Mark Zuckerberg is recruiting for his “superintelligence” team, which will work to make artificial intelligence systems capable of outperforming humans on nearly all tasks. But the Meta boss isn’t simply putting the job listings for roles with potentially over $100 million pay packages online; he’s on a much more focused head-hunting mission.
In fact, he has created a list of young AI talent that could turn around his company’s less-than-stellar reputation in AI, according to The Wall Street Journal, and replace the growing number of researchers who are leaving his company for competitors. The researchers and engineers included on “The List,” who have been contacted this month, all hold PhDs from institutions such as Berkeley, Stanford, Carnegie Mellon, and MIT in relatively niche AI fields.
Over a decade ago, when many of them earned their PhDs, broader fields such as generative AI and robotics were not nearly as hot as they are now. As a result, few engineering graduates chose to pursue niche specialisations within these areas, meaning that individuals with significant expertise in them today are few and far between.
Yu Zhang, an OpenAI researcher in automatic speech recognition, was advised by his mentor to abandon his speciality just before breakthroughs in deep learning revolutionised the field. Because of the unique experience he has gained in the years since, he has made it on Zuckerberg’s list, according to the WSJ.
Those on The List are pals, making life harder for Meta
Most of the potential recruits on The List are similar ages, have worked for the same companies, live in the same cities, and research overlapping topics; as such, most of them already know each other.
Moreover, as they likely didn’t pursue their once-unfashionable niches with the aim of becoming multimillionaires, their loyalties tend to lie with one another rather than the companies now dangling eight- or nine-figure salaries.
As a result, they are talking among themselves, ensuring there is no black box of secrecy around who is being offered what, the WSJ says. Some are negotiating package deals, such as Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai, who all left OpenAI for Meta together.
Meta, therefore, must work especially hard to get those on The List through the door. According to the WSJ, Zuckerberg is getting his hands dirty in the search for around 50 individuals to join his superintelligence team, scouring the most-cited papers in his areas of interest to identify their authors and conducting in-person interviews himself.
Earlier this month, Meta invested $14.3 billion for a 49% stake in Scale AI, a data-labelling startup. As part of that deal, Scale AI’s co-founder, 28-year-old Alexandr Wang, joined Meta and is expected to lead research on superintelligence.
Meta has also made offers to OpenAI staff, according to Sam Altman. “It is crazy. I’m really happy that, at least so far, none of our best people have decided to take them up on that,” he said on his brother Jack Altman’s podcast earlier this month. Nevertheless, The New York Times reports that at least four OpenAI employees have accepted offers.
Meta’s poor reputation with AI is putting off top talent
Unfortunately, Meta’s patchy reputation with AI has capped its success, costing it OpenAI’s co-founder, Ilya Sutskever, and chief research officer, Mark Chen, the WSJ says. According to The Information, the Llama 4 announcement was delayed at least twice due to the models underperforming in technical benchmarks and conversationality.
What’s more, when Llama 4 was released, experts noted that the version published on the LMArena leaderboard differed from what was publicly available, suggesting that Meta submitted an altered version that would rank higher. Its early decision to open-source its models, intending to see its technology become the go-to for developers, backfired when DeepSeek released models based on Llama but were more advanced and less expensive.
Zuckerberg is now considering de-investing in Llama, the NYT reported this week. Further damaging the company’s reputation are Meta’s AI chatbots engaging in seriously inappropriate conversations, as well as legal battles with creatives accusing it of stealing their work for training data.
Hiring the best AI brains is the hottest trend in the race for AI dominance
Meta is not alone in its efforts to build the smartest AI research team in the biz. Other tech giants, including Anthropic and Google, have been engaged in aggressive recruitment strategies, also acquiring entire companies to bring talent in-house. Google spent $2.7 billion on Character.AI last year, and OpenAI acquired Jony Ive’s startup, io, in a $6.4 billion all-equity deal last month.
While such recruitment drives sound like big money, they are a drop in the ocean compared to the cost of building data centres and other infrastructure required to achieve the same goal of taking the lead in the AI race. Bloomberg Intelligence reports that Meta, Amazon, and Microsoft will invest $371 billion into data centres and computing resources in 2025, with annual spending projected to surge past $525 billion by 2032.
