Re: [computer-go] Modern brute force search in go

[Don Dailey <drd@xxxxxxx>]

> >Just my thoughts.  All of this  is even more true of Go.  There exists
> >(in  principle) a  future GO  player than  can make  the very  best Go
> >player (of today or yesterday) look like a baby.
> >
> >- Don

> To be honest, I consider all of this pure speculation again.   

Yes, it was intended as  speculation.  What's wrong with that?  I have
looked at several of your last posts and they are full of opinions and
speculations too.   One of your speculations, was  that go programmers
are much nicer people than  chess programmers (which was an aggressive
attack) and  that Frank  had some ego  problems, which seemed  like an
aggressive attack to me.  Which one  do you consider yourself to be, a
nice go programmer or an evil chess programmer?

>                                                                   Top pro's
> estimate 'god's level' to be anywhere between 3 and 5 stones stronger than a
> top pro. I do think top pro's are biased in this respect, but they're not
> totally ignorant. And they're definitely not stupid either, so I'd say they
> are better qualified to make such estimates than anyone else. Until we have
> some hard data or good theories that say otherwise, why make such claims?

I am  always extremely  leery of  the advice that  comes in  the form,
"don't bother to reason about it, trust the experts who know more than
you and therefore must be right."

I trust  experts who know more  than myself more than  I trust myself,
but I don't automatically go into sheep mode either.  I also don't buy
into that this is their  domain arugment.  Great go players play great
GO, but as you admit, they tend to be biased.  I have rarely, if ever,
met an  unbiased chess  player, maybe GO  players are  different.  But
most chess players have inflated notions  of how close they are to the
top.

In the early  computer chess days, virtually every  human chess master
believed that there was a  conceptual barrier beyond which it would be
impossible for  chess computers to  break through.  The  "barrier" was
"speculated" to  be around  the 2000 ELO  rating level at  some point,
which is ridiculously low.

If the  assertion that computers  would someday play much  better than
that had  been made  back then, they  would have been  quickly slapped
down with  the argument, "shut up  because the experts  know more than
you."

The problem is that those masters were very good at playing chess, but
they didn't know ANYTHING about the deep nature of their own game, the
workings of the human brain, or computers.  When those statements were
made, I  knew right away they  were talking about  something they were
completely ignorant  about and  I trusted my  own intuition  more than
theirs, even though they  were the supposed experts.  Eventually their
point of view was proven to be foolishness.

Make anyone live inside a deep box their whole universe becomes a box.
The top  players are  used to being  in a  box sitting high,  but that
doesn't mean they know about things outside the box.  We can speculate
on this just as easily as they can.

Someone said this:

> It's not pure speculation. Youcan  try to determine god's rankd by looking
> at winning probabilities in even games between players of different
> strength -- at god's level, even a 0.1 rank difference amounts
> to a winning percentage of 100%. I think someone did this with the data
> in the EGF ranks, and concluded it would be about 2-3 stones above
> today's top pros.

I don't  know who  Youcan is,  but it's clear  he is  a mathematician.
Perhaps  he is  a  GO player  too?  But  in  either case,  I view  the
mathematician to be in a far better position to speculate on the basic
question, "how  high is high."  I  would never trust  anyone who says,
you shouldn't even think about the question.

- Don


_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://www.computer-go.org/mailman/listinfo/computer-go/