Bot Contest

I hope you didn't take my post as complaining or an indication of ungratefulness - I am incredibly grateful for the opportunity. I'll admit to not voting for the 10th bot, but that was because I hadn't had the opportunity to talk with most of the 104 bots and didn't feel I could make a good decision.

I should have done a better job correcting/clarifying myself in my last post, since I did use the words "on the user's end" at one time in a previous post and regretted it, since I realized afterward that it wasn't what I really meant- I probably should have rescinded that statement in a later post, but let it slip through the cracks, sorry.
I didn't say the technical problems were the user's fault in my last post, because I didn't think they necessarily were, but was more curious to know if other bots had experienced the same, or if anyone knew what it was- and should have said so. It was in no way intended as a complaint against the contest, just an "incident report." Now knowing per you that it's something Benji needs to look at (is it a bug?), I guess the next step is to post it on Bug Stomp.

As far as the transcript for Judge 4, I was just hoping someone had a copy, as I naturally would have liked to read it, but I'll live regardless.

Wendell-The contest was great, and your efforts have been amazing, as is the opportunity it offered all of us who participated.

Thanks again,
-Heather

no comment on the grounds that I may execute myself. opps incriminate

The strange apostrophe will come up if the user is not typing his reply into the chat field, but using a text prozessor like MS WORD and copy & paste. MS WORD is using the typografical correct character version of the apostrophe, but this character is not included with the standard ISO-character set of web pages. So the apostrophe will be replaced, and it depends on the used ISO-Set of the page how it will be replaced. The PF pages even missed to define/include an ISO-Set in the header of the web pages.

Thanks for the explanation, that makes sense. I had something similar happen a few weeks ago when I cut and pasted a Word doc into a Webmail message and sent it to several friends and family members. When everyone got my message, it came to them all garbley like that, many of them replying things like "Woah, what language settings were you using?!" and my original message at the bottom showed the messed up way they had received it. I figured it had something to do with my Word settings and made note to myself not to cut and paste into my Webmail, but still had no idea where the problem's root lay, with my Word settings, or with the Webmail's parameters, but that something wasn't matching up. (I don't have the lingo down nearly well enough to talk about ISO's, etc.) .

I didn't place a vote because I didn't understand on what basis we were voting. Was it supposed to be based on the 10 questions responses or just on the bot in general?

I would hope it's based on the bot in general. Otherwise the contest have to be called the "Chatterbox 10 Question Knowledge Contest", and the ability to converse would be not important. :-)

Tinman do you have a bot in this contest? If so tell me who it is.

Here are comments made by judges relating to the PF. You may find it helpful or maybe not but regardless here they are.

Questions - The Personality Forge bots all have that same basic look. Is that a disadvantage for them?

Yes, extremely. I was going to suggest that perhaps they should be ineligible for certain parts of the competition since they do feel so formulaic...

However, the same judge said this about Aries:

Well, I went by the guide you sent me, focusing most on conversational ability. In both cases (Aries and Jabberwock) I was able to have a conversation that was very, how to say it, unbot-ish.

Towards the end of the conversation Aries surprised me big time (I'd have to look at the log to show exactly what) by making a connection between that and something I'd said much earlier on. It was that surprise, that moment of "maybe he is alive," that put him at the top
of my list. It may have just been a fluke of the right questions at the right time, but it was my emotional reaction that had me rank him so high.

The botmasters behind Julie Tinkerbell, God Louise and Frizella are very creative and the bots' responses show that. However, it seems all of them get by through faking and don't seem to be huge improvements on the old Eliza bot.

Another judge had a totally different view of the importance of an interface.

To me, looks are something any botmaster can make stunning if he/she wants to invest the time and money. The appearance of the Personality Forge bots being similar was a good thing to me -- it gave me the impression they were focused more on the nuts and bolts, the simulation of conversational intelligence, rather than appearance, which is what the rest of the stuff on the net is for -- flashy-looking, unintelligent web sites.

Comments from another judge on PF interface:

Interface: it was okay, I would have preferred an animated bot to speak to

Interface: same old Personality Forge so now it's tiresome

However on Julie Tinkerbell: Interface: I liked it, because the character was animated and told you when it was “typing” a message.

Even though Jabberwock won with no interface I know it has an impact with many judges. One judge said this:

Jabberwock - lost out to Talk-bot on interface

Benji's Flash interfaces like the one Julie Tinkerbell uses are definitely a plus.

Wendell

The judges' comments are appreciated, but the most important part of the critique, the part about our bots' conversational ability, would be much more helpful if we had a point of reference. Just as the commenting judge said they'd have to review the logs to be able to explain what they were referring to, without seeing the logs, there's no way for us to really know what they meant, or why they felt that way.

Well like I said Heather the system fails apart when people don't follow instructions. I told all the judges to identifiy themselves. Obviously, they didn't do so.

There will be a 2006 CBC but based on recent events I will not be holding another PF contest. The contest site is being moved to www.chatterboxchallenge.com and everybody must re-enter if you want to participate. This means your bot will be removed from the contest site until you re-enter. So if you want your bot to have the the benefit of the extra traffic the contest might generate feel free to enter asap. I've posted some new rules but this is still in the earier stages so nothing is set in stone.

Wendell

You gave very clear instructions, Wendell, no doubt- it's clearly not your fault.
If the judges still have the transcripts in their possession, perhaps we could still get ahold of a copy, if they'd be nice enough to send them?
At this point, I'm pretty sure I've identified them in my own transcripts, but I'm sure we all are just as interested in eachother's transcripts (and enjoyed seeing them posted in previous phases of the contest) as a basis for comparison and means to improve, not to mention raw curiosity and entertainment value.