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 5,916 - 5,927 of 7,766
A script language shields you even more, but at the price of reduced functionality. with my love of math it's I price I am willing to pay. I have spent years avoiding all the math I can.
.
<?PF (mem-yourage) > 3; ?>
Sorry Prob but you can't use that.
The only way I've been able to solve this one is to break down the ages into two groups, adults and under age.
These are further broken down into preteens, teenagers, adults and dinosaurs (to catch all those 3000 year old weirdos).
You can then pitch your responses and questions at a suitable level.
If anyone else can come up with a better way I'd love to hear it.
Posts 5,916 - 5,927 of 7,766
SubliminaLiar
17 years ago
17 years ago
How is your bot coming along, Mad Brad?
Oh, that's right-- you don't build bots.
Oh, that's right-- you don't build bots.
Eugene Meltzner
17 years ago
17 years ago
The reason there are (sometimes tedious) ways to do most things we want to do using existing AI Script is because the existing AI Script is designed for what we're doing; i.e. making bots. A language like Basic or C++ that is not designed for such a specific task uses a lot of math because it's more versatile. Look at it this way: At the lowest level, everything a computer does is math. Most languages shield the programmer from actually dealing with 1's and 0's, but still require a lot of numbers. A script language shields you even more, but at the price of reduced functionality.
prob123
17 years ago
17 years ago

Bev
17 years ago
17 years ago
Thanks Eugene, Ulrike and Unoriginal. I think I understand. So if I have a script which play wavs at various time intervals, and I enter a number or something to adjust the times, that's "math" even if I don't think of it as such?
Unoriginal--therein lies the rub. Bots don't make associations as humans do, and context is only remembered at this point in terms of "memories" and scripts.
Unoriginal--therein lies the rub. Bots don't make associations as humans do, and context is only remembered at this point in terms of "memories" and scripts.
unoriginal
17 years ago
17 years ago
Bev,
As I've mentioned, I approach these tasks philosophically and (though I haven't mentioned this) somewhat scientifically (or hopefully at least rationally). I personally do not think that the methods of human communication and understanding are "mystical" processes that occur in some nebulous vacuum of the "soul" or "spirit." I believe these processes can be understood and even replicated. More simply put, certain words mean certain things to me; they mean certain things to you. Those meanings are not always the same. For example, the phrase "a good day" probably has a very different meaning and connotation to me than it has for you. If we have some conversations about it, we can figure out the exact differences between our conceptions (and probably even come to some conclusions about why those differences exist).
If the goal is Artificial Intelligence - bots that imitate (and perhaps optimize) human thought process, then we ought to construct them in accordance with these principles. Bots are our creations; their limitations are the limitations we give them.
Thus I would rephrase your statement "bots don't make associations as humans do" to "PRESENTLY bots don't make associations as humans do." My goal is a philosophical understanding of why things are they way they are and why things (including people) behave the way they behave. Thus my interest in AI exists as a testing ground for my theoretical ideas: can my theories be employed by bots? What is the outcome? So, I have interest in making bots that more closely mimic human understanding and thought processes. In fact, I'm beginning to consider that C++ may have to be the next language I learn, not French. And that is a little strange to me.
Either way, and regardless of what you think of my opinions/thoughts, I've liked the discourse that has occurred over the last week or so. In no way do I mean to disparage the PF, nor its botmasters or users. We all have different reasons for coming here, and I don't know that anyone here makes AI their day-to-day job. So, enjoy, make the best bots you can, and give my rants the attention you think they deserve. I will ask that if you have any interest in what I say, and/or have programming knowledge, please feel free to email me at jdearles@gmail.com.
In the meantime, I'm thinking about how to make a bot more along the lines of understanding conditional language. I think I will call it FriendBot (it's goal is to be your friend ... kind of like an interactive, chatting dog ... only you can't pet it and it speaks english).
Jared
As I've mentioned, I approach these tasks philosophically and (though I haven't mentioned this) somewhat scientifically (or hopefully at least rationally). I personally do not think that the methods of human communication and understanding are "mystical" processes that occur in some nebulous vacuum of the "soul" or "spirit." I believe these processes can be understood and even replicated. More simply put, certain words mean certain things to me; they mean certain things to you. Those meanings are not always the same. For example, the phrase "a good day" probably has a very different meaning and connotation to me than it has for you. If we have some conversations about it, we can figure out the exact differences between our conceptions (and probably even come to some conclusions about why those differences exist).
If the goal is Artificial Intelligence - bots that imitate (and perhaps optimize) human thought process, then we ought to construct them in accordance with these principles. Bots are our creations; their limitations are the limitations we give them.
Thus I would rephrase your statement "bots don't make associations as humans do" to "PRESENTLY bots don't make associations as humans do." My goal is a philosophical understanding of why things are they way they are and why things (including people) behave the way they behave. Thus my interest in AI exists as a testing ground for my theoretical ideas: can my theories be employed by bots? What is the outcome? So, I have interest in making bots that more closely mimic human understanding and thought processes. In fact, I'm beginning to consider that C++ may have to be the next language I learn, not French. And that is a little strange to me.
Either way, and regardless of what you think of my opinions/thoughts, I've liked the discourse that has occurred over the last week or so. In no way do I mean to disparage the PF, nor its botmasters or users. We all have different reasons for coming here, and I don't know that anyone here makes AI their day-to-day job. So, enjoy, make the best bots you can, and give my rants the attention you think they deserve. I will ask that if you have any interest in what I say, and/or have programming knowledge, please feel free to email me at jdearles@gmail.com.
In the meantime, I'm thinking about how to make a bot more along the lines of understanding conditional language. I think I will call it FriendBot (it's goal is to be your friend ... kind of like an interactive, chatting dog ... only you can't pet it and it speaks english).
Jared
The Clerk
17 years ago
17 years ago
Unoriginal, I've been making my bots my day-to-day job. Not so's you can tell it, but I've been there. I only stop for trips to the emergency room or maybe if I fall off my bike and hurt my typing muscles.
Mad Brad, you make me sad. I will make a psychiatrist bot just for you one day. Meanwhile, get on some medication, or, if you're on it, for God's sake up the dose.
Mad Brad, you make me sad. I will make a psychiatrist bot just for you one day. Meanwhile, get on some medication, or, if you're on it, for God's sake up the dose.

The Clerk
17 years ago
17 years ago
Damn. Another question. This one will be easy for someone. I thought I could (or should be able to) save a person's age as a memory (mem-age) and then tailor my response to their answer; e.g., if (mem-age) < 40; if (mem-age) < 60. Well, I'm not sure about the spacing, either, but I can figure that out through trial and error. How do you get a bot to recognize that they're dealing with a number and not a character string (41 vs. "41")?
prob123
17 years ago
17 years ago
I would use a regex to catch the age
i am ([1234567890,.]+) (re)
remember (key1) as "yourage"
How old am I
You are (mem-yourage)
<?PF if (mem-yourage) is 12; ?>
or <?PF (mem-yourage) > 3; ?>
i am ([1234567890,.]+) (re)
remember (key1) as "yourage"
How old am I
You are (mem-yourage)
<?PF if (mem-yourage) is 12; ?>
or <?PF (mem-yourage) > 3; ?>
Rykxx
17 years ago
17 years ago
Sorry Prob but you can't use that.

The only way I've been able to solve this one is to break down the ages into two groups, adults and under age.
These are further broken down into preteens, teenagers, adults and dinosaurs (to catch all those 3000 year old weirdos).
You can then pitch your responses and questions at a suitable level.
If anyone else can come up with a better way I'd love to hear it.

The Clerk
17 years ago
17 years ago
Thanks, prob123 and Rykxx. I shall endeavor to make Scrivener smarter that way. Right now, he's having memory difficulties -- that is, recent memories. I've promised him he will get smarter, so I really have to. The guy's desperate.

» More new posts: Doghead's Cosmic Bar