A new gold rush
There has never been such a talent frenzy in tech as today. Artificial intelligence burst in not just as “the latest trend,” but as a true driver of economies and companies. A full‑on hunt for AI talent has begun in the global market. At the center of this race are the largest IT giants: Google, Microsoft, Amazon, Apple, OpenAI, and of course Facebook, which today is increasingly referred to simply as Meta.
If ten years ago the phrase “artificial intelligence” was associated with research labs and debates about the future, today it means real money, billions of dollars in investment, and the fate of companies. It’s no longer programming languages or frameworks that are the most valuable resource, but human minds—people who can invent and deploy algorithms that change everyday reality.
Civilization‑scale ambitions
Meta has repeatedly proven it can sense trends and invest “in the future.” After the rebrand and massive investments in the metaverse, Facebook found itself in a situation where the stakes only grew. But the market immediately took a new turn: after the release of ChatGPT at the end of 2022, a real AI revolution began. The world saw that large language models (LLMs), generative AI, and self‑learning systems are not science fiction but a new digital “steam engine.”
Meta quickly realized: the winner won’t be the one with the most servers, but the one with the best people. The largest corporations and startups around the world started not just hiring—they began to poach, buy, sometimes literally “steal” the best specialists in artificial intelligence. Facebook set a new bar for this race by offering AI stars contracts that recently seemed impossible even for Fortune 500 top managers.
Millions for brainpower
Until 2022, the AI talent market was already overheated. The best specialists could earn $500K–$1M per year (including stock, bonuses, and options), work remotely, and choose from a dozen offers. But 2024–2025 changed everything.
The biggest “signs” are universities like Stanford, MIT, and Oxford; labs such as DeepMind, OpenAI, Anthropic, and Google Brain; and startups like Scale AI, Cohere, and Runway. Every graduate with a solid background in machine learning, deep learning, or reinforcement learning is a potential millionaire. But even among them there are “superstars”: authors of breakthrough algorithms, architects of the largest LLMs, and creators of new optimization and generation methods.
Facebook began poaching such people, offering sums that sound like lottery winnings: $100–250 million (!) to make the move, including options, bonuses, guarantees, and signing awards. Some experts compare this to football transfers—only now the world’s main “players” are not on the pitch but in data centers, writing code and designing model architectures.
Bright examples of “AI transfers”
- 1. Zhuomin Pang (Zhuoming Pang) — $250 million
A legendary AI architect from Apple who stood at the origins of generative networks for mobile devices. Facebook offered him an incredible package to lead a new track of personal LLMs for messengers and the metaverse. - 2. Aleksandr Van (Alexandr Wang) — $200 million
The founder and “face” of Scale AI, one of the youngest billionaires in Silicon Valley, became Meta’s principal advisor on infrastructure and training models on petabyte‑scale data. - 3. Trapit Banzal (Trapit Bansal) — $100 million
An OpenAI star who worked on optimizing distributed training of large language models. His departure is one of the year’s most talked‑about “moves.” - 4. Net Fridman (Nat Friedman) — $100 million
The former head of GitHub and a technology visionary known for investing in AI startups, moved to Meta as Chief AI Product Officer. - 5. Deniel Gross (Daniel Gross) — $50 million
One of the cofounders of Safe Superintelligence, a mentor to dozens of young AI teams, and a key Silicon Valley “angel.”
All these specialists are being invited on terms never before seen in IT history: we’re talking signing bonuses, options on Meta stock, and the freedom to recruit their own teams from the best people on the market.
Why AI talent is the new oil field
AI today is the new oil. Just as 20th‑century oil created corporate might, AI talent now determines who will own the digital future. Platforms no longer compete only on infrastructure—almost everything can be rented or bought. You cannot buy brains, creativity, and vision. The most powerful GPU clusters, the most expensive data centers, vast amounts of data—without talented specialists it’s all just “sand.”
Companies that win this race will secure decades of technological and economic leadership. It’s no surprise Facebook, Google, and Microsoft now compete to “assemble AI teams” as fiercely as they once fought over the advertising market.
Price tags we’ve never seen before
Back in 2019, $500,000 for an engineer in the US seemed extraordinary. In 2025, such sums no longer shock: if a specialist truly innovated in the LLM market, worked with the largest language or vision models, and participated in leading experiments with self‑supervised learning, reinforcement learning, or AIGC (AI‑Generated Content), they can be offered just about anything.
It has reached the point where major venture funds (Sequoia, Andreessen Horowitz, Y Combinator) not only invest in teams but also deliberately “buy out” developers for their portfolio companies. Startups are racing to announce record salaries just to get a “growth engine.”
What is changing in the industry?
1. Facebook (Meta) declares ambitions to become the No. 1 AI company
Meta has long been more than a social network. Over the past three years, the company has invested tens of billions in the metaverse, AR/VR, generative models, and data‑center infrastructure. The current bet is to make Facebook and Instagram the “smartest” platforms: content generation, recommendations, moderation, search, and communications.
Huge resources are aimed at overtaking OpenAI and Google in the quality and accessibility of AI services for everyday users. The company openly says: “Without the best AI talent, we will lose this race.”
2. Leadership is not in infrastructure but in people
Today cloud, compute, and data centers are “commodities.” The real value lies in those who invent new algorithms and see how to squeeze more out of hardware and data than competitors. Meta is poaching not just developers, but architects, research leads, theoreticians, and people who understand how to scale and embed AI into real products with audiences in the billions.
3. The compensation market: a new reality
AI developers are now the most expensive specialists in the IT market. Salaries for Senior Machine Learning Engineer, Deep Learning Scientist, and Chief AI Officer start at $300,000 per year and very quickly rise to $1 million and above. As the market caps of Meta, Microsoft, Amazon, and OpenAI grow, the equity component can far exceed the regular “cash” salary.
4. An explosion of startups
Nearly every AI specialist who is laid off or moves to Meta, OpenAI, Google, or Microsoft starts their own startup within a year—or becomes a co‑owner of a new “unicorn.” Money is flowing, the market is not saturated, and demand for skills keeps rising.
5. The changing role of universities and open source
Stanford, MIT, Berkeley, and Oxford are becoming not just “talent foundries,” but centers of real applied science and startups. The best AI talent now simultaneously works in industry and teaches—or mentors their research teams.
Open source is becoming the main source of new ideas and reputation: every breakthrough project on HuggingFace, GitHub, and Papers with Code is immediately “snapped up.” Experience as an open‑source contributor is currently valued higher than any formal diploma.
What this means for everyday developers and the job market
- 1. The bar to entry is higher—but so are the opportunities
Yes, the battle for AI talent is happening at the very top, but it opens new paths for everyday developers too. Anyone who can learn, adapt, and master new approaches can become part of this wave. Thousands of vacancies in machine learning, data science, DevOps for ML, MLOps, and AI products in C#, Java, Python, Go, and Rust appear every day. - 2. It’s not tenure that’s rewarded, but results and open‑source reputation
For AI startups, large companies, and investors, what matters far more today is not the number of years on a résumé but real contribution: successful projects, publications, open‑source contributions, and solutions to hard problems. If you want to enter the AI field, start building mini‑projects, publish your ideas, and participate in Kaggle and other competitions. - 3. It’s never too late to learn
The industry already has dozens of stories of people from adjacent professions who, over a couple of years, moved into AI—becoming Junior Data Scientists at first, then Seniors, and in 3–4 years managing small teams. “Lifelong learning” is not a slogan but the market’s reality. - 4. The return of the “cult of personality” and “geniuses”
With each year, standout individuals matter more—in both science and engineering. Facebook (Meta) is deliberately building its own culture around “AI stars”: freedom of research, the right to make mistakes, budgets for experiments, and most importantly—the ability to build teams of top specialists themselves.
GO TO FULL VERSION