The Thinking Game
Documentary tracing DeepMind's journey from a London startup to defeating world champions at Go and StarCraft, revealing the team's pursuit of artificial general intelligence.
Key Takeaways
DeepMind started in 2010 when Demis Hassabis and Shane Legg—both obsessed with AGI—decided academia wouldn't support their ambition. They founded a company instead. Peter Thiel became their first major investor, though he wanted them in Silicon Valley. Hassabis insisted on London to avoid the startup culture of churning through companies annually.
The breakthrough came with Deep Q-Networks (DQN). By combining reinforcement learning with deep learning, they created a single algorithm that taught itself to play dozens of Atari games. The system received only pixels and a score—no rules, no strategy. It discovered optimal strategies humans had never found, like tunneling around walls in Breakout. This was the first real example of something resembling general intelligence.
AlphaGo's match against Lee Sedol in 2016 changed everything. Move 37 had a 1-in-10,000 chance of being played by a human, yet it worked. When the system later beat Ke Jie (the world's top player), China cut the broadcast mid-game. That moment became their "Sputnik moment"—launching an AI arms race.
AlphaZero stripped out all human knowledge. Starting each morning playing randomly, by dinner it became the strongest chess entity ever. The system discovered attacking styles never seen before, proving that millennia of human strategy had barely scratched the surface.
Games weren't the goal—they were the training ground. Hassabis sees AGI as the ultimate tool for solving humanity's most complex scientific problems. He believes it's bigger than the internet, bigger than mobile—more like the advent of electricity or fire.
Notable Quotes
"Trying to build AGI is the most exciting journey that humans have ever embarked on."
"If you could somehow plug in those 300 brains into a system, you might have solved cancer with that level of brain power."
"The human brain is the only existence proof we have—perhaps in the entire universe—that general intelligence is possible at all."
Connections
- ex-google-officer-speaks-out-on-the-dangers-of-ai-mo-gawdat - Mo Gawdat worked at Google X (DeepMind's parent) and shares concerns about autonomous AI learning
- andrej-karpathy-were-summoning-ghosts-not-building-animals - Discusses the 2013 Atari/RL work that launched DeepMind's approach