Human defeated by Artificial Intelligence, from confidence to utter powerlessness
October 26, 2017 at 3:56 pm
(This post was last modified: October 26, 2017 at 4:00 pm by causal code.)
So even before AlphaGoZero, the world's strongest "Go" player, Deepmind had made something called AlphaGo.
AlphaGo utilized general learning algorithms, called neural networks, and it demolished Lee Sedol, one of the best humans back then.
In a funnily short period of time, his confidence before playing AlphaGo, slowly transformed into talk about feeling utterly "powerless" as he lost more and more games up until his eventual defeat.
![[Image: 24x3z3b.jpg]](https://images.weserv.nl/?url=i68.tinypic.com%2F24x3z3b.jpg)
This reminds me of people who tend to underestimate Artificial Intelligence, and I sometimes find myself being quite confident about some skill that humans do well, only to slowly face reality as that skill became overturned by artificial intelligence.
Interesting note: The latest version of AlphaGo, called AlphaGoZero, not only can easily beat Lee Sedol, and every other human, but also, AlphaGoZero played AlphaGo, winning 100 to 0.
This is a good reminder.
AlphaGo utilized general learning algorithms, called neural networks, and it demolished Lee Sedol, one of the best humans back then.
In a funnily short period of time, his confidence before playing AlphaGo, slowly transformed into talk about feeling utterly "powerless" as he lost more and more games up until his eventual defeat.
This reminds me of people who tend to underestimate Artificial Intelligence, and I sometimes find myself being quite confident about some skill that humans do well, only to slowly face reality as that skill became overturned by artificial intelligence.
Interesting note: The latest version of AlphaGo, called AlphaGoZero, not only can easily beat Lee Sedol, and every other human, but also, AlphaGoZero played AlphaGo, winning 100 to 0.
This is a good reminder.
