Inside Meta’s AI power play: How Andrew Tulloch’s move from Thinking Machines Lab changes everything

Meta Platforms (NASDAQ: META) has pulled off one of the most closely watched hires of 2025, bringing in Andrew Tulloch, the co-founder of the fast-rising AI startup Thinking Machines Lab. The move, first reported by The Wall Street Journal, underscores how Meta is doubling down on its race toward artificial general intelligence and deep research dominance. For investors and analysts, it’s another sign that Meta’s transformation from a social media conglomerate to a full-stack AI infrastructure company is well underway.

The hire also highlights a broader trend across Silicon Valley — the return of “AI megadeals,” where talent itself becomes the most prized asset in trillion-dollar market rivalries. Tulloch’s switch from a startup valued at over $10 billion to a public company with more than $1.2 trillion in market capitalization marks a subtle yet powerful vote of confidence in platform-scale research environments.

Why did Meta make such an aggressive play for Thinking Machines Lab’s co-founder in 2025?

Meta’s pursuit of Andrew Tulloch didn’t happen overnight. The company, led by CEO Mark Zuckerberg, has spent much of 2025 building its AI division into a formal entity known as Meta Superintelligence Labs. The structure is modeled after Google DeepMind, integrating researchers, engineers, and infrastructure experts under one umbrella to accelerate model development.

According to people familiar with the matter, Meta’s leadership had identified Tulloch as a key figure capable of bridging the gap between academic research and product-grade AI systems. Tulloch’s background spans years at both Meta and OpenAI, where he worked on scaling reinforcement learning and fine-tuning systems. His return, therefore, brings institutional memory and a track record of high-impact innovation.

Reports suggest that Meta made a multi-year offer rumored to exceed $100 million in compensation, a figure that, while unconfirmed, signals just how far companies are now willing to go to win top minds in artificial intelligence. Meta’s logic appears clear: if it cannot acquire the startup, it will secure its DNA through its people.

For Zuckerberg, who has openly discussed his ambition to move Meta “beyond social platforms toward AI infrastructure leadership,” Tulloch’s arrival is part of a broader effort to elevate internal research capabilities. It also fits into Meta’s historical pattern — from the acquisition of Instagram and Oculus to large-scale in-house R&D — of absorbing expertise early to dominate the next technology cycle.

What makes Thinking Machines Lab so influential in the new AI research landscape?

Thinking Machines Lab emerged in early 2025 under the leadership of former OpenAI CTO Mira Murati, attracting talent from major AI houses like Anthropic, Meta AI, and Mistral. Despite having no commercial product at launch, the company raised a record-setting $2 billion seed round led by Andreessen Horowitz, placing its valuation between $10 billion and $12 billion.

Its mission was ambitious: to build customizable, multimodal AI models that balance power with interpretability. The lab’s first internal product, reportedly named “Tinker,” focuses on developer-facing APIs for model fine-tuning — a sector where startups compete to make AI infrastructure more accessible to enterprises and researchers.

Tulloch’s exit undoubtedly leaves a leadership vacuum. However, given the lab’s depth of technical talent and backing from major investors such as Nvidia and AMD, it remains well-positioned to continue operations. Still, his departure symbolizes the growing fragility of even well-funded AI startups when competing with tech giants that can offer billion-dollar retention packages.

How does this compare with previous AI talent wars among Meta, Google, and OpenAI?

The race for top AI researchers has long resembled an arms race. Google DeepMind once cornered the market on elite talent, but OpenAI’s rise — fueled by Microsoft’s deep pockets — shifted momentum in 2023 and 2024. Now, Meta is staging its comeback.

This isn’t the first time Zuckerberg’s company has pulled off a strategic coup. Back in 2013, Meta acquired the Toronto-based Deep Learning Group, led by Yann LeCun, now Meta’s Chief AI Scientist. That early bet laid the foundation for LLaMA — Meta’s flagship open-source large language model family. Hiring Tulloch represents an echo of that move, signaling renewed focus on frontier research rather than consumer-facing AI products.

The difference this time is that the stakes are far higher. Generative AI models now underpin trillion-dollar valuations, national policy discussions, and hardware roadmaps. Every leading AI company understands that top researchers represent leverage — not just intellectual, but geopolitical and financial.

How are markets and institutional investors reading this Meta hiring move?

Investor sentiment toward Meta Platforms has been increasingly driven by its AI narrative rather than advertising metrics. Institutional flows show that large funds have continued to accumulate META shares, banking on long-term upside from AI infrastructure and compute investments.

As of October 2025, Meta’s share price has climbed nearly 27 percent year-to-date, outperforming peers like Alphabet and Amazon in relative growth. Analysts point to capital expenditure forecasts nearing $72 billion for fiscal 2025 — much of which will support AI data-center expansion and model-training infrastructure.

In that context, hiring Tulloch is seen less as a cost and more as a strategic investment. The move signals to Wall Street that Meta intends to stay relevant at the very top of the AI pyramid. Analysts describe it as a “bullish credibility signal” that strengthens the long-term growth story.

However, there are risks. Institutional investors remain cautious about whether such high-profile hires will translate into measurable breakthroughs. If progress stalls, the market may interpret Meta’s spending spree as excessive. For now, the sentiment leans moderately bullish — with META retaining a consensus “Buy” rating across major brokerages.

What does Tulloch’s arrival mean for Meta’s internal AI roadmap and competition?

Tulloch joins Meta at a pivotal moment. The company is integrating its AI models into nearly every product line, from Instagram’s recommendation engine to generative ad-creation tools for advertisers. His technical focus reportedly includes reinforcement learning, model fine-tuning, and alignment — areas critical to developing next-generation reasoning systems.

The hiring also consolidates Meta’s human-capital strategy: scale the world’s best models internally while keeping them open-source to win trust and developer adoption. This approach contrasts sharply with OpenAI’s closed-weight model strategy.

Zuckerberg’s challenge now lies in ensuring that elite hires like Tulloch can thrive inside Meta’s complex structure. Integrating startup-style innovation into a company of Meta’s size can be difficult, particularly when compensation disparities or team hierarchies come into play. Yet if successfully managed, it could give Meta a decisive edge in both speed and credibility.

Can Thinking Machines Lab maintain its momentum without one of its co-founders?

Thinking Machines Lab’s investor backing provides strong financial resilience, but its ability to retain momentum depends heavily on sustaining morale and research clarity. The company’s other co-founder, Mira Murati, remains deeply involved, with reports suggesting she is now consolidating leadership and restructuring technical teams to adapt.

The firm’s next milestones — likely including its first model release and expanded developer API — will serve as critical indicators of post-Tulloch stability. Investors are also closely monitoring whether the startup can differentiate itself from the crowded field of well-capitalized AI challengers such as Anthropic, Mistral, and Cohere.

While some observers believe Tulloch’s exit could delay early product timelines, others argue it might accelerate decision-making by streamlining leadership. Either way, the startup’s long-term survival will hinge on how effectively it can retain its remaining founding engineers and maintain research independence in a consolidating industry.

What does this hiring reveal about the future of AI competition and consolidation?

Meta’s latest recruitment underscores the growing concentration of AI capability within a few massive players. Despite record startup funding, the gravitational pull of trillion-dollar corporations remains irresistible when compensation, compute, and influence align.

This consolidation raises important strategic questions: will innovation continue to thrive in open environments, or will AI development become dominated by a handful of mega-platforms? The answer likely lies somewhere in between. Startups like Thinking Machines Lab will continue to push creative boundaries, while giants like Meta will industrialize those discoveries at scale.

For researchers, the tension between independence and impact will only intensify. For investors, the message is clear — AI remains both a capital and talent game, and the two are increasingly inseparable.

In recruiting Andrew Tulloch, Meta Platforms has once again demonstrated its unmatched appetite for scale — not just in compute power or data, but in human potential. The move signals confidence, ambition, and perhaps a touch of inevitability: that Meta sees the next phase of artificial intelligence not as a sprint of models, but as a marathon of minds.

The coming year will reveal whether Tulloch’s addition propels Meta closer to its vision of superintelligence, or if this will become another case study in Silicon Valley’s billion-dollar talent churn. For now, one thing is certain — the world’s largest social-media-turned-AI-infrastructure company has no intention of sitting out the next great tech revolution.