AI Zone Admin Forum Add your forum

NEWS: survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

My friendly pre-CBC contest
  [ # 16 ]

Hi Ken, I am trying to reply back to the message you sent me but comcast has blocked my message as spam. Do you have another address I can try?


  [ # 17 ]

Thanks for the alternative email but unfortunately, that was also a comcast address. It says the content of my mail may be spam and has decided that you don’t want to see it. Rather a strange thing to do for a supplier to not give you the option of seeing it. However, do you have say a hotmail or gmail I can send my message through?


  [ # 18 ]

That is odd. I have never needed it before. I have done a lot of international mailing. It has been so reliable that I have let all my other addresses go. I guess I’ll have to fire some back up.


  [ # 19 ]

Darn I wish the dog question was worded First letter in the word dog…Bildgesmythe would have been much more impressive LOL Funny how many ways to ask the same question


  [ # 20 ]

I just had a look at the conversation logs for Morti, and was saddened (but not surprised) ad how dismally he performed. It was to be expected, though, since I had so little time to work on his responses, due to the long awaited release of Program O version 2.0 - Running support and revising code for the project has taken up pretty much every minute that’s left over from building my back porch into a full-on room. downer But enough complaining and excuses. Next year will see Morti’s cousin, Morgaine, enter the fray, with a new platform, and all new responses. cheese I hope to have her up and running in a few days, once I get the last few remaining bugs out of the interpreter. smile


  [ # 21 ]

Thanks Ken, that address appears to have worked. Hope the information I have sent you is useful.


  [ # 22 ]

Hey Dave -
Yeah, Zoe did pretty bad in this contest, too.  I am

Here are some notes:
1) The “surreal moment” response that Steve saw is an error response related to how the conversation initializes.  It took me a bit how to figure that out, so a lot of people saw the “surreal moment” response too many times.
2) She actually had some really good conversations when I look through the logs - but happened to bomb out on the specific CBC questions
3) I won’t be entering “Zoe” in the CBC next year.  I’m working on Zoe’s sibling… which will be a clone of Zoe but will admit to being a bot.  One of the interesting things in the logs I saw is that people DO NOT like that she won’t admit she’s a bot… she was made for the Loebner contest, and programmed to lie in that area.  Wow… people DO NOT like that!  So her bot-admitting sibling will be in the CBC next year… once he/she is complete…!

In the meantime, she will have good days and bad days… hopefully when they judge for the initial round of the LPC, it will be a good day!  No way to know….!



  [ # 23 ]

Wow… people DO NOT like that!

People will get evil about a lot of things, just ignore them and make some sarcastic comebacks.  You will never please everyone. I do like the fact that the CBC gets you a better quality of chat, for the most part.



  [ # 24 ]
Adeena Mignogna - Mar 13, 2012:

Yeah, ... hopefully when they judge for the initial round of the LPC, it will be a good day!  No way to know….!


Without indicating, or implying any fault whatsoever with the Chatterbox Challenge contest or any of the judges… My impression for the sake of informal conversation only, is that it seemed this year, the 100th birthday of Alan Turing, that some of the top chatter robots, like Zoe, Skynet AI, Elbot as well as a few of the new chatter robots, like SlideSpeech and My Marie… May have been letting some of us win!  Last year, a tie eliminated me from round two in the LPC.  This year a tie saved me, for round two of the CBC.  This was my first time to round two of a chatter robot contest.  Thank you, everyone.


  [ # 25 ]

I personally believe that the bots have a mind of their own wink.  There are good days and bad.  I am always impressed at how good all the bots do in answering the CBC questions.  I actually find some of the “wrong” answers great.  The first year I got into the final round was on the botmaker vote they used to have…I was the only one that voted.  Worked out great, I won a bronze.  That’s what is so fun anything can happen.


  [ # 26 ]

Well, Zoe is truly “tuned” for the LPC sample questions. 

The issue with the open ended questions is that ... it’s so random.  And I didn’t both to do a *ton* of “cute” answers which would have done well in CBC because in LPC, it doesn’t always go over so well. 

But I am surprised that Zoe got 0 points for some of her responses… the CBC rules seem to me that an acknowledgement of the question was worth at least 1 point.  Oh well!  A learning experience, to be sure…!



 < 1 2
2 of 2
  login or register to react