Meta Offers $36 Million Minimum Salary to Senior Engineer at AI Startup
The My First Million podcast revealed that Meta has offered a senior engineer at Brett Adcock's AI startup a minimum salary of $36 million to entice them to leave. Adcock stated that every candidate contacted received some absurd offer.
Adcock believes that only about 20 to 30 people in California truly know how to build top AI models. The engineer rejected Adcock's offer of approximately $15 million to $20 million in startup equity in favor of Meta's four-year restricted stock units.
Over the past year, Meta has been "buying into" the AI race with high offers. Adcock thinks this strategy is smart and would do the same if in a similar position, but he is concerned about hiring "mercenaries" who are only in it for the money.
The discussion highlights the extreme scarcity of AI talent, with leading companies competing for limited experts with salaries far exceeding the norm.
The talent war is extending from model capabilities to organizational and compensation structures.
From a market mechanism perspective, ultra-high minimum offers are driving up the costs of top AI talent, with funds flowing from large companies to scarce experts; the event is driven by talent competition, benefiting large tech companies that can offer minimum guarantees, while startups must rely on mission and equity to attract talent.
Source: Public Information
ABAB AI Insight
Brett Adcock operates both humanoid robotics and AI-related companies, and his direct encounter with Meta's $36 million minimum offer reveals the extreme supply-demand imbalance in the current AI talent market.
On the capital front, Meta is converting resources into talent acquisition through high RSUs, motivated by the need to quickly fill gaps in cutting-edge model capabilities, specifically by offering minimum guarantees far exceeding startup equity to a few experts.
Similar cases can be seen in the high-priced talent competition between OpenAI, Anthropic, and Google, as well as historical expert raids in the semiconductor or search fields; the current AI industry is at a stage where experts who truly understand full-stack model building are extremely scarce.
This essentially represents capital concentration: scarce talent is rapidly concentrated by large company capital, with the mechanism being to replace the uncertainty of startup equity with the certainty of cash equivalents, thereby locking in a limited pool of experts.
ABAB News · Law of Cognition
On the day of the $36 million minimum offer, the talent market had already departed from conventional salaries.
There are probably only 20 to 30 people who can truly build models.
Mercenaries can be bought, but a sense of mission cannot.