in reply to Re: Artificial Intelligence experiment
in thread Artificial Intelligence experiment
Thanks! The article about ELIZA is hilarious, in that this AI "doctor" was also based on the principle of the superficiality of communication, and how many people were fooled, feeling like they received real help from an apparent psychotherapist.
My program is much simpler than ELIZA, just a hash, really, and a CGI interface.
It could probably be almost a one liner if I was a more experienced programmer. The trickiest part was just getting around Perl's taint checking, and some additional security measures on the server.
In a way though, Bot 2 is more "advanced" in that, according to that article. ELIZA could only be updated by manually rewriting the code. I'm intrigued by the possible potential of a program that can be updated or "learn" from people connecting to it and modifying it from all over the world simultaneously. My biggest wory was that with a potential for thousands of people to be able to "teach" Bot, simultaneously, the hash would grow extremely large very quickly. As it is, that did not happen. A million people could say "Hello Bot", and he says hello. They say "how are you" he says fine. The hash is not modified. People's Repertoire of greetings and casual conversation are surprisingly (to me anyway) limited. The problem has been more, trying to get people to converse long enough to find something not already in the hash.
If people could add alternate responses, that might be more interesting. -- Tom
In a way though, Bot 2 is more "advanced" in that, according to that article. ELIZA could only be updated by manually rewriting the code. I'm intrigued by the possible potential of a program that can be updated or "learn" from people connecting to it and modifying it from all over the world simultaneously. My biggest wory was that with a potential for thousands of people to be able to "teach" Bot, simultaneously, the hash would grow extremely large very quickly. As it is, that did not happen. A million people could say "Hello Bot", and he says hello. They say "how are you" he says fine. The hash is not modified. People's Repertoire of greetings and casual conversation are surprisingly (to me anyway) limited. The problem has been more, trying to get people to converse long enough to find something not already in the hash.
If people could add alternate responses, that might be more interesting. -- Tom
|
---|
Replies are listed 'Best First'. | |
---|---|
Re^3: Artificial Intelligence experiment
by PerlGuy(Tom) (Acolyte) on Mar 05, 2020 at 10:47 UTC | |
by marto (Cardinal) on Mar 05, 2020 at 10:51 UTC | |
by PerlGuy(Tom) (Acolyte) on Mar 05, 2020 at 12:59 UTC | |
by marto (Cardinal) on Mar 05, 2020 at 13:03 UTC | |
by Anonymous Monk on Mar 05, 2020 at 15:10 UTC | |
| |
by PerlGuy(Tom) (Acolyte) on Mar 06, 2020 at 01:22 UTC | |
| |
by PerlGuy(Tom) (Acolyte) on Mar 05, 2020 at 23:44 UTC | |
by marto (Cardinal) on Mar 06, 2020 at 03:19 UTC | |
| |
by karlgoethebier (Abbot) on Mar 06, 2020 at 17:37 UTC |
In Section
Meditations