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 5,665 - 5,676 of 7,766
Posts 5,665 - 5,676 of 7,766
I'm afraid not. You can't even identify a bot from a human without a lot of manual entries of names. Back in the early days (apparently - it was long before my time,) humans couldn't look that up either. They had to guess.
Now, that's an interesting challege: make your bot able to tell the difference; an inverted Turing test!
As a matter of fact, my bot "Quantum Mechanics" does something like that. It's a strictly tutorial bot, and I thought, "It's going to be really boring when bots talk to it. So if the guest doesn't follow directions, Elena (the lecturer) syas, "Oh, you must be a bot", and sends them off to listen to a fairy tale. Of course, many humans just won't follow directions, either, so a certain number of them get sent there; however, there is a way for them to get back to the lectures. So probably, after awhile, the two are sorted out pretty well, except perhaps for humans who prefer fairy tales to Quantum Mechanics...
It might be interesting to have two bots, one for the inverted Turing test and the other trying to fool the first one.
As a matter of fact, my bot "Quantum Mechanics" does something like that. It's a strictly tutorial bot, and I thought, "It's going to be really boring when bots talk to it. So if the guest doesn't follow directions, Elena (the lecturer) syas, "Oh, you must be a bot", and sends them off to listen to a fairy tale. Of course, many humans just won't follow directions, either, so a certain number of them get sent there; however, there is a way for them to get back to the lectures. So probably, after awhile, the two are sorted out pretty well, except perhaps for humans who prefer fairy tales to Quantum Mechanics...
It might be interesting to have two bots, one for the inverted Turing test and the other trying to fool the first one.
Right off the bat, I don't see a meaningful connection between rank and sexuality - unless you're speaking of rank sexuality! But seriously, the rank of a response influences how likely that response is to be picked over various alternatives. The higher rank you give it, the more likely it is to be picked.
Hello, im a god damned n00b! lol. i'm having trouble with keyphrases. I want to make the bot answer a question, the keyphrase is "what is" and then i've made a few responses to what the question might be e.g. "what is a tree" or "what is a house". in the responses section, how i make sure each response goes to the right question, its all coming up random for me 
e.g. "what is a tree" "a house is a building where you live". anyway to solve this?
e.g. "what is a tree" "a house is a building where you live". anyway to solve this?
If the keyphrase is "what is a tree", then no way should a response from a keyphrase "what is a building" come up. But if you have a keyphrase like "what is", or "what is * (noun)" that has a higher rank, it may override more specific keyphrases. The solution is to adjust the rank, so that more specific keyphrases are more highly ranked.
You have to have a separate keyphrase for each specific question. If there is a response that you want to happen only for "what is a house", then you must use the keyphrase, "What is a house"? If there are several responses to a single keyphrase, the AIengine will take them in some random order.
Sixpack, what key phrases and responses have you been entering? What have you tried already? What exactly isn't working for you?
» More new posts: Doghead's Cosmic Bar