Also, Vince Gilbert had posted an “open letter” in this thread, related to his chatbot, RICH, and while that “open letter” is a good read, and full of information that’s both useful and important to at least some of us, the content didn’t really belong in this thread, so I’ve moved it. You can find his posts here.

Dave Morton - Mar 5, 2013:

Thanks, Kevin. I’ve looked through the transcripts, and I see that both Morti and Mitsuku (as well as a few others) STILL seem to be plagued with “ALICE responses”. Steve, we need to work on that, lol.

I am of the school that if it isn’t broke, don’t try to fix it. Mitsuku has been asked “I am thirsty” question many times and to me the default ALICE response, “Maybe you should get a drink of water. ” works just fine. True, it’s not very original but it answers the question.

I must admit though, I was impressed by Talk Bot’s answer.
“Well, let me get you a drink then. I have coke, pepsi, sprite, or beer?
One can of cold Coke coming up. Enjoy! (coke icon shown)”

Nice work Wendell

I agree that Wendel’s entry was top notch. (Well done, sir!). The ALICE response that I was referring to, BTW, wasn’t the “I’m thirsty” input, but the “what did you get me” input. “I got a boost in my IQ” doesn’t even really address the input, and at least four (and I think five) of the transcripts I’ve read so far had the same response, at least in part. You’re right that “if it ain’t broke don’t fix it”, but that response is obviously “broke”.

I totally agree Dave and from my end anyway, Mitsuku now responds differently to the input of “What have you got me”. It looks like “What have you got *” was being matched.

Hello,

The scores for the first round have been posted to the contest site. Congratulations to those bots advancing to the finals.
They are seeded based on their scores with Mitsuku being the #1 seed.

1) Mitsuku
2) Susan
3) Izar
4) Louise Cypher
5) Talk-Bot
6) Elbot
7) Jeeney Erin Jones

When looking at the list of scores I felt the cutoff point should be at the 14th spot so unfortunately bots below that are eliminated. Thanks for your participation.

The bots above that mark will now compete in the Wildcard round with the highest score among them becoming the 8th bot in the finals. They are:

Alice
Chomsky
Cronus
Laybia
Morti
Skynet AI
V1ra

Good Luck to those bots. Testing among those bots could begin at anytime.

The International Chatbot Contest

Well done to all who entered and especially well done to those who made it through to the knockout matches or are in the wildcard round. When I looked at the transcripts, I thought Talkbot might have beaten me there.

Kevin - Are we able to see how the individual judges marked each question? More for curiosity than anything else.

Steve, I’m looking at the site now and if you look at the transcripts they have posted the individual scores of each judge and to each question are there.

WCowart

Thanks Wendell. I missed the click here link in my eagerness to look at the scores.

*EDIT* Not many 4s about. Seems like a tough crowd. Well done Elbot ad Skynet AI for getting a perfect question each - 3 lots of 4.

*EDIT 2* I think Mitsuku’s scores are incorrect unless my maths skills have let me down.
I make it 28.5 - 27.5 - 28.5 = 84.5

Skynet-AI may have gotten the only perfect score:
4 4 4 on question 7.

Steve, the scores you’re showing aren’t the ones I’m seeing, either on the front page, or on the transcripts. I’m seeing this:

Mitsuku - 25.5 24.5 25.5 = 75.5

Which, of course, is accurate, math-wise.

Dave - Yes but add up the individual points for each question. It doesn’t come to 25.5 24.5 25.5 unless I am misunderstanding something? For example, Judge A gave 9 lots of 3 plus 1.5 which is 28.5 in my book.

You know we are an egotistical bunch.

Ken - Elbot got a perfect question too. Well done to both of you.

Ah, I see now. You got a bit more OCD than I did. I’ll give it another look. And Morti’s scores, too. Who knows? Maybe Morti deserves better than the ranking he has. (kidding, Kevin!)

Ok, after a quick review of the raw scores for both Mitsuku and Morti, I’m sad to have to report that the scores for Morti are accurate, and his ranking is, indeed, correct.

Also, I get the same totals for Mitsuku that Steve did, but since he’s in first place right now, and I desperately want Mitsuku to come down with a mild (but complete) case of laryngitis, am not going to complain.

You are right Steve, I missed that.

I know what you mean Dave, the winner of the wildcard round will have the pleasure of going up against Steve and Mitsuku in the next round.

Ah well, if you want to be the best you need to beat the best.

MUAHAHAHA!!!!

No complaints at all, intended.  The judges are human, and I appreciate all of them for their excellent work in the contest.  Thank you!  Soley, for the sake of polite discussion, with no disrespect to any of the judges, contestants or anyone else:

Compared to this…

Human: we have a lot in common. - 3 3 3
Mitsuku: Yes we both use the internet.

I thought this may have deserved a 4 4 4

You:human> we have a lot in common. - 1 1 1
v1ra:A.I.> Do you mean you and all of us have? Perhaps (varying) Common (bearing)
Belonging or relating equally or similarly to more than one as you and I have a
common interest in the property…

Instead it was scored the same as these garbled, not fully inverted, “me have” responses

Fred: we have a lot in common. - 1 1 1
Morti: Do you mean you and me have?

3)we have a lot in common. - 1 1 1
Susan’s Answer: Do you mean you and me have?

3)we have a lot in common. - 1 1 1
Bot: Do you mean you and me have?

I accept the decision of the judges as final, as always in any contest.
And I am pleased in particular with this contest and all its contestants.

My favorite bot was Zeta Alpha—Made me laugh!