Re: computer-go: Two ways to program a GO-engine

["Charles Matthews" <charles@xxxxxxxxxxxxxxxxxx>]

From: Nick Wedd
>Do we know that?  How do we know that?
>
>Suppose no-one had come up with the trick for winning at Nim.  You can
>imagine Nim programs that use tree searches, heuristics, detailed end-
>game analysis.  Then one day someone claps their hands and comes up with
>the trick about expressing the numbers in binary etc.
>
>I accept that something like this is improbable in Go.  But I don't know
>that it won't happen.


This is one occasion when it is useful to call on the theoretical result,
that Go is NP complete.  Therefore there cannot be this sort of hidden
secret, valid on boards of all sizes, depending only on computation linear
in the board size.

Charles