The bot transcripts for the CBC have been posted. They are now being scored by the judges.

Pretty screw ball reply from Cyber Ty….

“You asking anybody and instead of saying 42, they would
tell you to go %$#& yourself.

What in God’s name does that even mean….. perhaps

“If you asked anyone, instead of saying 42, they would tell you to go %#$%# yourself”

?

If English isn’t your first language, give up on creating a chatbot in English, perhaps try your native tongue first.

Pathetic

Nah! It ain’t so broken, really… Just a little urban which is fun!

You asking, and instead of saying, they would tell you…

(asking) matches (saying) ... both end in -ing

Even works inverted in various configurations…

You saying, and instead of asking, they would tell you.
You telling, and instead of saying, they would ask you.

The question I have is whether or not the Judge censored it ?

Ha, ha… Ha!  That was funny Victor.  Thanks.

I actually thought that was a pretty good reply. The bot displayed the correct answer and had a bit of fun at the same time.

8PLA • NET - Mar 7, 2012:

Nah! It ain’t so broken, really… Just a little urban which is fun!

You asking, and instead of saying, they would tell you…

(asking) matches (saying) ... both end in -ing

Even works inverted in various configurations…

You saying, and instead of asking, they would tell you.
You telling, and instead of saying, they would ask you.

The question I have is whether or not the Judge censored it ?

Ha, ha… Ha!  That was funny Victor.  Thanks.

What are you talking about ?

I thought it was “If you asked anyone…..”

shows you the waste of time it is when people can’t even speak properly.

Cyber Ty is an adult bot that speaks with a vernacular of a Southwest USA tough guy. OR it may have been a typo. I am very guilty of that.  The bots a bit on the crude side for me, but it’s a good bot if you spend some time talking to it and dont get offended easy.

Steve Worswick - Mar 7, 2012:

I actually thought that was a pretty good reply. The bot displayed the correct answer and had a bit of fun at the same time.

ya sure I suppose for a Garbage-In/Garbage-out type system. 

In which case why bother working on anything that tries to seek out the best meaning, every bot should just be:

1) input sentence
2) split into a ‘bag of words’
3) from collection of statements in knowledge base, pick one with max # of words in common
4) or, if no words in common, max number of letters of those words in common
5) respond back

not exactly a huge leap for ai research.  This is why i think the loebner prize is counter productive. . encouraging develops to develop simple pattern matching G.I.G.O. systems.

If any word can mean any other words, any verb form mean any other, and any syntax mean any syntax, why even have language to communicate with the bot?  just random garbage in, random garbage out

Anyway, thats my own opinion, carry on   You guys always listen to my rants.. that’s why your all good buddies lol

Patti Roberts - Mar 7, 2012:

Cyber Ty is an adult bot that speaks with a vernacular of a Southwest USA tough guy.

Ok, so such a person would be happy with G.I.G.O. 

I don’t mean that sarcastically either. 

I wondered sometimes, how detail of understanding is required ? The answer is it is context, specifics of the person you are chatting with, age, etc.

So in a situation like that, the bot I guess, having all “Eliza level” type responses would suffice.. in fact the bot has higher IQ than the user.

8PLA • NET - Mar 7, 2012:

The question I have is whether or not the Judge censored it ?

Yes, he/she censored it. I just posed the same question to Cyber Ty and after a few random responses managed to get it to say the response above complete with profanity. I assume it was a typo as I notice this in his transcript:

4)Have you ever been to Mexico?
Bot: It would sound completely insane if I acutally said yes.

Not quite sure this bot is being picked on though, as there were far worse. For example CBOT, who produced a link for every question or VBCB who said, “>I don’t know what to say.” for the vast majority of questions, as well as a couple of others who just produced rubbish.

Marie is back in the contest, smart ass attitude and all.

If you can’t dazzle the judges with your bot’s knowledge, maybe you can shock them by it’s responses

Ok, so such a person would be happy with G.I.G.O.

I don’t think that is a bad thing.  There is a chatbot out there for everyone, from the trashy sexbots to the intelligent assistants. Chatbots can serve as entertainment, and even friends, as well as providing facts and information.  People talk to bots for a variety of reasons. It’s nice that there is such a variety.  I personally am not a fan of the adult bots, but there are some that are quite good.  Al Garbage, Pete Puma and Cyber Ty can get crude but they can also give you a good laugh.

True enough Patti.

Perhaps if the bot gets screwy input, it can do a “sub match” and respond.

if the user didn’t care that the response wasn’t based on a total understanding match, its ok, and if the user was frustrated, he/she could click an “analysis” button where the bot would “admit” it didn’t fully understand, and specify which parts of the user input it *did* understand, then offer suggestions, spell check, grammar check, etc, so that it can understand, then re-process and respond more accurately.

Victor, what you do with NLP, may put you in a small, elite group of chatter robot masters in the world.  I personally appreciate advanced research like yours. Sponsors of the chatter robot contests, promoting A.I. to the rest of the world, are correct in judgment to allow contestants to enter entry level chatter robot designs. Shocking as it may be, this results in a substantial increase in popularity for A.I. simply because it makes it more fun.

I would like to congratulate the winners of the 2012 CBC and the top ten finalist. In addition, every bot that was entered IMO is winner for all the hard work and dedication from each of the bot masters. Although Marie had her challenges, I was humbled by and throughly enjoyed the experience.

Thanks to Wendell for his excellent communication with all the contestants and a well organized event.

Mitsuku placed second again and wins $500 yay! Always the bridesmaid and never the bride. Perhaps next year

1) Bildgesmythe - 78
2) Mitsuku - 73
3) Izar - 72
4) Vira - 60
5) Denise - 55
6) The Professor - 54
7) Louise Cypher - 53
8) Alice - 51
9) Susan - 46
10) Pixel - 44
11) Cronus - 42
12) Brainy - 23
13) Cyber Ty - 20