DeepMind’s Journey from Games to Fundamental Science
Demis Hassabis, CEO and co-founder of DeepMind, dreams of using AI to solve fundamental problems in science. He joins Azeem Azhar to explore his own journey from world champion gamer to neuroscientist to building AI systems that can train themselves to solve real-world engineering challenges and, eventually, make Nobel-prize winning discoveries.
They also discuss:
Why games are the perfect training ground for AI algorithms. How DeepMind combines the dynamism of a startup with the “blue sky” creativity of a lab. Why “deep reinforcement learning” is the basis for artificial intelligence systems.Further reading:
“How GPT-3 Is Shaping Our AI Future” (Exponential View podcast, 2020) “Inside DeepMind’s epic mission to solve science’s trickiest problem” (Wired, 2019) “Artificial general intelligence: Are we close, and does it even make sense to try?” (MIT Technology Review, 2020)