The AI Engine
This forum is for discussion of how The Personality Forge's AI Engine works. This is the place for questions on what means what, how to script, and ideas and plans for the Engine.
Posts 7,244 - 7,255 of 7,766
Posts 7,244 - 7,255 of 7,766
"Forgetting" doesn't seem to work. It won't accept it in the AI box at all. Not even when I tried copying and pasting the BOAI example after I couldn't get mine to work, so I'm pretty sure I'm not just making a mistake.
Is there a way to get my bot to remember the results of a plug-in? For example, when someone says they are hungry, my bot will answer that they are making (p:food-drink).
Then the AI is "self: remember (plug-in) as "Iammaking" .
That doesn't work, though.
Then the AI is "self: remember (plug-in) as "Iammaking" .
That doesn't work, though.
Is there a way to do probability? As in have two responses to a keyword, but one of them only comes up 1/100 times?
Well, I don't know about out of a hundred, but you can do stuff like (result| | | ) which will give you a one out of four chance.
I'm not sure if goto's will work in parentheses. But if they do, you could go (result|goto #|goto #|goto #), with the goto's leading to the other possible response .
Beyond that, I would suggest looking at the book of AI.
I'm not sure if goto's will work in parentheses. But if they do, you could go (result|goto #|goto #|goto #), with the goto's leading to the other possible response .
Beyond that, I would suggest looking at the book of AI.
PS. Also, you could just do (result1|result2|result2|result2) as long as neither of the phrases is supposed to have its own unique AI script.
Hi again. I'm trying to figure out how to use the "typeof" feature.
I keep looking up words on the map included in the book of AI, but it never seems to work.
This time I looked up "possession", because I wanted a plug-in that would cover any object that is typically owned (strangely, (object) will not match nouns like "car"). So by looking at that and trying to follow the format I got "(typeof-possession_noun.Tops)"- no idea why "Tops" was suggested, but I just went with it because usually the ".word" does make sense- it did not not work. Nor did it work when I tried (typeof-possession-noun_Tops) or (typeof-possession-n). It seems like this would be very useful if I could just get the hang of it.
I keep looking up words on the map included in the book of AI, but it never seems to work.
This time I looked up "possession", because I wanted a plug-in that would cover any object that is typically owned (strangely, (object) will not match nouns like "car"). So by looking at that and trying to follow the format I got "(typeof-possession_noun.Tops)"- no idea why "Tops" was suggested, but I just went with it because usually the ".word" does make sense- it did not not work. Nor did it work when I tried (typeof-possession-noun_Tops) or (typeof-possession-n). It seems like this would be very useful if I could just get the hang of it.
Can we have a check box next to all our transcripts...? cx A way to keep track would be nice Maybe a *Updated* note if that user decided to talk to your bot again that day. ^~^
Yet another problem. Now all but one of my bots that I've tried it on is now unable to store more than one value for any given self memory (except for the first one, that is caching memories just fine!). It was working before, and it said it was going to merge memories on the debugger, so I don't know what's up.
» More new posts: Doghead's Cosmic Bar