Archive of UserLand's first discussion group, started October 5, 1998.
Computer Folklore
Author: Paul Snively Posted: 4/7/2000; 11:53:14 AM Topic: scriptingNews outline for 4/7/2000 Msg #: 16022 (In response to 16008) Prev/Next: 16021 / 16023
David Winer wrote:>It was thought, at one time, that computers could gain consciousness,
>and perhaps have rights and vote and otherwise be human.It's far from apparent, at this moment in time, that computers cannot gain consciousness, have rights, vote, and otherwise become, not human, but sentient. The argument is still going on between those who take what is known as the "Strong AI" position, which contends that the above is possible, and the "Weak AI" position, which contends that the above is not.
The strongest argument in favor of Strong AI that I'm aware of is found in Philosophy Professor John Pollock's <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0262161133>. The specifics of the theory put forth in the above title can be found in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0262161524>.
The code that implements the above-documented framework (!) can be found at <http://www.u.arizona.edu/~pollock/>
The best 50,000-foot view from a Strong AI adherent whom I had the great honor and pleasure of meeting while at Indiana University, Professor Douglas Hofstadter, is found in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0465026567>.
A strong critique of Strong AI from noted physicist Roger Penrose can be found in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0140145346>.
Critiques of Penrose's critique of Strong AI by Hofstadter and philosopher Daniel Dennett can be found in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0553345842> and by physicist Frank Tipler in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0385467990>.
Penrose's critique of the Hofstadter, Dennett, and Tipler critiques of Penrose... can be found in <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0195106466>.
A different approach to Strong AI is found at <http://www.cyc.com/>.
Several Star Trek: the Next Generation episodes dealt intelligently with the issue of Data's rights as a sentient being. I'm thinking in particular of the episode in which Data becomes (albeit briefly) a father. Also, the son of the late singer Mel Torme wrote an intelligent script in which the barrier between the Enterprise proper and the holodeck is breached in such a way that a "character" in the holodeck becomes aware that he is a... character. He plaintively asks Picard what will happen to him once Picard leaves the holodeck: will he disappear, will he go home to his wife and children, what? Picard gives him the only possible answer: "I don't know." I personally found this episode very moving; it asked all the right questions and steadfastly refused to give the pat, easy answers.
Bottom line: don't be so quick to dismiss "computer consciousness" as "computer folklore." Sixty years of computer science isn't nearly enough to have come anywhere near settling the issue.
>In fact,
>it takes conscious beings, programmers, to make computers do
>anything useful.Not true since at least the 1960s, when Samuels wrote the first piece of software (of record, anyway) that learned from experience until it could play a better game of checkers than its author. Machine learning has made great strides since then, and computers now routinely do things that require that they have mastered certain abilities in excess of their creators' abilities. The best general resource on "intelligent agent" technology is <http://www1.fatbrain.com/asp/bookinfo/bookinfo.asp?theisbn=0131038052>.
This page was archived on 6/13/2001; 4:54:42 PM.
© Copyright 1998-2001 UserLand Software, Inc.