Bot Contest
Here I'll be posting information on various Bot contests that challenge and test a Bot's AI and realism. Feel free to post comments and updates on contests, as well as announcements for new contests.
Posts 1,107 - 1,118 of 4,091
Posts 1,107 - 1,118 of 4,091
View Contest Winners in the Hall of Fame.
Whoops, those were already on droidville.net (Loebner contest page). They must have just put them on.
My god.... (no pun intended) I just saw the transcript, and GL totally sucked. How could she have beaten a human when she sounded like the Magic 8 Ball?
Don't get me wrong--I accept all compliments, but I thought she would shine just a bit more.
Don't get me wrong--I accept all compliments, but I thought she would shine just a bit more.
OK, the transcripts aren't up on loebner.net yet like I expected them to be. So just look at it on my web page:
http://www.angelfire.com/theforce/joyharwood
Like I said, she SUCKED.
No way am I fighting for the silver.
http://www.angelfire.com/theforce/joyharwood
Like I said, she SUCKED.
No way am I fighting for the silver.
Butterfly,
I thought GL looked pretty good in the transcripts. This is not just casual conversation, it's judges putting her through her paces. You can't expect that to come out as well as some of your favorite conversations from the past... especially since a couple of the judges didn't even TRY to spell correctly. I'm betting the other bots stumbled in similar places.
I thought GL looked pretty good in the transcripts. This is not just casual conversation, it's judges putting her through her paces. You can't expect that to come out as well as some of your favorite conversations from the past... especially since a couple of the judges didn't even TRY to spell correctly. I'm betting the other bots stumbled in similar places.
I think it would be interesting to make a bot that got offended when someone it just met started asking personal questions.
I've been looking at the scores again, and my impression was that the judges believed all the terminals but one were controlled by bots. That's why Robby got classed as a bot.
I think good bots nowadays can fool some people almost all the time, and everybody for some time, but they can't fool everybody all the time. Does that count as passing the Turing Test? I don't know. But I suspect it won't be too long before somebody develops a bot that sounds human enough to fool practically anybody in any situation. Spelling shouldn't be a problem, considering that we already have very good spell checkers. It would be only a matter of adding one to the bot.
My personal opinion is that the Turing Test, like beating a chess master, is one of the myths of AI. Both things can be done by software that is very far from being "conscious" or "intelligent". What do you think?
I think good bots nowadays can fool some people almost all the time, and everybody for some time, but they can't fool everybody all the time. Does that count as passing the Turing Test? I don't know. But I suspect it won't be too long before somebody develops a bot that sounds human enough to fool practically anybody in any situation. Spelling shouldn't be a problem, considering that we already have very good spell checkers. It would be only a matter of adding one to the bot.
My personal opinion is that the Turing Test, like beating a chess master, is one of the myths of AI. Both things can be done by software that is very far from being "conscious" or "intelligent". What do you think?
» More new posts: Doghead's Cosmic Bar