts with machine learning skills typically received a higher premium than software engineers with the same skills.
• Silicon Valley’s AI-fuelled madness has echoes of the dotcom crash
Jody Thelander, chief executive of J Thelander Consulting, said these hires were “critical for their company’s success, so they have to pay because without that person, they can’t get the job done”. She added she was surprised by the “difference between the median and the top people” in terms of compensation levels.
Not all about the money
David Chie, chief executive of the talent acquisition company Maple Drive, said AI researchers were not motivated solely by money.
“I’m seeing people being more driven to solve greater societal problems that can be assisted with AI, versus just around the dollars,” he said. “The top researchers, I think, are being attracted to interesting problems that weren’t solvable previously.”
Even so, he expects companies to continue to pay huge sums to lure the best talent. Not many individuals, he said, “have been working in machine learning and AI prior to it becoming a popular talking point two years ago. There’s a lot of taming of the lion … AI requires a lot of brilliant engineering expertise that honestly just isn’t available and so these people are able to command large salaries.”