[squarebear]

Do you have a Pandorabot?

Do you think it could win the Loebner Prize?

If so, why not enter it in this year's Divabot contest?

More details here:
http://www.divabot.com

[Don Patrick]

Results!: http://www.divabot.com/enter.htm

Although I recognise that the winners are commonly strong competitors, I can't say I was impressed by their responses here. Their success rate comes down to 50%. I can see why Izar did well in last year's Loebner selection though, clearly much effort has been put into making it very chatty. One thing I appreciate in these chatbots is that they sometimes remark on the change of topic, showing some awareness.

It's only fair to mention that my non-AIML program wouldn't have fared better. I tried the questions and the few it got right were mostly the kind of questions I have hardcoded (math and such). One question I did like was "Is the sea wet?", which pointed out a hitch in my program, in that "the" sea wasn't automatically inheriting the properties of "a" sea, and I'm quite happy to have opened that door now

[squarebear]

Which is why the transcripts in chatbot contests are so important. Even if you don't take part, you can still learn a lot.

[Freddy]

Thanks Don. 

I think at some point I will try to gather links to as many transcripts as I can find. I agree with SB, they would prove useful.

[Don Patrick]

Oh? Is one of the winners yours, Freddy?

I'm curious about last year's Divabot questions. Perhaps mostly so because I find it remarkable that these are still the kind of questions people are expecting at the Loebner Prize.
On another note, I appreciate the Divabot contest for preventing some swamping of the Loebner Prize, though that was more a concern when there was still a limit on the number of accepted entries. I'm glad I don't have to rush things this year.

[Freddy]

No I don't have a bot in the Loebner, nowhere near. But I will be working on a bot once I have my AIML interpreter going.

I will try a few competitions though one day.

My main thought was to collect them for the good of all and put them in an article or something here.

[squarebear]

I find it remarkable that these are still the kind of questions people are expecting at the Loebner Prize.

These are exactly the kind of questions that the Loebner Prize judges love. I have been at the last 3 contests and have witnessed things like: "Which is bigger, a small mountain or a large lion?", "how many plums can I fit in my shoe?", "howww r youuuuuuuuu?" and "Would it hurt if I stabbed you with a towel?" among many others

The friendly conversational type questions asking about its favourite colour/sport/movie or personality based questions seem few and far between.

[Don Patrick]

I meant to refer mostly to questions that involve the alphabet, numbers, and other questions to which the answer is too obvious to ask, and too easy to rig a "insert subject here" pattern for.
Personally I prefer difficult-ish questions over the "favourite this and that" from last year's Loebner Prize, simply because I didn't build an AI for chatting. The wet sea question I consider to be a valid basic knowledge question, otherwise possible to arrive at with a little inference, I also would accept Noel Sharkey's trick questions because they're clever but not impossible. But a lot of other questions are only there to unmask the computer rather than perform actual "testing" of anything, neither intelligence nor conversation.

Truthfully though, I'm just a little frustrated that the answers my program gets right here are due to some really dumb methods that I rigged in half an hour, rather than the advanced abilities that I spent 2 years programming. If I end up ranking high in the selection round on similar grounds, it would be a dishonour.

[Art]

...
Truthfully though, I'm just a little frustrated that the answers my program gets right here are due to some really dumb methods that I rigged in half an hour, rather than the advanced abilities that I spent 2 years programming. If I end up ranking high in the selection round on similar grounds, it would be a dishonour.

Only in your eyes Don.

[Don Patrick]

That makes it worse!
Oh well, I should be glad to pass to the real event anyway if I do. The selection round is more of a technicality.