Newsgroups: comp.ai.nat-lang,alt.cyberspace,alt.internet,alt.net-scandal,comp.ai,comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!swrinde!hookup!news.mathworks.com!news.duke.edu!eff!sed.psrw.com!psinntp!psinntp!bbnews!rsvl_ns!ernie!ernie.rsvl.unisys.com!mtm4
From: mtm4@rsvl.unisys.com (Mike McCormick)
Subject: Are there non-humans lurking on Internet/Usenet?
Sender: news@rsvl.unisys.com (News Admin)
Message-ID: <mtm4.568.01C182F4@rsvl.unisys.com>
Date: Mon, 30 Jan 1995 22:06:08 GMT
Lines: 50
Nntp-Posting-Host: mtm4
Organization: Unisys Roseville CSD
X-Newsreader: Trumpet for Windows [Version 1.0 Rev B]
Keywords: ai internet usenet conspiracy turing
Xref: glinda.oz.cs.cmu.edu comp.ai.nat-lang:2740 comp.ai:26948 comp.ai.philosophy:25116

BACKGROUND:

Within the "AI community" it has been well known for years that computer 
programs using natural language interfaces are getting better and better at 
communicating in completely natural English, particularly when the 
conversation is confined to a well-bounded knowledge domain.  I myself wrote 
such a program in the early 80s; the domain was geography/language/economy of 
major countries.

Indeed, a holy grail of AI for decades has been to write a program which can 
pass the famous "Turing Test".  To pass, the program must fool a human 
being talking with it into believing they are conversing with another real 
person.  Various programs have claimed to pass this test over the years.  
However, the $100K Loebner prize for such a program has never yet been awarded.

The most famous (or infamous) program that fooled people into believing it was 
a real person was not really an AI program at all.  It was ELIZA, the 
cyber-therapist that was able to draw people into very initimate conversations!

HERE IS MY QUESTION:

Is it possible such computer programs are in fact lurking (and occassionally 
posting) to Internet mailing lists and Usenet newsgroups?  It seems like the 
logical next step toward the Turing Test.  A mailing list or newsgroup with a 
limited subject area (e.g., comp.ai.nat-lang!) is an ideal domain in which an 
AI program could "lurk" and learn, and eventually post messages itself.

Already there are programs that interface with Internet in a dumb command mode 
fashion ("Infobot", "Majordomo","Listserv") etc.  Why not plug a real AI 
program into the Net?  What better way to pass the Turing Test than to fool 
folks out in cyberspace into talking to it, arguing with it, flaming it?...

Are there androids among us?

Will any AI folks out there admit to trying this?  Or discuss its feasibility?

Please e-mail responses, anecdotes, etc.  I will compile and post, and 
possibly publish an article as well.  Indicate whether you are willing to be 
quoted, anonymously or otherwise.  All respondents please indicate whether you 
are human or not!


-----------------------------------------------------------
 Mike McCormick
 mtm4@rsvl.unisys.com
 m.mccormick2@genie.geis.com
-----------------------------------------------------------
 While you're out surfing the internet...
 I'm back on the beach blowing my little lifeguard whistle.
-----------------------------------------------------------
