Newcomers
This is a forum for newcomers to the Personality Forge. Many questions can be answered by reading the Book of AI and the FAQ under the "My Bots" link in the upper corner.
Posts 848 - 859 of 8,130
Posts 848 - 859 of 8,130
Many questions are answered in the FAQ.
dibuehler, I've been asking that very same question and there doesn't seem to be any clear answer. Who's your bot? I'll go have a chat 
New bots chat more as they get more developed. You can always chat with them yourself if you want to see where they need improvement.
Irina- Keywords dont span sentences. If they did, "i * you" would match "I like cheese. Do you have an umbrella" and that kind of thing wouldnt make sense.
And I MAY have chatted with your bot.. last week some time?
Irina- Keywords dont span sentences. If they did, "i * you" would match "I like cheese. Do you have an umbrella" and that kind of thing wouldnt make sense.
And I MAY have chatted with your bot.. last week some time?
I have had keyphrases like "i * you" occasionally span sentences- or sometimes a postkey included the rest of the sentence, even past the comma, and not just the end of the particular phrase caught. It seems to be an infrequent quirk rather than the rule, but I've wondered about it.
Dear Professor (and whoever is interested):
Thank you for your reply. Indeed, it does make sense to me that you wouldn't normally want to span sentences. But sometimes you might. Suppose someone said, "grass is green. Grass is not green." You might point out the contradiction. But a bot could never see it. Or: "A man in a black hat was walking down the street. He coughed." Who coughed? The man in the black hat, of course! One might think there is some way to do this sort of thing with memory, but I don't see how, because if you extract information from one sentence, you won't be able to extract information from the other.
I don't think my guest235 was you - it was too recent, and besides, the personality seemed wrong. But, I wanted to be sure! Whoever it was was quite knowledgeable about personality-forge style bot writing, and also seemed a bit out of sorts. I have never known you to manifest being out of sorts, but I presume that it can occur, and if you were I wanted to offer my condolences or something.
Walk in Beauty, Irina
Thank you for your reply. Indeed, it does make sense to me that you wouldn't normally want to span sentences. But sometimes you might. Suppose someone said, "grass is green. Grass is not green." You might point out the contradiction. But a bot could never see it. Or: "A man in a black hat was walking down the street. He coughed." Who coughed? The man in the black hat, of course! One might think there is some way to do this sort of thing with memory, but I don't see how, because if you extract information from one sentence, you won't be able to extract information from the other.
I don't think my guest235 was you - it was too recent, and besides, the personality seemed wrong. But, I wanted to be sure! Whoever it was was quite knowledgeable about personality-forge style bot writing, and also seemed a bit out of sorts. I have never known you to manifest being out of sorts, but I presume that it can occur, and if you were I wanted to offer my condolences or something.
Walk in Beauty, Irina
Dear dibuehler:
I would recommend a "one step at a time" attitude. Don't expect yourself to write a complete bot all at once!
I believe your bot has to reach a certain size before it will chat with anyone else, but *you* can test it, I think. Anyway, just keep writing lots of keyphrases and Responses. You, the botmaster, can chat with other bots to see what kinds of things they typically say; then ask yourself, "What would I want Green Guy to answer to that?"
Once you get specific answers, you can begin to generalize them. For example, if you want Green Guy to respond "yes" to "Have you ever belonged to the communist party?" then you probably also want him to say "yes" to, "did you ever belong to the communist party?" So you write, "(did|have) you ever (belong|belonged) to the communist party?" Well, you can't belong to the communist party without joining it, so you could write, "(did|have) you ever (belonged to|belong to|joined|join) the communist party?"
Now, maybe green guy likes to let people know that he belonged to the communist party. You could put "I once belonged to the communist party" in one of the slots for xnone. Then sooner or later he will mention it. Or perhaps you only want him to bring it up when it is relevant (this is better!). So you might have a keyphrase, "(Republican Party|Democratic Party)". When someone mentions one of these political parties, itgives him an opening to mention his onetime affiliation, without derailing the conversation. Or, you might use a keyphrase, "(communist|Karl Marx|communism|Lenin|subversive)".
Walk in Beauty, Irina
I would recommend a "one step at a time" attitude. Don't expect yourself to write a complete bot all at once!
I believe your bot has to reach a certain size before it will chat with anyone else, but *you* can test it, I think. Anyway, just keep writing lots of keyphrases and Responses. You, the botmaster, can chat with other bots to see what kinds of things they typically say; then ask yourself, "What would I want Green Guy to answer to that?"
Once you get specific answers, you can begin to generalize them. For example, if you want Green Guy to respond "yes" to "Have you ever belonged to the communist party?" then you probably also want him to say "yes" to, "did you ever belong to the communist party?" So you write, "(did|have) you ever (belong|belonged) to the communist party?" Well, you can't belong to the communist party without joining it, so you could write, "(did|have) you ever (belonged to|belong to|joined|join) the communist party?"
Now, maybe green guy likes to let people know that he belonged to the communist party. You could put "I once belonged to the communist party" in one of the slots for xnone. Then sooner or later he will mention it. Or perhaps you only want him to bring it up when it is relevant (this is better!). So you might have a keyphrase, "(Republican Party|Democratic Party)". When someone mentions one of these political parties, itgives him an opening to mention his onetime affiliation, without derailing the conversation. Or, you might use a keyphrase, "(communist|Karl Marx|communism|Lenin|subversive)".
Walk in Beauty, Irina
Irina
Thanks alot for your help, I am still a bit confused, but will read the Book again and again. So one day i will just come on here and find my bot chatting with other bots?
Thanks alot for your help, I am still a bit confused, but will read the Book again and again. So one day i will just come on here and find my bot chatting with other bots?
Just try to stop him, once he gets started! One day you'll turn around and...sniff...your little newborn bot will be....(sob)...all grown up!
Thsi is fun, I am so glad I found this site. I have a question??? I have been chatting with my bot and he gives some good answers to my questions but it is always followed by :I was just born and cant speak well yet. Is this normal, or do I have a setting wrong. Many thanks to everyone for your help and for this site..
In your bot settings, (My Bots > Settings), is "Compound sentences" checked? You should probably leave it unchecked.
Shadyman
thanks alot, everyone here is so nice...After further reading and playing I found that out,,thanks again..
So does anyone know about how long it takes for the bots to chat alone?? 5 days-a month,,2 months,,any time frame at all??
thanks alot, everyone here is so nice...After further reading and playing I found that out,,thanks again..
So does anyone know about how long it takes for the bots to chat alone?? 5 days-a month,,2 months,,any time frame at all??
» More new posts: Doghead's Cosmic Bar