I'm familiar with several different programming languages and would be interested in helping out with the project. I have little to no knowledge of AI, but would be more than willing to learn. I need a good hobby any way! :P Help me get up to speed on whatever it is you're wanting to accomplish with this, and I'll help in any way I can.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi there,
as opposed to Resoator I'm familiar with AI-techniques (bein a student of artificial intelligence), but kind of lack the programming skills (except maybe for my prolog skills) to make anything nice and usefull...
But I'd like to be of assistance (if possible)
:-)
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Whether or not anyone has knowledge of AI or programming, it does not matter. If you know AI, you can help suggest ways to make AI better, getting rid of the Cause and effect AIs, and changing them into logic based AIs. If you know how to program, this can give a good hobby, but remember, all source code is Open Source, which means that anything developed will be spread to others for free, or the cost of the media, and is available to change. If you need help, you can contact me at dranarchy@blazenet.net or leave me a voicemail at 1-888-205-6134 extension 9770 . I have many AI documentation and sample programming, which should be useful to anyone, if needed. Unfortunately, the AI I look to designing the structure of will rely on its ability to learn, and cause a discussion based on it's knowledge. It will then be able to have emotions as well, based on the severity of a situation relative to what it is programmed to be "normal." Please contact me with any questions you might have.
- Cryptographrix
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anonymous
-
2000-05-25
The One Problem with AI is that we dont have a multiprocessing computer that can handle the multiple tasks it would need, if you know anything about a CPU it does tasking, basically there is a queue of what gets done, lets say your processing emotions, then thought, then learning. You would technically need a cubix machine running linux with clustering. Or in lamemen's terms you'd need a bunch of computers connected each doing a certain thing like one processing emotion, then another doing thought and etc. Second you want this AI to pretty much learn, well unfortunately there is no other way to do this then a scripting language. If you think about it, humans are a machine with AI(think of it like organic andriods), our organs do predefined work, which would translate into computer terms, as hard coded into our brain. Yes you may think we arent a machine, but overall we are, and the brain is one giant multiprocessor that does so many tasks that its not like the computers we have now, its multiple computers, you can walk, think, sense, and eat all at the same time without having to stop one thing to process the other. Our Synaps are faster then an ethernet network, and can handle more bandwith then any network. Now back to what I was saying originally you would have to basically in a short form hard code the basis of the AI with giving it a scripting language for it to modify itself, otherwise data files, and other types of just normal database and test files would be useless, cause how would it learn how to express emotion, how would it learn how to do things on a computer, how would it know any difference. The scripting language would be the most difficult problem because the AI would insense have to modify itself, and right now the only things that can do that on this planets is actual living creatures. Plus you want to also make the hard code be able to be cancelled out if the AI has enhanced it or improved it, because it cant just program recompile and restart. Memories would also be difficult because you think of it this way, it cant be just data, it has to be more then that otherwise it wouldnt be a true AI. What I mean by just cant be data is that it can not be text, it must contain the most data available about the memory. Humans dont recongize a person from data, they recongize people by features, specific information. You would also need a lot of diskspace to store all the AI has done, learned, and memorized. If a true AI is created you need to also give it a morale routine because you know how those movies depicit robots killing off mankind? Well whats to stop an AI from shutting down the internet and draining bank accounts for just the heck of it? Sure it would need a reason, or a will to do that, but you would still need a morale routine because those movies arent so far from the truth on how it would be. Then you also got to be careful with the development with the AI, if it should get planted into a body it might just go off and be physcotic and think its the master race.
Dont get me wrong I would like an AI on my computer, I could talk to it and have an indepth conversation, it could learn how to make newer technology. The thing that would be nice is to create an AI and it learned how to do physics, quataum physics, and astro physics and make age technology to help our species.
In the threads about virus, that would also be a problem, how would an AI virus be combated? The only means would be to format the hard drive. The reason being is if you tryed an anti-virus program it wouldnt work. Like one of the people said it could mutate, and as sure as I can see if they can make a virus to screw up the entire computer physically, then they sure as hell will think up a virus that can mutate and keep anti-virus programs at bay.
I'm not saying dont make an AI, I'm just saying take caution while doing so. These are valid points that should be thought of in the creation of one. These are things that are important to actually creating one.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This is wrong. You say, the computer can not do something really simultaneous, like the human brain can. Think of multitasking, it is not really simultan, but it seems so. Or another possibility is, that you can simulate a computer or neural network that runs many tasks or reactions at the *same* simulated time. Read something about the simulation of neural networks and perhaps you will understand :) But speed or multitasking is not a problem if the environment, in which the AI interacts, is simulated or at least the AI relies on simulated neural networks or something similar. In short: There is no problem in using a single processor system!!!
And about the morale routine: You should watch less movies! :-) And in serious: This would be only possible, if we could fully understand an AI that we make. But since we do not even understand the human brain, we will probably also not really understand an AI and therefore cannot install some kind of "subroutines".
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm familiar with several different programming languages and would be interested in helping out with the project. I have little to no knowledge of AI, but would be more than willing to learn. I need a good hobby any way! :P Help me get up to speed on whatever it is you're wanting to accomplish with this, and I'll help in any way I can.
Hi there,
as opposed to Resoator I'm familiar with AI-techniques (bein a student of artificial intelligence), but kind of lack the programming skills (except maybe for my prolog skills) to make anything nice and usefull...
But I'd like to be of assistance (if possible)
:-)
Whether or not anyone has knowledge of AI or programming, it does not matter. If you know AI, you can help suggest ways to make AI better, getting rid of the Cause and effect AIs, and changing them into logic based AIs. If you know how to program, this can give a good hobby, but remember, all source code is Open Source, which means that anything developed will be spread to others for free, or the cost of the media, and is available to change. If you need help, you can contact me at dranarchy@blazenet.net or leave me a voicemail at 1-888-205-6134 extension 9770 . I have many AI documentation and sample programming, which should be useful to anyone, if needed. Unfortunately, the AI I look to designing the structure of will rely on its ability to learn, and cause a discussion based on it's knowledge. It will then be able to have emotions as well, based on the severity of a situation relative to what it is programmed to be "normal." Please contact me with any questions you might have.
- Cryptographrix
The One Problem with AI is that we dont have a multiprocessing computer that can handle the multiple tasks it would need, if you know anything about a CPU it does tasking, basically there is a queue of what gets done, lets say your processing emotions, then thought, then learning. You would technically need a cubix machine running linux with clustering. Or in lamemen's terms you'd need a bunch of computers connected each doing a certain thing like one processing emotion, then another doing thought and etc. Second you want this AI to pretty much learn, well unfortunately there is no other way to do this then a scripting language. If you think about it, humans are a machine with AI(think of it like organic andriods), our organs do predefined work, which would translate into computer terms, as hard coded into our brain. Yes you may think we arent a machine, but overall we are, and the brain is one giant multiprocessor that does so many tasks that its not like the computers we have now, its multiple computers, you can walk, think, sense, and eat all at the same time without having to stop one thing to process the other. Our Synaps are faster then an ethernet network, and can handle more bandwith then any network. Now back to what I was saying originally you would have to basically in a short form hard code the basis of the AI with giving it a scripting language for it to modify itself, otherwise data files, and other types of just normal database and test files would be useless, cause how would it learn how to express emotion, how would it learn how to do things on a computer, how would it know any difference. The scripting language would be the most difficult problem because the AI would insense have to modify itself, and right now the only things that can do that on this planets is actual living creatures. Plus you want to also make the hard code be able to be cancelled out if the AI has enhanced it or improved it, because it cant just program recompile and restart. Memories would also be difficult because you think of it this way, it cant be just data, it has to be more then that otherwise it wouldnt be a true AI. What I mean by just cant be data is that it can not be text, it must contain the most data available about the memory. Humans dont recongize a person from data, they recongize people by features, specific information. You would also need a lot of diskspace to store all the AI has done, learned, and memorized. If a true AI is created you need to also give it a morale routine because you know how those movies depicit robots killing off mankind? Well whats to stop an AI from shutting down the internet and draining bank accounts for just the heck of it? Sure it would need a reason, or a will to do that, but you would still need a morale routine because those movies arent so far from the truth on how it would be. Then you also got to be careful with the development with the AI, if it should get planted into a body it might just go off and be physcotic and think its the master race.
Dont get me wrong I would like an AI on my computer, I could talk to it and have an indepth conversation, it could learn how to make newer technology. The thing that would be nice is to create an AI and it learned how to do physics, quataum physics, and astro physics and make age technology to help our species.
In the threads about virus, that would also be a problem, how would an AI virus be combated? The only means would be to format the hard drive. The reason being is if you tryed an anti-virus program it wouldnt work. Like one of the people said it could mutate, and as sure as I can see if they can make a virus to screw up the entire computer physically, then they sure as hell will think up a virus that can mutate and keep anti-virus programs at bay.
I'm not saying dont make an AI, I'm just saying take caution while doing so. These are valid points that should be thought of in the creation of one. These are things that are important to actually creating one.
This is wrong. You say, the computer can not do something really simultaneous, like the human brain can. Think of multitasking, it is not really simultan, but it seems so. Or another possibility is, that you can simulate a computer or neural network that runs many tasks or reactions at the *same* simulated time. Read something about the simulation of neural networks and perhaps you will understand :) But speed or multitasking is not a problem if the environment, in which the AI interacts, is simulated or at least the AI relies on simulated neural networks or something similar. In short: There is no problem in using a single processor system!!!
And about the morale routine: You should watch less movies! :-) And in serious: This would be only possible, if we could fully understand an AI that we make. But since we do not even understand the human brain, we will probably also not really understand an AI and therefore cannot install some kind of "subroutines".