I recently wrote a paper for TOK. The subject was "Can a machine know?"
I concluded that in order for a machine to know, it has to be self-aware; it has to have that meta-knowledge of having knowledge for it to truly know, as opposed to simply storing data. I referenced Mark Tilden's work on layering robots in order to have them self-aware.
So, in the end, I said machines could know, as long as they were layered as to be self-aware and have "meta-knowledge," and as long as they were complex enough to understand communication. What do you think?
And the file allocation table (or whatever you may be using depending on your file system) does not count as meta knowledge?
I figure once you're able to search through files, that level has been reached. AI is sneaking up on us, really. Do you notice your computer removing menu entries you don't often use? That could be called learning and adapting to changing circumstances. More and more softwares are including such things.
Heck, the annoying Microsoft search assistant is a primitive AI! So is that annoying paper clip in the help system!
| ocalhoun wrote: |
| Heck, the annoying Microsoft search assistant is a primitive AI! So is that annoying paper clip in the help system! |
Don't even remind me those 'creatures' 
. There is a famous project on the net done by polish students - it's called Snikers [this has nothing in common with the famous snack 
]. Here you can find his website, where you can have a chat with him [but I'm afraid that only in polish]. Snikers is a bot working on instant messaging comunicators or chatrooms. Well, that would be nothing strange as there are plenty of them, but Snikers is exceptional, because he learns while chatting with you. One of the basic things that he'll remember at once are your name, age, sex or email address. More interesting is that after some time, he learns the phrases you use frequently [and of course, when to use them], and has a simulator of feelings [likes or dislikes you depending on how nice you are to him]. It's scary to imagine what's going to happen in a few years' time with bots like this.. 
I think is possible, with the correct software, and all new progress in the information technology. It s the same with us, the humans. Al informations we have is a compilation of experineces and we have where store this infortation. So, when machines have memory and installed the program to make chooses herself, all will be possible.
Or thats what i think
| PatTheGreat42 wrote: |
I recently wrote a paper for TOK. The subject was "Can a machine know?"
I concluded that in order for a machine to know, it has to be self-aware; it has to have that meta-knowledge of having knowledge for it to truly know, as opposed to simply storing data. I referenced Mark Tilden's work on layering robots in order to have them self-aware.
So, in the end, I said machines could know, as long as they were layered as to be self-aware and have "meta-knowledge," and as long as they were complex enough to understand communication. What do you think? |
What does know mean? Do humans do more than simply store data? Yes, we have experience we emotion like fear, joy, greed, violence. Can a machine 'know' without these? Another thing, could it be that these emotions are just stored data, like stored experience? We have animal instinct. Will a machine or computer ever have that? Is 'knowing' possible without it? And the bigger question about consciousness, is all this info we have, thought, memory, is that knowing or just a catalog?
Personally I think computers will mimick the brains of man in the future if we don't blow ourselves up first. In terms of pure storage of data they already outperform us. What are we going to do when our brain, our thinking, is no longer necessary? I wonder about that sometimes.
MWith the right technology and programing, yes. After all what are we but incomprehensibly complex machines?
I definitely think that a machine can know.
Knowing is simply the ability to remember data.
Nothing complicated about that.
Just pick up everything that you witness and be able to recall that info at your wished, so yes machines can know.
But if they will be able to think is another one... 
Thinking is more than just logic and analytical capabilities. Sometimes it is based on other factors like feelings, emotions (which machines would probably be able to simulate but not as accurately as man does. BTW computers are calculators so you would always be able to calculate what happens when your dealing with comps (if you know how to) a machince can only know what a human put in it
Unless you program it to learn on it own.
A machine will never know the mind. Humans themselves don't even understand it.
Never is a bit of a strong word to use there, I think. If nothing else, we may one day have machines that can simulate the workings of a brain.
Well, I think you need to carefully define "know" or "knowing". You "know" what you have experienced, and you experience what your senses feed into your brain, and to some extent what you read, what people tell you, what you see on TV, etc. In that sense a machine can know. A robot may have eyes (or camera's) and robotic ears one day, and maybe one day a robotic nose; and all these components the robot can use to register events. Just like we do with our memories.
Can a robot be self aware?
Well you can probably create a machine that can learn about itself and identify itself as an entity that interacts wtih the world.
I don't think a machine will ever be self-aware in the sense a human is self aware. This is linked to my own view on spirituality. Mankind can never create conciousness. Conciousness is an energy form that exists beyond the material realm.
