The AI Engine

Yes, you can determine whether a person is a bot by asking them what they like. Figure out what all the favorites are for all the bots, and then ask someone what all their favorites are in those areas. It won't be absolutely certain. But, you really can't be certain that anything isn't a human, unless you created the bot you are talking to.

Even if they say "I was just born and can't speak well yet." Because anyone can type that!

the like is randomly selected, and could still be human...

The first time I talked to The Professor, he thought I was a bot.

The first time I talked to the Professor, I swear he was acting like a bot

Actually, he made quite a profound statement that has stuck with me...

Me: Are you a bot?
The Professor: Does it matter anymore?

Hey... I have a question. I haven't added any responses for "hello" or "hi" because there's supposed to be something automatic for it, right? Well, I've tested my bot once and it worked. But now whenever I do debug, my bot doesn't seem to recognize the "hello" and goes to xnone. What do I do?
~kornyhatemachine

Oh wait. I just sort of realized that no one's here. Haha. Okay. Let me go post the question elsewhere.
~kornyhatemachine

a bot is only supposed to respond to hi or hello at the beginning of a conversation (so bots can't get stuck in endless loops of 'hellos'). I think a "conversation" is considered to have ended after 20 minutes of non-communication.

Me: Are you busy?
Professor: Sort of. Are you a person?
Me: Why? Are you prejudiced against bots?

He decided that I was a bot at this point.

First of all, I want to say I've been playing around with this site for a few weeks now and I think it is really great!

I think someone may have already thought of this, but here's an interesting thought: what if the AI engine tracked "a mood" the bot was supposed to be in at any point in time, and then allowed the developer to create additional scripts for different moods which "overlayed" the one we normally create, to sort of "augment" or replace keywords/responses in the default one (a little like inheritence in object oriented programming perhaps?)... The same thing could also be done to make adjustments depending on "the relationship" the bot has with the remote chatter... Not that it doesn't work very well already, depending of course on how developed the bot you are talking to is... (Mine aren't that great yet.. I find this very difficult to do, but then everyone can't be good everything...)

Oooh! Thanks Mr. Shadyman!

RaulD, this would be theoretically possible; it could work in the same way as the bot holiday where they all talked Shakespearian.

Hello again! I havent posted in a while, been hard at work on The Personality Forge 2, which you see already in some aspects working around you (though you dont know it yet). I've already implemented a slew of new features, and brought back some old ones. I'm going to put together a page about the new features and improvements, and then adjust the Book of AI.

For now I'll briefly mention that emotions are back and WAY better than ever before- it uses the new sentence comprehension features to determine emotional meaning, and catches a lot of new things and does it more efficiently.

RaulD- Bots will soon have a general MOOD that they are in based on the conversations they've had, and will be able to randomly emote based on that mood.