Personality

OK,that sound like mine, I said thought I was getting AI-Script, Not that my bots were getting any smarter,just like my kids I have very little control over what they say or do.

Exactly!

This is driving me crazy. Ok I have this keyphrase: 'do you like better * or *' the responses work when I follow that exact pattern, but when I say 'do you like better: item1 or item2' this colon keeps messing it up. I've tried everything, including making a separate keyphrase for when they use a colon but that still doesn't work. Can someone please tell me what I'm doing wrong?

What does it look like in debug? What happens with that colon in the pre-processing?

I'm trying to have a new keyphrase that will add the following memory:

remember (key1) as "youneed"

Should the keyphrase be:

"I need *"

I entered "I need *" as the keyphrase and the responses were:

* Why do you need (key1)?

* That's funny. I also thought that I needed (key1), but then I changed my mind.

But then, it was coming up like this:

Me: I need a doctor!
Duo Maxwell: Why do you need a?
Me: I need to write a letter!
Duo Maxwell: That's funny. I also thought that I needed to, but then I changed my mind.

So, I went ahead and changed it to:

* Why do you need (key1) (postkey)?
* That's funny. I also thought that I needed (key1) (postkey), but then I changed my mind.

And then, he got the right responses:

Duo Maxwell: Why do you need a doctor?
Duo Maxwell: That's funny. I also thought that I needed to write a letter, but then I changed my mind.

Only problem is, the memory that will be saved as (key1) is "a" and not "a doctor"

Should I change the keyphrase to:

"I need (adjartnounprep)"

(I actually did try that, but then the bot was totally ignoring it, and going to his responses for "i")

I hope that wasn't too confusing.

Ok here's what I said: 'which do you like better: chocolate or vanilla'

Now here's some of the stuff I saw in debug (it changed it)

New Sentence: ' which do you like better , chocolate or vanilla '

Find: do you like better * or * (28) Time: 0.93

And here's his response: Bot: I'll stick with (key2)

But I still can't figure out what to Do about it

I keep getting a line in debug about the word "or". It says that "or" was ignored...I can't understand why an important word like that would be ignored, but it has been cuasing me great difficulty too.

Good heavens, Frizella is interpreting the strangest things as xcommands. Recent example:

Frizella: (to Little Mu) Little Mu! Gosh darn it, it's good to see you again.
Little Mu: Great to see you!
Frizella: Try a little sugar. When you just say " great to see you!" I get all contrary.

It makes her sound quite a bit more prickly and cranky than she really is. I put "Great to see you!" through debug and it sure enough finds it to be an xcommand. Any ideas why?

The verb "see".

Thanks, Eugene! I made a higher-ranked keyphrase and now she's being a little more polite. I'll have to watch her transcripts and do the same thing for other phrases when I see her getting snippy again.

Here's a related question: is it possible to create a keyphrase that allows her to skip xcommand if the apparent command includes "please"? I don't want her to use her "Try a little sugar" line if the request was made politely in the first place. Can an AIScript for a particular xcommand response specify that it's only used in the absence of "please"?

Why not just write a keyphrase for "please * (verb)" -- and another one for "(verb) * please" ... That wouldn't take very long, and if it's ranked high enough it should trump xcommand.