[ExI] Google's Go Victory Is Just a Glimpse of How Powerful AI Will Be

spike spike66 at att.net
Tue Feb 2 17:26:37 UTC 2016

-----Original Message-----
From: extropy-chat [mailto:extropy-chat-bounces at lists.extropy.org] On Behalf
Of Robin D Hanson
Subject: Re: [ExI] Google's Go Victory Is Just a Glimpse of How Powerful AI
Will Be

Alejandro Dubrovsky <alito at organicrobot.com> wrote: 
>> The games were played back in October, and insiders would have known that
it was doing remarkably well even before that...

The AI hipsters may educate me on this please.  We in the chess world
watched in amazement as computers overtook humanity.  Club players know the
typical trajectory of a hardcore player: some advance quickly, then level
off, some keep going and level off, but computers advanced far more steadily
over the years than humans do.  

We who were watching that in the 90s kept assuming the software would level
off indefinitely at about the low end of the grandmaster level, where
arbitrary additional computing power just couldn't overcome human intuition.
Well, we were wrong.  Computer strength kept climbing right through
grandmaster and is still climbing.  Even those who watched the whole process
from back in the 1970s were astonished and were consistently wrong in our
predictions.  I blew it over and over.

Our appalling conclusion: chess does not require intelligence.

>...(And why do my posts keep taking a day to appear here?)

Robin Hanson rhanson at gmu.edu

I have heard this before from some servers, and we have never discovered
why.  Robin does it always take that long, or is it inconsistent?

Sometimes my posts are delayed by a few hours but seldom or never a full
day.  I don't see anything funky with your account.



More information about the extropy-chat mailing list