Newsgroups: comp.ai
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!portc02.blue.aol.com!howland.erols.net!news-peer.gsl.net!news.gsl.net!ix.netcom.com!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: tiring of mentifex
Message-ID: <nagleE2zKFo.FGx@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <19961225155800.KAA01730@ladder01.news.aol.com> <nau-ya02408000R2512961239200001@news.wam.umd.edu>
Date: Wed, 25 Dec 1996 20:18:59 GMT
Lines: 25
Sender: nagle@netcom6.netcom.com

nau@cs.umd.eduu (Dana S. Nau) writes:
>In article <19961225155800.KAA01730@ladder01.news.aol.com>, i3uddha@aol.com
>(I3uddha) wrote:
>> ... Becoming bored with your manifestoes. ...

>Unfortunately, comp.ai often seems to attract people with offbeat ideas
>that most serious AI researchers would consider implausible.  Because of
>this, the signal-to-noise ratio in comp.ai has never been very high; and
>during the last year or two, it has decreased to nearly zero.

    However, over on comp.ai.games, the content is much better.
Games have become sophisicated enough simulated worlds that they need 
good behavior-based AI to drive the non-player characters.  Progress
there is verifiable by others; you can get the game and see how well
the computer-run characters behave.  

    Game AI is almost entirely non-linguistic, which is good; top-down
AI, as a field, is sort of stuck, so we may as well push bottom-up
AI upwards for a while.

    I'm currently working at the physics and control level, as I have
been for the last few years, but I look forward to getting back into
behavior in 1997.  See "www.animats.com".

					John Nagle
