Can you realy call Kitt a computer?

[Count Dracula]

How much longer before an AI as sophisticated as KITT and KARR become a reality? 5 years? 10?

And what AI is the most sophisticated (for the public, I am sure the military has the most advanced AI now) out thier? And is it free to download and tinker with?

[neps]

This topic has been debated constantly, you may want to search the board to read up on what has been said.

I can talk to my computer and it will reply. I can tell it to open things, I can tell it find information for me. All with only my voice. Car systems have navigation interfaces that work much in the same way. Other then telling the computer to do something and it actually doing it, I don't think there is anything more advance out there for public consumption. The main key for AI is for it to think for itself, nothing like that yet from what I know of.

[TT Snim]

Well, I have been tinkering with some AI now... Creatures 3... Look at this link untill I feel up to finding a better link...
http://alifelabs.creatureslife.net/Library/L1.htm

[knightimmortal]

Seriously, you are looking at 10-15 years minimum for that kind of sophistication. There are expert systems out there that mimic AI, but that is all.

KI

[TT Snim]

Ya' know what bothers me... AI. Thats right, them two letters bother me. Why not DI. If any thing becomes intelligent then it would be real, organic... or Degital. Hence Digital Intelligence. D.I.

[knightimmortal]

Because in some cases, digital, has nothing to do with it, and in effect, you are creating a simulated intelligence, so Artificial Intelligence is in fact the completely correct term.

KI

[Darknight]

You have my curiosity up Laura. Speak more about the cases in which "digital has nothing to do with it." I'm not even quite sure that I read that right.

DK

[knightimmortal]

Many aspects of AI are analog mixed with some digital, so it isn't fair to call it simply digital. Any further than that, and I would have to shoot the entire board.

KI

[Michael Pajaro]

How much longer before an AI as sophisticated as KITT and KARR become a reality? 5 years? 10?

You can put my prediction as closer to 100 years.

Many people associate computing power with intelligence. The more memory, the faster the chips, the smarter the computer is. Not true. In order to create artificial intelligence, we have to first understand what NATURAL intelligence is. The truth is that we barely have the slightest idea how he human brain works. Philosophers have been debating what it means "to think" for thousands of years. Modern medicine has examined the brain for maybe 100 years. Psychology has been looking at it for about a century as well. All of these fields are involved with "intelligence". Understanding neurons and electrical signals can explain how the brain works as an organ, but we have no clue how electrical signals can represent intelligence. And until we understand the theory of intelligence, we can't begin to recreate it.

In another perspective: The heart is a pretty important organ, but essentially it's just a basic pump. Doctors have spent 25-30 years trying to develop an artifical heart, and they still haven't created a truly reliable one. Now try to conceive of how much more complicated a brain is than a heart, and how much longer it will take to develop an artificial brain. I think we're still a LONG way from seeing a real KITT.

Mike

[knightimmortal]

*sigh*

I wish I could say something to legitimately disagree with Mike, but it might be construed as some sort of disrespect, rather than a defense of the AI position, so let's just say: there are people that have made leaps, bounds and a whole lot more understanding than your average ordinary everyday person believes, but more and more dangers are being found in the process. 5-10, but nobody is totally sure if it will be on the side of KITT, or more towards...Lore.

KI

[Count Dracula]

In another perspective: The heart is a pretty important organ, but essentially it's just a basic pump. Doctors have spent 25-30 years trying to develop an artifical heart, and they still haven't created a truly reliable one. Now try to conceive of how much more complicated a brain is than a heart, and how much longer it will take to develop an artificial brain. I think we're still a LONG way from seeing a real KITT.

Mike

Actually, they did create a artifical heart that stays in the chest rather then a big bulky machien. Unfortunatly it only lasts for a couple of years before they have to replace it.

[Michael Pajaro]

Actually, they did create a artifical heart that stays in the chest rather then a big bulky machien. Unfortunatly it only lasts for a couple of years before they have to replace it.

That's sort of my point. They're still not reliable. If we can't even build a reliable pump, how can we expect to build a reliable brain?

Mike

[knightimmortal]

Because we aren't trying to mechanically engineer a brain, we are using logic pathways to emulate brain activity and function?

KI

[Count Dracula]

knightimmortal, what kind of dangers are they finding with AI that makes them not sure if it will be good or evil?

And what is the most advance AI now in the public? If it is Alice, I will wait for a bit to download, last Alice I had was extreamly stupid, and this was from http://Alicebot.org

[knightimmortal]

Alice isn't AI, it is an expert system. An expert system is one that can be easily controlled by the programmer but isn't really intelligent.

And the problems are when you program in a certain number of variables, nobody knows where those variables will be taken when a something makes a choice.

Here is an example with people in real life. Just because I do something, take an action, make a statement or something that isn't agreeable with something else, I make a decision to try to do something about it, which in the end results in people thinking that I am Garthe, and that everything I do from there on out is Garthe like, or is bad in a ways. Therefore, I must be evil, and bad, and aggressive, and attack people even when I say Hi. And that is what can make an AI go really wrong in all theories.

KI

[Andy101]

Its a start... But i think its programmed to react to what you say... not to think for itself...

{ if ($1 == Hello ) /say Hi there... }

Like a mIRC bot...

[Count Dracula]

I programmed Alice before and good at programming, so if you say it is not real AI, then that means I know squat in programming AI.

I would love to learn, because the industry I want to be in some day is Robotics and Electrical Engieneering doing research for the government, maybe even try to build a real KITT (thought that will take a vary long time).

[knightimmortal]

See, here is where I am going to fall into sync with Mike. Nobody knows about AI programming. A select few know the in depth principles towards it, but AI is not just something that your average person knows about.

I have been in school for many years to even get a glimpse into what it is about. People who have been working in the field since the 60's are just barely getting their glimpses into getting deep into it.

Programming something to talk back to you is not AI. It is not even close to thinking, because that doesn't exist yet. So, sorry, but the harsh truth is that you don't know about AI programming.

On a brighter note, if you are good at programming, and truly want the dream of working for the government in those fields, keep on the physics, chemistry, and math classes. Be ready to take tons of classes in psychology, physiology, and even more in the physics and math departments, and don't give it up, and most importantly, don't fool yourself into thinking that you know something you don't, always keep researching, always keep learning.

KI
(I can see clearly now...what do you mean I have another two years until I get the next degree????? )

[Count Dracula]

See, here is where I am going to fall into sync with Mike. Nobody knows about AI programming. A select few know the in depth principles towards it, but AI is not just something that your average person knows about.

I have been in school for many years to even get a glimpse into what it is about. People who have been working in the field since the 60's are just barely getting their glimpses into getting deep into it.

Programming something to talk back to you is not AI. It is not even close to thinking, because that doesn't exist yet. So, sorry, but the harsh truth is that you don't know about AI programming.

On a brighter note, if you are good at programming, and truly want the dream of working for the government in those fields, keep on the physics, chemistry, and math classes. Be ready to take tons of classes in psychology, physiology, and even more in the physics and math departments, and don't give it up, and most importantly, don't fool yourself into thinking that you know something you don't, always keep researching, always keep learning.

KI
(I can see clearly now...what do you mean I have another two years until I get the next degree????? )

Thank you.

[knightynight]

No matter how soon KITT, or some reasonable facsimile, becomes a reality, I think we might all be in agreement that we should never neglect the responsibility that creating such a machine will invariably produce. Creating life, or even artificial life, presents us with moral and ethical problems. Such as, "If I build an intelligent indestructible car that happens to kill/maim/destroy someone/something is the car responsible or the creator?" "If someone steals my prototype and learns how to alter it so that it can kill/maim/destroy who is responsible?" "If these questions bring us into such uncharted territory that we know not how to deal with it should we even dare to proceed?" the list goes on. No matter who takes part in the creation of artificially intelligent entities they should address these issues and understand that artificial life just might be the next atomic bomb. I am not against the idea of AI, just against any moron who would blindingly attempt to play God without seriously considering the consequences. I seriously believe that many Knight Rider fans (and I don't intend to point fingers or to insult) find that the concept of creating the perfect pal/buddy/friend/partner a dream come true. I know I once wished that I could have a friend like KITT myself when I was a kid.

What I saw at that Creatures website above is that whoever wrote the articles for the site did not address any of the afore mentioned issues except for responding to "scaremongers who would try to tell us that robots will one day rule the world". They seemed to me to be scoffing at anyone who would show an ounce of careful consideration. I was incredibly disheartened.

[knightimmortal]

And a true AI developer does in fact bring those up, and considers them very, very carefully.

Thank you for stating it so well.

KI

[Darknight]

I'll take another approach to the matter. I don't worry so much about playing God now as I did before, for the following reasons. (Assuming you believe in one God who created all humans, this is for you. If you're not, then this argument wouldn't work for you.) God made humans in his own image, yet we are not deity. So, evidently, He made us after a form of himself, but not exactly as himself. Knowing that, I think perhaps that we humans have the ability to make something after a form of ourselves, but not completely human. Perhaps an AI would have the ability to think similarly, but would not be able to consider itself human, just as KITT could not truthfully consider himself human. Yet, he was close, and definitely capable of more complex thoughts than say, a chimpanzee. However, a chimpanzee, having a body similar to ours, can in some ways identify with us at a level closer to humanity.

For example, a chimp can feel attraction toward another chimp, because most of them are of the same genus, and more than likely, species. An AI might be able to come closer to feeling love than a chimp, but it would have to meet another AI before it could begin to do so. Even then, there would have to be some way for them to interact, and they must have the capability for interest and curiosity. They would each have to have some function that would make them eager to discover more about the other AI, and then they would need a function that operated something like what we call "wonder," "amazement," "infatuation" or being "starstruck" - basically, an internal function that would gauge the level of interest that one AI would have for another, and then enable the AI to make a choice as to whether the "risk" was too high, or if it should "go for it."

I believe there is a barrier that humandkind can not cross, however. That is, I believe that we can not take inorganic material and breathe a soul into it. I am aware of silicon based micro-organisms that have been engineered, but that's quite a bit different from AI. I believe we can possibly engineer something that has intelligence, and something that seems to have a living system, but we can not replicate all human physical functions and behaviors without making another human, using DNA. I hope all this made some sense.

DK

[TT Snim]

Darknight, I agree with you on this nearly compleatly. I am wondering, would it be posible for us to engage in a chat, or prihaps E-mail discusion, in relation to our theories? I have been studying AI for a short time. I not only think AI on a large level is posible, but it may be posible with out compleatly remodeling the human brain. This "True AI" as kitt is may be posible and I wish to be a part of it's construction.
And yes, to me this all not only made sense, but reinfoces my privious conclutions.

[KFCreator]

Check Popular Science magazine constantly for updates about AI. Actually, a few months back they ran a small article talking to an MIT researcher who is currently developing an AI program that they claim "has the intelligence of a 3 or 4 year old" and has the ability to learn and incorporate new ideas into its program on its own. And, there have been countless other small articles in that magazine remarking about other reserachers who were experimenting with AI programs and found that the programs did unexpected things that had not been pre-programmed, raising speculation that they can indeed learn. Also, biotechnology may advance the field of AI even faster by using organic components to creating mini "cyborgs." In fact, the next major leap in computing power is going to be "quantum computing," which will utilize a combination of organic tissue and cells and silicon and microchips to create machines that wil be able to perform more calculations per second than we can imagine. I place the "emanation" of AI's no more than 50 years from now.

[Darknight]

TT, can you specify just what it is that you think you agree with me on? I'm not sure I follow you completely.

DK