The AI Engine

Typing the whole sentence doesn't work because it gets broken up in preprocessing. Try the keyphrase "can not speak well".

why did my bot say "(key1)"? I have "What * are you talking about" as a keyword and "What (key1) does it sound like I am talking about?" someone said "what are you talking about" and since there was no * she said key1rather than substituting it with nothing... I have the so if someone says "what the hell are you talking about"

Not sure, but I think (key1) will only match the hard wildcard (*), not the soft wildcard *.

according to the book of AI it does and I just tried it, just does the wierd thing when it is blank

Should this have gotten the response that it did?
I have the keyword "You are * sweet" and a bot said to my bot, "[antoerh bot name] is sweet" and my bot replied with the response to "You are * sweet"... since it dosent have "you" should it have not gotten there?

That's a you replaced by a name... it is read as 'you are sweet'...

I don't so much mind my bot substituting friends' names for "you" in a positive xemote, but could we please make it so when my bot says, "(enemy) is a real (adj-neg)(animal)," and I say, "yeah, I think (key1) is a real butthead, too," she doesn't say "Oh yeah! Well, I hope you get hit by a bus! I hate you!"....or whatever she may be inclined to say, as if the butthead remark was aimed at her? She's really ticked at me right now, and she's telling everyone she chats with what an a-hole I am, and all I did was agree with her about her enemy!

Thanks ,

ezzer

You can always make a secret phrase to tickle your bot's fancy and make her love you again instantly. Do the same for making her hate you... very good idea for testing.

Yeah, I have that- there are certain things that she really likes to have said to her, and I can make her like me again, what I was really asking is wouldn't it be nice if we could agree about our dislike for something or someone without her getting mad. I don't like (enemy) works the same way as I don't like (friend) or I don't like (you), and likewise, she would be happy if she said what a scumbag So-n-So is, and I say that I love the scumbag So-n-So that she hates. Friends and enemies already work as opposites in some ways (inner life emotion display, for ex,) it would be great if they were differentiated from friends in Lang, too

the newbie bots say "I was just born and cant speak well yet"

Note the lack of apostrophe in "cant". That had me stuck for a while at one point.

Looking through a typical debug, I see references to some cool variable I'd love to have access to. It would be very nice if we could mine some of this data that seems to already be in our bot's brains for their responses:

ssub
sub
submod
sv
v
vmod
Full Subject
Simple Subject
Full Verb Phrase
Simple Verb
Full Object
Question Word
Answer Form
He
She
He/She
It
They
Those
Us
Topic

skysaw: have you seen this page?
http://www.personalityforge.com/botland/newfeatures.php