Why Demis Hassabis Ignored Elon Musk and Sold DeepMind to Google for $650 Million Instead

The origin story of Google buying DeepMind is worth understanding if you own Alphabet Inc. (NASDAQ:GOOGL | GOOGL Price Prediction) or if you are following AI whatsoever. (All of us?)

The Funding Crisis That Changed Everything

In January 2014, Demis Hassabis was facing a wall. VC backers had reneged on funding commitments to DeepMind, leaving the lab without the capital it needed to pursue what Hassabis described as his core ambition: “get a shitload of computers and solve intelligence.” That mission statement does not fit neatly into a venture fund’s five-year return window.

This is the overlooked part. Blue sky research and venture capital are fundamentally misaligned. VC funds need exits. AGI research needs decades and infrastructure that costs billions. When the money dried up, Hassabis had to find a different kind of partner.

Larry Page’s Pitch and Elon Musk’s Warning

Larry Page pitched Demis on selling to Google rather than going independent. The offer was approximately $650 million. Google had the compute, the patience, and the distribution that no VC could match.

Elon Musk made a competing pitch, warning Hassabis about the risks of corporate control over AGI and offering an alternative path through Tesla and SpaceX. Hassabis declined. His reasoning came down to resources and mission fit. Google could actually fund the research at the scale required. Musk’s alternative could not guarantee the same compute access.

What That $650 Million Bought

Twelve years later, the numbers tell the story. GOOGL has risen roughly 1,060% since January 2014. The DeepMind team is now central to Alphabet’s AI strategy, responsible for the Gemini model family.

In Q4 2025, Alphabet posted $113.83 billion in revenue, rising 18% year over year. CEO Sundar Pichai said: “The launch of Gemini 3 was a major milestone and we have great momentum. Our first party models, like Gemini, now process over 10 billion tokens per minute via direct API use by our customers, and the Gemini App has grown to over 750 million monthly active users.”

Google Cloud revenue reached $17.66 billion in Q4 2025, rising 48% year over year. Alphabet is guiding for $175 to $185 billion in capital expenditures in 2026, almost entirely aimed at AI infrastructure. That is the patient capital Hassabis needed in 2014 and could only find at Google.

When a mission requires massive compute and a multi-decade time horizon, strategic acquirers with existing infrastructure are often the only viable path. Alphabet made that bet for $650 million. The Gemini models processing over 10 billion tokens per minute are the return on that investment. If you believe AI infrastructure spending translates into durable revenue growth, the DeepMind origin story is exactly the kind of foundational advantage worth keeping in mind.