AI Talent War: $250M Salaries Eclipse Historic Scientific & Sports Deals

Silicon Valley’s AI talent war has reached an unprecedented peak, with compensation packages far exceeding historical benchmarks for scientific and technical achievements. Recently, Meta reportedly offered AI researcher Matt Deitke a staggering $250 million over four years, averaging $62.5 million annually, with potentially $100 million in the first year alone. This figure shatters all known precedents for scientific and technical salaries, including those from the most significant scientific milestones of the 20th century.

Deitke, who co-founded the startup Vercept and previously led development of the multimodal AI system Molmo at the Allen Institute for Artificial Intelligence, possesses expertise in systems that process images, sounds, and text—precisely the technology Meta is keen to develop. His recruitment is part of a broader trend; Meta CEO Mark Zuckerberg reportedly extended an offer of $1 billion over several years to another unnamed AI engineer. These astronomical figures underscore what major tech companies believe is at stake: a high-stakes race to develop artificial general intelligence (AGI) or superintelligence—machines capable of performing intellectual tasks at or beyond human levels. Companies like Meta, Google, and OpenAI are betting that the first to achieve this breakthrough could dominate markets worth trillions, driving compensation to unprecedented levels regardless of whether this vision is realistic or speculative.

To contextualize these sums, consider historical scientific endeavors. J. Robert Oppenheimer, who led the Manhattan Project during World War II, earned approximately $10,000 annually in 1943. Adjusted for inflation, this amounts to about $190,865 in today’s dollars—comparable to a senior software engineer’s current salary. In stark contrast, 24-year-old Deitke, a recent PhD dropout, is poised to earn roughly 327 times Oppenheimer’s salary during the development of the atomic bomb.

Even the salaries of legendary figures from the Space Race pale in comparison. Neil Armstrong, the first person to walk on the moon, earned approximately $27,000 annually, equivalent to about $244,639 today. His crewmates, Buzz Aldrin and Michael Collins, earned even less in today’s terms. Current NASA astronauts earn between $104,898 and $161,141 per year. Remarkably, Meta’s AI researcher stands to earn more in just three days than Armstrong made in an entire year for his historic “giant leap for mankind.”

Engineers who designed the rockets and mission control systems for the Apollo program also earned modest salaries by today’s standards. A 1970 NASA technical report analyzing engineering salaries revealed that a newly graduated engineer in 1966 started with $8,500 to $10,000 annually (about $84,622 to $99,555 today). An engineer with a decade of experience earned around $17,000 ($169,244 today). Even the most elite, top-performing engineers with 20 years of experience peaked at approximately $278,000 per year in today’s dollars—a sum a top AI researcher like Deitke can now earn in a matter of days. Furthermore, Armstrong’s moon mission included an $8 per diem, roughly $70.51 today, before deductions for onboard “accommodations.”

The collaborative spirit of past innovations also contrasts with today’s individual mega-deals. During Bell Labs’ golden age, when foundational technologies like the transistor and information theory were developed, the lab’s director earned only about 12 times what the lowest-paid worker did. Claude Shannon, the creator of information theory, worked on a standard professional salary in 1948 while laying the mathematical groundwork for all modern communication. Similarly, the “Traitorous Eight” who left William Shockley to found Fairchild Semiconductor—the company that effectively birthed Silicon Valley—began with seed funding of $1.38 million ($16.1 million today) for the entire company. This amount is a mere fraction of what a single AI researcher now commands.

Even the highest salaries from the early tech era pale in comparison. Thomas Watson Sr., IBM’s legendary CEO, received $517,221 in 1941—the third-highest salary in America at the time, equivalent to about $11.8 million in 2025 dollars. A modern AI researcher’s package represents more than five times Watson’s peak compensation, despite Watson building one of the 20th century’s most dominant technology companies.

This compensation trend has also surpassed the earnings of many top athletes. Steph Curry’s most recent four-year contract with the Golden State Warriors was $35 million less than Deitke’s Meta deal, leading observers to liken this to an “NBA-style” talent market, where AI researchers now out-earn even basketball stars, though soccer superstar Cristiano Ronaldo’s recent $275 million annual earnings stand as a high benchmark in the sports world. While premium prices for technical talent are not entirely new—in 2012, three University of Toronto academics auctioned themselves to Google for $44 million ($62.6 million today), and by 2014, a Microsoft executive was comparing AI researcher salaries to NFL quarterback contracts—today’s numbers dwarf even these recent precedents.

Several factors explain this unprecedented explosion in compensation. The current landscape features a level of industrial wealth concentration not seen since the Gilded Age of the late 19th century. Unlike past scientific endeavors, today’s AI race involves multiple companies, each with trillion-dollar valuations, fiercely competing for an extremely limited talent pool. Only a small number of researchers possess the specific expertise required for the most advanced AI systems, particularly in niche areas like multimodal AI, Deitke’s specialty. This scarcity is amplified by the pervasive “AI hype,” positioning it as the next major technological revolution.

The economics fundamentally differ from past projects. While the Manhattan Project cost a total of $1.9 billion (approximately $34.4 billion adjusted for inflation), Meta alone plans to spend tens of billions annually on AI infrastructure. For a company nearing a $2 trillion market capitalization, the potential payoff from being the first to achieve AGI dwarfs even Deitke’s substantial compensation package. As one executive candidly told The New York Times, “If I’m Zuck and I’m spending $80 billion in one year on capital expenditures alone, is it worth kicking in another $5 billion or more to acquire a truly world-class team to bring the company to the next level? The answer is obviously yes.”

This intense competition has empowered researchers. Young AI professionals are known to maintain private chat groups on platforms like Slack and Discord to share offer details and negotiation strategies. Some even enlist unofficial agents to navigate this lucrative market. Companies are not only offering massive cash and stock packages but also significant computing resources; some potential hires have reportedly been promised allocations of up to 30,000 GPUs, the specialized chips vital for AI development.

Tech companies view themselves as engaged in an “arms race” where the winner could profoundly reshape civilization. Unlike the Manhattan Project or Apollo program, which had specific, finite goals, the pursuit of artificial general intelligence ostensibly has no upper limit. A machine capable of matching human intelligence could theoretically improve itself exponentially, potentially leading to an “intelligence explosion” and cascading discoveries—if such a scenario truly materializes.

Whether these companies are truly building humanity’s ultimate labor replacement technology or merely chasing an unprecedented wave of hype remains an open question. What is clear, however, is the vast distance traveled from the modest $8 per diem Neil Armstrong received for his moon mission. After Deitke accepted Meta’s offer, Vercept co-founder Kiana Ehsani lightheartedly quipped on social media, “We look forward to joining Matt on his private island next year.”