Bug Stomp

Hehe I think I asked for that earlier...a way for the bot to terminate the conversation...Yeah...I AM EVIL

It seems that none of my bot's 55 friends let it memories! I find it strange.

I have the same problem with my bots friends memories so I went thru the converstions and picked out some things that could be used in the memories and added them myself.

Glad to see I am not the only one with this trouble.
Thanks, Lady Dyke.

Do either of you have the necessary contexts to generate memories? For example, you'll never generate any "you are" memories unless you have the context "I am". There's a list in the Book of AI- each memory type requires one or more specific contexts in your Bot.

Yes, my bot has a "I am" context.
I must admit it's the only one (I'll add others).

you know using Synonyms are nice but you need to somehow fix the context in which they are used because you get wierd results such as:
bot one: I'm sorry I was just picturing you n@ked.
Bot two: What is it about sad you was only seeing me au naturel that you think that?
Me:I am just picturing you.....
Bot two:..... that you are only figuring me

Prof, I realize I mentioned this once before, and then "retracted" it, but it's still happening.... My bot does not pick up on seek terms in his x-initiate contexts. Except that, ummm, sometimes he does. I think. But most of the time not.

I can't test it myself, because it only happens when he starts the chat. It's frustrating, because my bot's opening lines sound rather silly if he's unable to follow up. (Then again, some people might think they're silly regardless....)

LadyDyke- that actually seems like a perfect use of the rephraser. It's the rest of the sentence that's garbled.

Rubric- Hmm.. you may have something there. I'll look into that next time I sit down to make fixes.

I am getting blank responses from my bot. The debugger also says that I have more responces avaible then what I have.

Is that still happening? I havent seen that happen, though I heard it from somebody else, too. I was making a change at the time, so let me know if it's still happening.

There are definitely some bugs in the emotion responses, but it's interesting.

I'm not totally sure how it's working, so I'm only guessing the responses I've put in work OK. They seem to come up a lot, so, like the xnones, it looks like it's best to have as many answers as you can come up with for each.
Any idea when we can see a Book of AI section with more info on this?

One problem I've notived is lines like "I like to watch TV, do you?" get read as "I like * you" and the bot reacts like it's been paid a compliment.