home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.ai.philosophy
- Path: sparky!uunet!tcsi.com!iat.holonet.net!ken
- From: ken@iat.holonet.net (Ken Easlon)
- Subject: Re: Drawing the entity/environment boundary
- Message-ID: <C0BCnA.1A7@iat.holonet.net>
- Organization: HoloNet National Internet Access BBS: 510-704-1058/modem
- References: <C052B5.95t@spss.com>
- Date: Mon, 4 Jan 1993 05:01:55 GMT
- Lines: 59
-
-
- In article <C052B5.95t@spss.com> ,
- markrose@spss.com (Mark Rosenfelder) writes:
-
- >Neither does Searle, and that's precisely the problem. Searle never seems
- >to question the notion that you can determine whether understanding occurs
- >in the Chinese Room by asking if the man in the room understands Chinese.
-
- I think the Turing test, and the whole Chinese room argument WITH responses
- is a crock.
-
- All who agree say aye.
-
- If you take a Chinese room or the winner of a Turing contest and subject it
- to the real world, the idea simply breaks down. Could either one "pass" an
- employment interview, or pick up girls at a bar.
-
- My point here, is that we humans are constantly subjected to all kinds of
- tests, and as likely or not we fail. How is any machine using only symbols
- and syntax going to beat us at our own game.
-
- My opinion is that people who seriously think a word mongering machine can
- duplicate any but the most superficial activities of the human mind
- probably haven't taken a close look at the working of their own mind.
-
- >The corollary of his position would be that if a Turing Test-passing AI
- >*were* intelligent, the intelligence must lie in the *CPU*. The Systems
- >Reply more sensibly draws the entity boundary around the
- >computer/algorithm combination (and emphasizes the need to think more
- >carefully about what systems exhibit what cognitive properties).
-
- Did I read this right? Are you saying there are people in responsible
- positions that think a computer sans algorithm is intelligent? I probably
- read it wrong.
-
- I think you're still leaving out a vital component, the computer/algorithm
- combination should read computer/algorithm/programmer combination. At the
- current state of technology, a programmed computer is simply an extension
- of the programmer (admittedly a faster running, better remembering
- extension).
-
- If computer technology ever gets to the point where Bill Gates can fire all
- his programmers, then I might concede that computer/algorithm possesses
- some degree of cognitive ability. That won't be the point however, where
- I'll agree that computers can pass for human.
-
- Am I saying that programming a consciousness emulating computer is
- impossible? Not at all, but I do think any attempt to do so without taking
- a good long subjective look at the human mind is bound to result in
- failure.
-
- That was refreshing, I feel much better now.
-
- --
- Ken Easlon | Hello! I'm a .SIC Virus. Copy me and spread
- ken@holonet.net | the fum.
- Pleasantly Unaffiliated |
-
-
-