Skip to content 
Meta CEO Mark Zuckerberg is offering unprecedented pay packages to attract elite artificial intelligence researchers as he builds a team to lead Meta’s superintelligence ambitions. Reports suggest some of the offers go as high as $300 million over four years, including more than $100 million in total compensation for the first year alone. This aggressive recruitment strategy is primarily aimed at poaching top talent from rival firms, especially OpenAI. The move signals Zuckerberg’s belief that nothing matters more for the future of Meta than achieving breakthroughs in AI.
Also read: U.S. Leads the Global AI Race
Meta’s Billion-Dollar Bet on Superintelligence
Zuckerberg has formed a new division called Meta Superintelligence Labs (MSL), which is tasked with developing artificial general intelligence (AGI). In a note to employees, he confirmed that this team will lead the company’s push toward next-generation large language models (LLMs). Alexandr Wang, the 28-year-old former CEO of Scale AI, and Nat Friedman, former GitHub CEO, are co-leading MSL. Though neither are traditional researchers, they are overseeing a unit now filled with top-tier AI talent.
According to Wired and The Wall Street Journal, Meta has extended at least 10 massive compensation offers to researchers from OpenAI, Anthropic, and Google. Among the new hires are notable figures like Trapit Bansal, co-creator of OpenAI’s o-series models, and Shuchao Bi, known for his contributions to GPT-4o’s voice capabilities. Others include Huiwen Chang from Google Research, Ji Lin, Joel Pobar, Jack Rae, Johan Schalkwyk, Pei Sun, Hongyu Ren, Jiahui Yu, and Shengjia Zhao.
The pay packages reportedly include immediate stock vesting, a rare perk even in Silicon Valley. One senior engineer at Meta confirmed to Wired that their pay was around $850,000 per year. Engineers in Meta’s top bracket (E7) average about $1.54 million annually, according to Levels.fyi data. However, offers made to leadership candidates in MSL go far beyond that. Meta spokesperson Andy Stone stated that some claims regarding compensation have been exaggerated, adding that “some people have chosen to greatly exaggerate what’s happening for their own purposes.”
OpenAI Pushes Back Amid Talent Exodus
Meta spent heavily on hiring employees and has caught the attention of OpenAI. In a note to staff entitled How to think about Meta, Mark Chen, the chief research officer of OpenAI, wrote: It was a bit like someone has broken into our home and stolen something. This was reinforced by CEO Sam Altman in a podcast, who added that Meta had offered some of his team members a 100 million signing bonus but had been advised that none of his better talents had taken them up.
Even after these claims, Meta has already succeeded in poaching at least seven employees of OpenAI. The backlash has been so pronounced that OpenAI has thought of making changes to its internal compensation systems. Nonetheless, Chen stressed that the change would not be made in favour of some future development at the expense of equity. He also mentioned that OpenAI is committed to having more supercomputers online towards the end of the year, and that was one of the pain areas that the organisation had among its researchers: the lack of access to GPUs.
Access to Compute and Resources Becomes Key Incentive
One of the major draws for AI researchers joining Meta is the promise of unlimited compute access. Unlike at OpenAI, where staff have previously complained about unfulfilled hardware promises from leadership, Zuckerberg has assured new hires that they will not run out of resources. In a field where cutting-edge GPUs are essential for training powerful models, this pledge carries significant weight.
According to the Wall Street Journal, Zuckerberg has positioned Meta’s AI unit as a place where engineers can experiment freely without worrying about infrastructure bottlenecks. In Silicon Valley, where high compensation is common, access to resources can be the decisive factor for elite talent. Despite some researchers denying the scale of the compensation offers on social media, the incentives—monetary and otherwise—remain unparalleled.
Meta Executives Explain the Extent of Offers
In a recent company-wide meeting, Meta Chief Technology Officer, Andrew Bosworth, squashed the rumor mill within the company. He cleared the air that not all employees are collecting packages of 100 million. The market is smoking! Look you fellows. It is not very hot. Okay?” Bosworth remarked, and such plasma is only used at a very small number of top spots in leadership.
He underlined that $100 million includes a number of elements and was not a simple sign-on payment. Bosworth also attacked Altman over the latter spreading a false impression of the scope of Meta hiring spree since “we are winning the war of recruiting talent at OpenAI.” As part of its recruitment drive, some of the new employees have discredited such touted pay levels of up to $300 million as nothing but fake news, but the extent of its hiring exercise is difficult to ignore.
The vision of Meta promoted by Zuckerberg is getting more precise. He hopes to make Meta a formidable competition in the AGI race by loading his AI department with some of the smartest people and eliminating resource limitations. Deal books also say that the deep pockets and readiness to take risks by the company have already ruined the established balance of talent in the industry, even though its majority still remains averse to the model.
The other tech giants Google and Microsoft are monitoring the AI talent war closely. Compensation adjusted by OpenAI, and freedom and resources promised by Meta have put the industry on alert at the prospect of a long-lasting battle over the best artificial intelligence minds.
Comments 1