Although this blog is about the blockchain and its “cryptoassets” this post is about what happened last week in the game of Go.
The Chinese game of Go is a way that Google’s AI team from DeepMind is testing the edges of what’s possible.
There are many great pieces written about Google’s latest incarnation of AI called AlphaGo Zero, like this one in the New Scientist, but I don’t think anything captures what happened like the reactions of human Go masters who were soundly defeated by the AI.
What distinguishes AlphaGo Zero from AlphaGo is it trained from scratch, playing against itself, rather than learning in part based on watching human games.
As the New Scientist piece above explains: “Three days and 4.9 million such games later, the result is the world’s best Go-playing AI.”
After losing to AlphaGo Zero this is what Ke Jie, the world’s reigning Go champion, posted:
After humanity spent thousands of years improving our tactics, computers tell us that humans are completely wrong. I would go as far as to say not a single human has touched the edge of the truth of Go.
And, separately, this is what Gu Li, a professional Go player from China, posted:
AlphaGo has completely subverted the control and judgment of us Go players. I can’t help but ask, one day many years later, when you find your previous awareness, cognition and choices are all wrong, will you keep going along the wrong path or reject yourself?
Go is at least 2,500 years old so for the reigning Go champion to say “not a single human has touched the edge of the truth of Go” is flat out remarkable.
And for a highly skilled and respected player like Gu Li to feel that everything he knows about the game is “all wrong” and question how to even continue is equally so.
Playing AlphaGo Zero clearly seems to have pushed these two Go experts into the most existential state of reflection possible.
Why am I writing about this on a blog about the blockchain? Because as goes AI so goes the blockchain.