Sir Demis Hassabis may be the most powerful CEO alive. Nobel Prize winner at 48, he now runs Google's DeepMind AI division. His methodical path to AI is as brilliant as it is unsettling: 🧵
I've been watching AI development for years. Nobody worries me more than Demis Hassabis. Not because he’s reckless... But because he’s brilliant, patient, and building something that could outthink us all:
Most people think Hassabis got lucky with DeepMind. They're wrong. He's been preparing for this moment since childhood. By 13, he reached chess master level with a 2300 rating. But chess was just training for something bigger...
At 17, Hassabis co-designed Theme Park, a game that sold several million copies. But he wasn't interested in entertainment. He was testing whether AI could simulate complex systems. Each project was building toward his ultimate goal...
Hassabis founded Elixir Studios to push AI further. His game Republic: The Revolution featured AI simulating an entire country's politics. The scope was so ambitious that it caused delays. Hassabis was just getting started:
Then he made a move nobody expected. He paused his career to get a PhD in neuroscience. He wanted to understand how the human brain creates intelligence. This would become the foundation for everything that followed:
In 2010, Hassabis combined everything he'd learned. Chess strategy. Gaming simulation. Neuroscience insights. DeepMind was born with one mission: to solve intelligence, then use it to solve everything else...
2016: AlphaGo beat world champion Lee Sedol 4-1. Go was considered impossible for AI. The game had been resistant to existing programming techniques. Hassabis's AI mastered it through pure learning. This proved his approach worked:
2020: AlphaFold solved protein folding. This 50-year scientific challenge stumped entire generations of researchers. Hassabis's AI cracked it. Result? 2024 Nobel Prize in Chemistry. But his most dangerous project is still ahead...
Hassabis believes artificial general intelligence is coming within the decade. True AGI that reasons, creates, and solves problems like humans. He calls it "one of the most beneficial technologies of mankind ever." But there's a dark side...
Hassabis also signed the AI extinction warning. He knows what he's building could destroy humanity. "We need urgent research on how capable and controllable new AI models are," he says. He's racing to build safe AGI before others build dangerous versions.
Video/ Image credits: • Google DeepMind CEO Demis Hassabis: The Path To AGI, Deceptive AIs, Building a Virtual Cell https://www.youtube.com/watch?v=yr0GiSgUvPU
