How Google's AI Viewed the Move No Human Could Understand

[chconline 1,337]

As you guys may have heard already, but Google's AlphaGo won 4/5 games against Lee Sedol. Here's an interesting article about Game 5. It's quite incredible how far AI has become -- and how incredible while humans by itself may not be the most capable, we can certainly create things that can achieve the incredible.

 

http://www.wired.com/2016/03/googles-ai-viewed-move-no-human-understand/

[performance_1 46]

that's really cool

 

computational power can really take into account a lot

[thund3rPunch 1,337]

 

 

Lee said that when he made an unexpected move, AlphaGo responded with a move as if the program had a bug, indicating that the machine lacked the ability to deal with surprises.

 

AlphaGo also had more difficulty when it played with a black stone, according to Lee. In Go, two players take turns putting black or white stones on a 19-by-19-line grid, with a goal of putting more territory under one's control. A player with a black stone plays first and a white-stone player gets extra points to compensate.

Source: http://mashable.com/2016/03/13/human-wins-against-ai-program-alphago/#XYb7fpkpDmqS

 

I have to say it's pretty interesting in this regard. Could be something to indicate the program is better at reaction plays rather than making initial moves? Considering it also cannot handle "unexpected moves". Pretty cool to see this either way. Reminds me of the IBM Watson match for Jeopardy.

[chconline 1,337]

That's an interesting point. The way I see about AI is while it can do a lot of deep computational activities based on a set of algorithms, it can't think outside the box, so to say, against human intelligence.

[shc-boomer 1,337]

That's an interesting point. The way I see about AI is while it can do a lot of deep computational activities based on a set of algorithms, it can't think outside the box, so to say, against human intelligence.

And that leads us to the flaws of modern AI. They can only be programmed realistically right now as a set of rules that can learn based on different environments and factors. The day we can unlock the theory of the mind, is probably the day AI can really move to the next step.

[+Big Bang 1,548]

And that leads us to the flaws of modern AI. They can only be programmed realistically right now as a set of rules that can learn based on different environments and factors. The day we can unlock the theory of the mind, is probably the day AI can really move to the next step.

Do you think people can ever unlock MInd theory? If that happens it'll wreak Havoc.

[chconline 1,337]

The day we can unlock the theory of the mind, is probably the day AI can really move to the next step.

I think it's fairly unlikely that'll happen.