AlphaGo and what was not understood, I think.

I saw the Netflix documentary on AlphaGo, an AI/Neural Network program to play Go. I can highly recommend it, even though the events tool place in 2016. It is a very human story. 

The long and the short of it is that, as usual, experts in a certain domain believe that a computer will never be able to do what they do. And then AlphaGo beat the world champion of Go, Lee Sedol , 4 games to 1. 




Lee Sedol seems to me to be a nice bloke who fought an heroic fight against the machine.  But what struck me was what he said at the end. That AlphaGo had made him rethink his way of playing, he would learn from the machine. But I think that is impossible, AlphaGo's "reasoning" is nothing like ours. Which is why AI / Neural Networks can uncover hidden patterns and strategies. Because they do not think like humans. Our experience sometimes limits us. Well designed neural networks have no such limits.

One of the Go commentators said during the documentary that move 37 by Alpha Go was creative and beautiful. What better review a creative being ask for?

PS: I think it is harder to build a robot which does your plumbing than a program which beats you at chess or Go. Plumbers will not be out of work any day soon, honest financial advisers and other intellectual workers will be. 

PPS: Dishonest financial advisers will always flourish.


PPPS: I've just found a statement in Clever Algorithms, Nature Inspired Programming Recipes by Jason Brownlee  which seems very relevant to this blog post. Talking about the application areas of AI in general one is "The person solving the problem is inadequately prepared or imagines some psychological barrier that prevents them from discovering a solution." In other words AI discovers things we are too ignorant or too psychologically blocked to find.





Comments

Popular posts from this blog

Lithophanes, I can only learn by experience.

Go! Stop! Stroop!

How I don't learn from experience