AI Zone Admin Forum Add your forum

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

Unofficial AISB Loebner Prize 2017 Finalist selection
 
 
  [ # 61 ]

“Rose’s first “huh?” does immediately follow after another output in every round except 4. It also doesn’t appear to be her normal, capitalised “Huh?” response.”  —- good catch Don. You are correct that Rose does not have a lower case one to emit. But it would be strange that it arose from anywhere else in the system configuration.

 

 
  [ # 62 ]

It seems to originate from your old LPP interface:

LOEBNERVS2010/loebner_chatbot/loebner_chatbot.cpp
char word[MAX_WORD_SIZE]
;
charSkipWhitespace(ourMainOutputBuffer);
ReadCompiledWord(p,word);
if (!*
wordstrcpy(p,"huh?"); // in case we fail to generate output 
 

 
  [ # 63 ]

right. down in loeber copy of engine if script itself fails.

 

 
  [ # 64 ]

I’ll write up my experiences with this year’s contest when I get chance but I can possibly help with Bruce’s query.
In round 4, the bridge program stopped running altogether. Andrew restarted it and so this may be related.

 

 
  [ # 65 ]
Don Patrick - Sep 20, 2017:

It seems to originate from your old LPP interface:

LOEBNERVS2010/loebner_chatbot/loebner_chatbot.cpp
char word[MAX_WORD_SIZE]
;
charSkipWhitespace(ourMainOutputBuffer);
ReadCompiledWord(p,word);
if (!*
wordstrcpy(p,"huh?"); // in case we fail to generate output 

We should patch that right now with something like…

char word[MAX_WORD_SIZE];
charSkipWhitespace(ourMainOutputBuffer);
ReadCompiledWord(p,word);
if (!*
wordstrcpy(p,"WHO DID A REBOOT DURING A LIVE CONTEST ROUND?"); 


Hey Don, 

What were you saying before, something about unnecessarily supporting both protocols?

 

 

 

 
  [ # 66 ]

He had no choice but to restart it. The bridge stopped working and was no longer on the screen. Although the route of:

Judge -> new protocol -> bridge -> old protocol -> chatbot -> old protocol -> bridge -> new protocol -> judge

for each interaction, did seem to be tempting fate a little.

 

 
  [ # 67 ]

I am thinking, Dave’s thumbdrive solution for ChatScript would have made the (AISB) bridge unnecessary.

And, would have provided log files

But, the thumbdrive disappeared (after being signed for) and the log files were deleted.

A bridge from character-mode to message-mode bi-directionally, is kludgy, I think.

It is like trying use email to interact with a chatroom.

 

 
  [ # 68 ]

As I say, I’ll write up my experience once I get a moment to breathe but I clearly saw a text file of instructions that Andrew was following where the first line read something along the lines of “If the USB drive doesn’t arrive…..”. There was no USB drive attached to Bruce’s allocate laptop.

Daniel Burke took a few photos of the contest. Check out this one he took of Bruce’s laptop. No USB drive.

Image Attachments
36441239324_d8047fe372_z.jpg
 

 
  [ # 69 ]
Steve Worswick - Sep 20, 2017:

As I say, I’ll write up my experience once I get a moment to breathe but I clearly saw a text file of instructions that Andrew was following where the first line read something along the lines of “If the USB drive doesn’t arrive…..”. There was no USB drive attached to Bruce’s allocate laptop.

Daniel Burke took a few photos of the contest. Check out this one he took of Bruce’s laptop. No USB drive.

I think I found the thumbdrive (Thanks to Daniel, and Steve)...

That looks like a green thumbdrive, in between the laptop and the laptop power supply.

From the empty plate, it looks like someone was too busy eating to bother reading the instructions.

Maybe they ate the thumbdrive! LOL

 

 
  [ # 70 ]

Wrong color green, Tom. raspberry

 

 
  [ # 71 ]
∞Pla•Net - Sep 20, 2017:

Hey Don, 

What were you saying before, something about unnecessarily supporting both protocols?

That is was unnecessary for you, because you already had an interface that worked with the new protocol. Just as for Bruce, if they had used the usb drive that worked with the new protocol, they wouldn’t have had to use the bridge that works with the old protocol. I don’t appreciate your comment or your frustrations.

 

 
  [ # 72 ]

Am I going to have to shut this thread down here? 8pla, your posts are getting a bit trollish, and that’s not a good thing. Let’s try to work together here. And Don, sometimes the best response is no response at all. Poking trolls only encourages them smile

 

 
  [ # 73 ]

I’m mostly attempting to adjust an autistic person’s skewed view of reality, seeing as he’s been insinuating various injustices on Richard and Steve’s end since the selection round, to which the wiser silence of others has been no obstacle. I’ll try not to try.

 

 
  [ # 74 ]

I’m sure you’ll all be glad to know the USB Stick of Rose arrived in Andrew’s office on tuesday.  It took 2 business days for it to travel via fedex from California to the college in london, and then 3 business days to travel down some corridor.

 

 
  [ # 75 ]

Thank you for helping to incubate my hypothesis with your interpretations. Apparently, some contestants have become emotional.  If so, set those emotions aside and take a hard look at the mathematics to avoid future consequences.  Suggestion: Don’t blame your chatbots, they are not the problem. Blame yourself if you were not aware of the, Hypothesis Of Undefeatable Odds. Another interpretation of this hypothesis is to consider that no rules were broken. If there is an unwritten rule that qualifying round contestants are permitted entry to the final round, then next year, contestants had better prepare to team with a qualifying round contestant in the final round or avoid lowering their winning probability.

Now, if the Chatbots Organization members don’t object, I would like to continue discussion of my Hypothesis Of Undefeatable Odds.  At this point in my hypothesis, I am considering Probability. In this Experiment, the events are non-mutually exclusive. The addition of A (Alice hypothetically speaking) causes M (Mitsuku hypothetically speaking) to be counted twice. So let’s plug in the contest data to the Additional Rule: When two events, A and M, are non-mutually exclusive, the probability that A or M will occur is: P(A or M) = P(A) + P(M) - P(A and M).

Probability

1. Qualifying Round data:

Mitsuku
Rose
Uberbot
Midge
Tutor
Colombina
Arckon
Johnny & co
Aidan
Alt Inc
Talkme
Izar
Simplex
Alice
Momo
P.A.M.

2. Final Round data:

Calculate Probability for final round without Alice in the final round:
P(Mitsuku) = 1/4 = .25
P(Rose)  = 1/4 = .25
P(Uberbot) = 1/4 = .25
P(Midge)  = 1/4 = .25


Calculate Probability with Alice in the final round:

P(Alice) + P(Mitsuku) = 2/5 = .40
P(Rose)  = 1/5 = .20
P(Uberbot) = 1/5 = .20
P(Midge)  = 1/5 = .20


Disqualify Alice and Mitsuku in the final round (hypothetically):
P(Rose)  = 1/6 = 0.166666667
P(Uberbot) = 1/6 = 0.166666667
P(Midge)  = 1/6 = 0.166666667
P(Arckon)    = 1/6 = 0.166666667
P(Columbina)  = 1/6 = 0.166666667
P(Tutor)    = 1/6 = 0.166666667

So, the entry of A (Alice) to the final round doubles the winning probability of M (Mitsuku). And, the winning probabilities of Rose, Uberbot and Midge are each lowered by over 5%. In theory, disqualifying A (Alice) and M (Mitsuku) adds three contestants to the final round. This may suggest Rose, Uberbot and Midge all lost a competitive edge.  In short, if there is an unwritten rule: Qualifying round contestants are allowed in the final round, then the other teams had better figure that out, to be competitive.

 

‹ First  < 3 4 5 6 7 > 
5 of 7
 
  login or register to react