The competition for AI talent in Silicon Valley has become fierce. Despite massive salary and stock option offers, many researchers are choosing loyalty to the leaders and organizations they believe in. This article covers in chronological detail the background behind Meta CEO Mark Zuckerberg's push to recruit OpenAI alumni, the choices of key individuals, and the culture and values of the AI industry.

1. The Billion-Dollar Offer, and the Rejection

In early 2025, Mark Zuckerberg approached OpenAI's former CTO Mira Murati in his push to get ahead in the generative AI race. He tried to acquire Murati's fledgling AI startup Thinking Machines Lab, but she firmly declined.

"Thank you for the billion-dollar job offer, Mark Zuckerberg. But I've decided to say no."

After Murati's rejection, Zuckerberg reached out to about 10 of Thinking Machines Lab's roughly 50 employees to gauge their interest in switching, launching a full-scale "talent war." A primary target was co-founder and AI research luminary Andrew Tulloch. Zuckerberg offered Tulloch an extraordinary compensation package worth up to $1.5 billion, but Tulloch also turned it down.

"None of his colleagues left the company."

Meta spokesperson Andy Stone dismissed the reported compensation package as "inaccurate and absurd," adding that "all compensation depends on stock price appreciation."

2. Background of the AI Talent War: Loyalty and Legendary Leaders

While rejecting offers worth hundreds of millions or even billions of dollars is rare in Silicon Valley, as the AI talent battle has intensified, money alone is no longer enough to attract talent. Some AI researchers do move for higher pay and more authority, but many demonstrate loyalty to the leaders and organizational culture they believe in.

"These leaders are legends of the tech world, whose names carry the appeal of rock stars."

In particular, OpenAI and companies founded by its alumni rally researchers with the historic mission of "AGI (Artificial General Intelligence)." OpenAI co-founders Sam Altman, Elon Musk, Ilya Sutskever, and others have championed this mission with an almost religious fervor.

3. Meta's Aggressive Recruiting and Companies' Defense Strategies

Meta approached more than 100 OpenAI alumni and succeeded in recruiting at least 10. On July 25, 2025, Zuckerberg appointed Chinese researcher Shengjia Zhao, an OpenAI alumnus, as head of Meta's Superintelligence team.

However, many OpenAI researchers turned down Meta's offers. Their reasons included:

• Belief that OpenAI is the closest company to achieving AGI
• Desire to work in smaller organizations
• Not wanting their research used solely for ad-based products

Anthropic is also defending against Meta's recruiting offensive. Founded by OpenAI alumnus Dario Amodei, all seven of Anthropic's co-founders remain at the company. United through the "effective altruism" social movement, they prioritize the risks of AI and responsibility to humanity.

"All seven of Anthropic's co-founders are still at the company."

Meta managed to recruit only two people from Anthropic.

4. Startup Culture and Leadership: The Cases of Murati and Sutskever

Mira Murati spent six years at OpenAI, managing nearly every aspect of the company as CTO. She earned researchers' trust through high emotional intelligence and a flat leadership style. Thinking Machines Lab, which she founded, similarly pursues a flat organizational culture, with all researchers' titles unified as "technical staff member."

"Even senior researchers only use the title 'technical staff member.' This is modeled after Bell Labs' egalitarian and collaborative culture."

More than 20 OpenAI alumni have joined Thinking Machines Lab, including John Schulman, a key figure behind ChatGPT.

Meanwhile, Ilya Sutskever's Safe Superintelligence (SSI) discourages employees from listing the company name on LinkedIn and minimizes external exposure to prevent talent poaching. Sutskever also rejected Zuckerberg's offer to acquire SSI.

5. The Choices of Talent and Changes in Silicon Valley

Andrew Tulloch studied at the University of Sydney and Cambridge before working on Facebook's (now Meta) AI research team. In 2016, he received a recruitment offer from OpenAI but declined due to the salary gap. Seven years later, after ChatGPT became a massive hit and OpenAI's value skyrocketed, he ultimately joined Thinking Machines Lab.

"Andrew is truly a genius among geniuses." -- Former Facebook executive Mike Vernell

In this way, Silicon Valley's AI talent is increasingly prioritizing mission, leadership, and organizational culture over simple financial compensation.

6. The Future of AI Startups and Investors' Perspective

Thinking Machines Lab recently raised $2 billion in funding with the goal of "making AI systems more understandable, customizable, and powerful." Murati plans to unveil the first product soon, though the specifics remain secret even from investors.

"We're building multimodal AI that matches the way you interact with nature and the world. We'll reveal our first product within the next few months."

Anthropic is also receiving significant investor attention, with assessments that it has entered an AI growth "flywheel."

7. Competition and New Opportunities in Silicon Valley

Unlike the PayPal Mafia, OpenAI alumni are founding new startups or joining competitors, creating a unique ecosystem where competition and collaboration coexist.

"Unlike the PayPal Mafia, OpenAI alumni keep competing with each other."

There are also voices emphasizing that jumping quickly into new platforms and opportunities when they emerge can be life-changing, even for ordinary people.

Conclusion

The Silicon Valley AI talent war has moved beyond simple salary competition. Loyalty, leadership, organizational culture, and a sense of mission toward humanity have emerged as crucial variables. How talent movement and startup innovation in the AI industry will shape the future remains well worth watching.