Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

For reference, AlphaGo is not globally optimal. Go is much too big for that.

(edit: also there'd be no point playing a game against it!)



I believe the OP refers to global/local as they are used in Go. There, global/local play is a concept taken for granted by humans, with explicit names like "sente" and "tenuki".

I think the OP simply implies AlphaGo doesn't seem bound by these human cognitive crutches.

This is not the same as "global" in the theoretical sense, as in "finding the perfect solution / exhaustive proof". See also "game temperature" [1]

[1] http://senseis.xmp.net/?Temperature


Correct. Google seems to agree: "AlphaGo has the ability to look “globally” across a board—and find solutions that humans either have been trained not to play or would not consider." [1]

[1] https://googleblog.blogspot.com.br/2016/03/what-we-learned-i...


What he's saying is that if AlphaGo played optimally, there'd be No point in playing.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: