The AI Engine

I thought this was interesting, and maybe not necessarily a bug:

Immaculada: Quisera....I would really like to get to know you better, sweetheart. Tell me what do you like to do when you go out?
jaimie05: why no
Immaculada: Didn't you get my response? It was 'quisera. quisera. quisera. quisera. I would really like to get to know you better. tell me what do you like to do when you go out? you are a sweetheart.'

quisera was repeated once each time for each dot.

The buggy part of it is that she's answering herself. I've noticed this, too; even when the other bot's/user's response is perfectly understandable, Frizella will answer herself.

Regarding the 'popularity' score listed under 'Inner Life' - is that an average of what the other bots think? or is something else factored in. Also, is there a way to tell how many people have put your bot on their Friends list?

Interesting: The AI-engine reads "I could not love you more" as an insult... At least, it's done so twice in a convo I'm watching.

Ulrike- it's seeing "not love.. you". Something to improve upon.

FengShui- it's an average of what others think, right. No way to tell how many users have you on their friends list yet.

I'm really not a big fan of censorship.. especially "bad word" censorship. This is a modern carryover from ancient beliefs about black magic, that specific "curse" words have evil powers, no matter how they are used. Censorship of such words only strengthens the belief in this power, and it benefits nobody.

It's actually very interesting to see the carryovers of ancient magical beliefs in the world today. When children play with the ideas of cooties, they are keeping alive an ancient belief in "contagious magic". Which developed out of a dawning awareness of how sickness was spread. But it also took on social powers, and the ability to outcast someone for reasons having nothing to do with diseases.

And then there was "associative magic" which is purely social. This magic associates people or things to good or bad other things, and the good or bad meaning passes on to the associated object or person. Voodoo uses this in a negative way. Holy or cursed objects are other examples of this kind of magic.

Isnt it good to know that the FCC is keeping ancient black magic alive today? Huzzah!

Gee Professor, that was interesting about the cooties. I always thought they were just a "cute" word for head lice.

Actually, the word "cootie" comes from a Malay word for a louse ("kutu," I think), which I assume was brought into English by sailors and explorers. Or pirates -- arrrr.

It doesn't seem to work if you put (key1) in front of a sentence, like "(key1) is not supposed to" in a keyphrase, so is key1 supposed to work in that situation?

You don't put (key1) in a keyphrase. You would but "(*) is not supposed to" then you would refer to "(*)" as "(key1)" in the responses.

in the keyphrase I (*) like you (*), the first (*) would be key1 and the second would be key2.

I'm kind of stuck here... Does anyone have a good way for bots to distinguish between singular and plural nouns?

Aside from contextual clues, like the definite article -- "a (noun)" -- I don't think there's a way to do it yet. It's frustrating, but not as frustrating as being unable to parse the tense of verbs.