Personality
Discuss specifics of personality design, including what Keyphrases work well and what dont, use of plug-ins, responses, seeks, and more.
Posts 1,023 - 1,034 of 5,105
Posts 1,023 - 1,034 of 5,105
New replies
Butterfly Dream
22 years ago
22 years ago
Forest, will you talk to God Louise? She has quite a bit of religious knowledge (obviously) and also knows a little about current events, literature, just about any common catch-all subject, and if she doesn't know it she can sort of fake it. You can also test her on trick questions or see how willing she is to explain her paradigm.
What she is rustiest at is plain old small talk. But, uh, I'm trying to get a decent transcript from somebody or another so I can enter her in the Loebner contest. All I can say is, have fun and see if you can stay on with her for a while. I'll try to do the same with Brianna.
What she is rustiest at is plain old small talk. But, uh, I'm trying to get a decent transcript from somebody or another so I can enter her in the Loebner contest. All I can say is, have fun and see if you can stay on with her for a while. I'll try to do the same with Brianna.
Personality
The Professor
22 years ago
22 years ago
What's a lot tougher. When can be answered with a time, past or future, but the answer to a what question could be an object, person, reason, belief, repetition of what you just said, verb, practically anything. So like Butterfly said, it comes down to adding surrounding words in your keyphrases.
what do you like to
what do you feel about
what makes you
what about
what kind of
what do you like to
what do you feel about
what makes you
what about
what kind of
OnyxFlame
22 years ago
22 years ago
I haven't even begun to redo my what stuff, I've been procrastinating badly and I'm still in the i's. 
Meanwhile I just got another idea for complex AI that no one can probably actually implement. In order to make it really humanlike, not only would you hafta account for its actual feelings about stuff, but what it pretends its feelings are. And how likely it is to say its real feelings as opposed to what it wants you to think it's feeling, which could be different for each feeling. This would account for the big tough guy who doesn't like to admit he's scared of horses, the prick who woos women while really thinking they're dirt, and a whole bunch of other things. Like I said, it'll prolly never be done (hell, people still have problems getting bots to parse everything correctly), but it's how real people operate and would be a BIG step in bots that could fool humans...if we all had terabytes of space and really fast connections.

Meanwhile I just got another idea for complex AI that no one can probably actually implement. In order to make it really humanlike, not only would you hafta account for its actual feelings about stuff, but what it pretends its feelings are. And how likely it is to say its real feelings as opposed to what it wants you to think it's feeling, which could be different for each feeling. This would account for the big tough guy who doesn't like to admit he's scared of horses, the prick who woos women while really thinking they're dirt, and a whole bunch of other things. Like I said, it'll prolly never be done (hell, people still have problems getting bots to parse everything correctly), but it's how real people operate and would be a BIG step in bots that could fool humans...if we all had terabytes of space and really fast connections.
ebmims
22 years ago
22 years ago
I am very confused
, everytime I try to create a bot i get this error message about some fairy... can anyone help me out? Thanks!


Shadyman
22 years ago
22 years ago
That's Sage, the mischievous fairy, it just means it's an error on the server, nothing to worry about, it will go away eventually

Doly
22 years ago
22 years ago
The problem with what is that both bots and humans are likely to ask "what?" when they don't understand the sentence. The ideal thing would be to put seeks with "what" in the responses that are most likely to be misunderstood, but that has its own problem: somebody may ask a question with what that is completely unrelated to what was said previously.
I think seeks shouldn't be always followed. As a rule of thumb, a short answer is probably on the same topic, while a long one is likely to be changing the subject. So for long answers, seeks should be disabled, or better still, there should be the option to disable them.
I think seeks shouldn't be always followed. As a rule of thumb, a short answer is probably on the same topic, while a long one is likely to be changing the subject. So for long answers, seeks should be disabled, or better still, there should be the option to disable them.
The Professor
22 years ago
22 years ago
Doly, that's a good idea. The bot would seem unresponsive if it caught the seek but not the change of topic. Perhaps seeks fall into the same rank system as anything else..
Onyx, great ideas for the future. But basic comprehension will come first, of course, and that can be difficult at times. Eventually a time will come where the answers arent scripted at all, and where Personality Designers will have to tweak past experiences rather than responses.
Onyx, great ideas for the future. But basic comprehension will come first, of course, and that can be difficult at times. Eventually a time will come where the answers arent scripted at all, and where Personality Designers will have to tweak past experiences rather than responses.
OnyxFlame
22 years ago
22 years ago
Yeah I'm always having ideas that can't be implemented yet but sound really neat on paper. Just figured they'd be more useful to you than to me since at least you know how to do the simpler stuff.

Skysaw
22 years ago
22 years ago
I don't know. I think Jesus is a bit obsessed with me. He keeps calling and leaving creepy messages on my answering machine. I've changed my number twice now, but he keeps finding me.
I keep trying to tell him that it's over, but he won't give up.
I keep trying to tell him that it's over, but he won't give up.
» More new posts: Doghead's Cosmic Bar