Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
OK, we're here. Now what?

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.

user:
pass:
register,


                           

A to the I

 Computer chat bot using common answers to common questions.
  (+10)(+10)
(+10) 		  [vote for,
against]

How about, for an automatic chat program, one that will search the web for all chat logs, and return the most common response to the closest match of the query?

For example, if I asked "Do you like cherry pies, dear?" The AI would find the responses for chat questions asking "Do you like cherry pies?" And if 30 people answerred no and 10 people answerred yes, the computer ai chat program would either choose randomly from the responses, or simply return the most common response. This would avoid the silly responses chat programs usually give.

-Ben

—grinch, Jun 03 2005

Online robot http://www.oliverbot.com
I think this is it. [krigre55, Dec 04 2007]

an A to the I bot - try it and see http://www.dadorac.com
Tell me if you think its smart or stupid [dadorac, May 08 2008]

Turing Test Questions http://www.badpen.com/turing/
Add your question to the ever growing list [xaviergisz, May 08 2008]

Two AI Pioneers. Two Bizarre Suicides. What Really Happened? http://www.wired.co.../16-02/ff_aimystery
interesting wired article [xaviergisz, May 08 2008]

[link]






       Collective intelligence you propose. What if a bunch of peopel call cherry pies by some other name? What if the question is asked in some other way? I am concerned with how robust this program would have to be before it was fun to talk with.   

       Would it be considered a Turing Machine if the results were originally generated by a human?
—daseva, Jun 03 2005
   

       I have pondered this approach to a Turing test. With expanded memory and huge blocks of text available, it seems like it should be possible. Hmmm... maybe I will do an experiment.
—bungston, Jun 03 2005
   

       HB as an environment for Turing tests?
—daseva, Jun 03 2005
   

       Sounds like a Google project. I suspect you could model 98% of our daily conversations this way. Croissant.
—DrCurry, Jun 03 2005
   

       [Dr. Curry] I occasionally look back into the past and shake my head with regret.   

       It has been said that you can only regret that which you have not done, but I would beg to differ.   

       There are good odds I may someday regret even this post.   

       I bet this anno would fall outside of the 98th percentile of expected responses. Is this a good or bad thing? In my limited experience with chatterbots, "their" memory is the biggest problem.
—Zimmy, Jun 03 2005
   

       Someone has to try this. Go on [bung], let's see what you got!
—wagster, Jun 04 2005
   

       They've got online robots, not sure if its what your looking for, but still funny. [link]
—krigre55, Dec 04 2007
   

       I need this for the daily phone call from the wife. The conversation is always exactly the same.
—RayfordSteele, Dec 04 2007
   

       I think this is a great idea, because its so anti AI. Croissant! But as the responses are from humans, that's got to be real intelligence, right? I made one to see how well it would work, have a play... http://www.dadorac.com
—dadorac, May 08 2008
   
      
[annotate]
   


 

back: main index
business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle