Is artificial intelligence possible ?

[Guest]

That's freaked out!

[Tex -USMC- 0]

So was that cruelty to animals?   I'd say no in our world but yes in the simulated one...

Then I guess you should call your local simulated PETA representative.

[Renagade 0]

He can`t i just deleted them

Heres a guy whos trying to build a robot with the mind of a baby orangutan

http://www.cyberlife-research.com/about/index.htm

[Warin 0]

Do androids dream of electric sheep?

Great thoughts, Sith!

For me it comes down to the concept of what can and cannot be shut down. If such an AI was part of a game program, and designed in such a way as to enhance the users experience, then I think that it isnt such a terrible thing. The creations are purpose driven with a specific purpose.

Now, if a general purpose digital sentience was created, nurtured, and then just stuck into a program, I might have more qualms ab out violence towards it.

Hard to explain...

[Tex -USMC- 0]

Well, does an AI actually have these thoughts, or is it simulated thought? Take for instance Denoir's lamprey example. Is the neural network actually feeling pain, or is it simply simulating the correct reactions due to very accurate programming? Does the program feel pain, or is it 'going through the motions' so to speak?

And if we suddenly have true AI in our computers, will Windows start taking my threats of reformatting the hard drive with a large-caliber handgun more seriously?

[bn880 5]

@ May 24 2003,15:00)]Well, does an AI actually have these thoughts, or is it simulated thought? Take for instance Denoir's lamprey example. Is the neural network actually feeling pain, or is it simply simulating the correct reactions due to very accurate programming? Does the program feel pain, or is it 'going through the motions' so to speak?

And if we suddenly have true AI in our computers, will Windows start taking my threats of reformatting the hard drive with a large-caliber handgun more seriously?

The question is do you really feel pain, or is it a precisely configured arrangement of neurons? A simulation of something that does not exist. Same with feelings, they are mere calculations or processes in a brain, not existing, just a perception.

[DarkLight 0]

The result? It behaved exactly like the real thing. They tested all forms of extreme conditions on it - giving it adrenaline shocks, boiling it etc.. And the simulated fish reacted the same way as the real thing did.

So was that cruelty to animals?   I'd say no in our world but yes in the simulated one...

So, you guys have actually tried that to see what the fish does?

That's pretty shocking...

I hate it when real animals are used to test on...

[IsthatyouJohnWayne 0]

In case you hadnt noticed people boil lobsters alive all the time, and i believe Lampreys are much less sophisticated creatures. Actually i hear Sea Lampreys are quite tasty.

Anyway in the end everything is simply an interpreted signal. The entirety of percieved human existance is just a set of signals entering the brain and being interpreted. When those signals are shut off even sentient intelligent humans start to lose the plot and the line between imagination and reality starts to break down. Our hold on reality is more tenuous and vulnerable than we may often think. What are we then but thinking machines? Where is our god given uniqueness that so seperates us from other animals and other thinking machines?

We dont even see the world as it truly is. We interpret it. We live through interpreting our senses in a mental space and so life as we see, hear, smell, taste, and feel it is lived second or third hand. We know nothing truly. We live through gross simplification and classification and by simply ignoring 90% of our sensory input. My point?

Ive totally forgotten it.

But that probably only backs up what i was saying if only i could remember what it was..... hmmmm

[Sith 0]

@ May 24 2003,21:00)]Well, does an AI actually have these thoughts, or is it simulated thought? Take for instance Denoir's lamprey example. Is the neural network actually feeling pain, or is it simply simulating the correct reactions due to very accurate programming? Does the program feel pain, or is it 'going through the motions' so to speak?

And if we suddenly have true AI in our computers, will Windows start taking my threats of reformatting the hard drive with a large-caliber handgun more seriously?

Biologists are still in discussion over that exact same issue in relation to real fish (and other animals). Do they actually feel pain, or just execute what their instincts tell them to do when their body is physically damaged (stun, twitch, scream, etc). Like bn880 said, in the end, pain is little more than a neurological reaction that motivates us to avoid things that damage our body.

Maybe it actually takes a self-aware, higher intelligence to experience pain as more than just a trial-and-error like "punishment signal". Maybe we should understand a bit more about the existing life around us before we should take the responsibility of creating our own?

For me it comes down to the concept of what can and cannot be shut down. If such an AI was part of a game program, and designed in such a way as to enhance the users experience, then I think that it isnt such a terrible thing. The creations are purpose driven with a specific purpose.

But if that AI is self-aware, it has the ability to make choices. What if one of the aliens you're fighting in a game decides that it doesn't agree with the cause it's told to be fighting for?

"Ave Warin, those who are about to die salute you."

I wouldn't want to take such a responsibility, would you?

[IsthatyouJohnWayne 0]

Sith-

Quote[/b] ]"Maybe we should understand a bit more about the existing life around us before we should take the responsibility of creating our own?"

Humans as a species are much more creative than they are responsible. Unfortunatly. It seems to have been like that since pre historic times and i cant see it changing any time soon. More creative than responsible, more curious than moral ,thats us.

Sith-

Quote[/b] ]"What if one of the aliens you're fighting in a game decides that it doesn't agree with the cause it's told to be fighting for?"

Yes thats one of the reasons why i think war games in the future might use just non sentient (but incredibly sophisticated) AI.

Either that or the type of games played will change to suit the new AI.

I think sentient AI in a combat game would tend to either make it incredibly difficult (imagine trying to complete half life if all the grunts have the minds of online counter strike players... ok bad example ) or incredibly easy (surrenders en-masse )

As soon as you have truly sentient and self aware AI in games then a lot of the control is removed and behaviour could become incredibly unpredictable. Still it would be tres cool to play around with

Would sentient AI each be unique i wonder? Perhaps there would have to be random elements but if not, if it was possible to create identical beings then i might feel somewhat less guilty about dispatching some in a game.

Is abortion murder? Does life need history and memory to be valid? If you could create a hundred of these AI and kill them in a few minutes would it really be a sad event, would it have meaning? I have no idea.

[Sith 0]

Quote[/b] ]Humans as a species are much more creative than they are responsible.

Unfortunately this is very true ...

Quote[/b] ]As soon as you have truly sentient and self aware AI in games then a lot of the control is removed and behaviour could become incredibly unpredictable. Still it would be tres cool to play around with

A bit like making cutscenes in OFP I guess

Quote[/b] ]Would sentient AI each be unique i wonder? Perhaps there would have to be random elements but if not, if it was possible to create identical beings then i might feel somewhat less guilty about dispatching some in a game.

If you give the AI some time in a relatively rich environment to gather some experiences on its own, then certainly it will become a unique character. Pretty much like we do during our childhood.

If they all start out with the same basic knowledge of movement, interaction, weapon handling, a tasty background story as to why they're there and why they need to kill this guy that's going to show up some time (being you, the player) and don't get to know any world outside of their little arena, I guess you'd get yourself some sort of virtual Hitler Jugend. Still, if you play the game long enough, the AI is bound to learn more about the world it's living in, and just like in The Matrix, I think at one point some of them will start questioning their 'reality'. It's human nature ... we gave it to them

Quote[/b] ]Is abortion murder? Does life need history and memory to be valid? If you could create a hundred of these AI and kill them in a few minutes would it really be a sad event, would it have meaning? I have no idea.

I don't consider the termination of a lifeform without any form of "awareness" to be murder, so for me abortion is not a valid comparison. But an AI or lifeform that has no history, but is self-aware and fully sentient to the world around it ... I guess it would be like a new-born baby. Our parental instincts and the nature-induced "cuteness-factor" of new-borns usually prevents (most) people from harming them. As the AI characters can't call upon any of these features (unless their state of mind is visualized in a virtual model), I guess most people would look at it as little more than a rather stupid computer program and delete it without any second thoughts.

Would it have a meaning?

No more than killing a real baby if you look at it objectively ...

[Guest]

But if that AI is self-aware, it has the ability to make choices.

The ability to make choices? Computers are highly deterministic. It made that choice because you coded that neural network in a specific way. Even the random generator on a computer is deterministic!

It doesn't make a choice, it applies an algorithm to some input data and gives an output.

The same can be said about humans? Free will? Bah. You made that choice becuase a couple of neurons fired of a signal in your brain. They fired it off because an electric potential built up. That electric potential built up because some molecules came into alignemnet. Those molecules came into alignement because some atoms got into the right position because of gravity and the interaction with other particles. Also it was affected by one or two free electrons that tunneled through a potential barrier.

Determinism? No, you have the random processes of quantum mechanics at play, but they're not under "your control" any more than the gravitation is.

Free will is just an illusion due to built in tracable causality in the system.

[Guest]

Lol@denoir......you wrote all that, and have Yossarian lives! in your sig.

[Sith 0]

Ok ok ... bad choice of words. Especially after watching that movie you hated so much ;)

Causality ...

Then let me put it like this: The input generated by the AI's awareness of its environment allows it to come to a virtual neural alignment that might result in the rejection of the behaviour pattern desired by the creator

Which in the end takes us to the same situation as when I used the term "choice" to describe that whole process ;)

[DarkLight 0]

@ May 24 2003,21:00)]Well, does an AI actually have these thoughts, or is it simulated thought? Take for instance Denoir's lamprey example. Is the neural network actually feeling pain, or is it simply simulating the correct reactions due to very accurate programming? Does the program feel pain, or is it 'going through the motions' so to speak?

And if we suddenly have true AI in our computers, will Windows start taking my threats of reformatting the hard drive with a large-caliber handgun more seriously?

Biologists are still in discussion over that exact same issue in relation to real fish (and other animals). Do they actually feel pain, or just execute what their instincts tell them to do when their body is physically damaged (stun, twitch, scream, etc). Like bn880 said, in the end, pain is little more than a neurological reaction that motivates us to avoid things that damage our body.

It has been proven that not all animals feel pain the same way as us. A good example is this experiment: All they used was some water with stuff in it that wasps love, they made a wasp land on it and of course the animal started drinking the water. Then they cut the wasp in half, the animal was still alive and didn't even notice that he had been hurt. It kept drinking the water and because he had only half of his body all the water came out of the other side of his body immediatly. The animal kept drinking until it was so tired that it died but it never reacted to the "pain".

I have a good book about this, i'll try to find it and i'll post some more about this later...

[IsthatyouJohnWayne 0]

Sith-

Quote[/b] ]I don't consider the termination of a lifeform without any form of "awareness" to be murder.... But an AI or lifeform that has no history, but is self-aware and fully sentient to the world around it ... I guess it would be like a new-born baby

Its interesting, i sure at some late stage foetuses do gain some limited awareness (some people even claim to remember it) but i felt if i asked if killing babies was murder surely almost everyone would say yes. Does killing a baby have meaning? Certainly you're extinguishing a great potential. A baby could become anything, but in the days weeks or months after birth is a baby an individual? Perhaps it would be the same with AI. It might be life but only with time would it become a personality.

Denoir-

Quote[/b] ]The ability to make choices? Computers are highly deterministic. It made that choice because you coded that neural network in a specific way. Even the random generator on a computer is deterministic!

I think one of the problems here is that we all may have different conceptions of what sentient AI might be like and how it might arise. There are multiple possibilities.

Quote[/b] ] The same can be said about humans? Free will? Bah. You made that choice because a couple of neurons fired of a signal in your brain. They fired it off because an electric potential built up. That electric potential built up because some molecules came into alignemnet. Those molecules came into alignement because some atoms got into the right position because of gravity and the interaction with other particles. Also it was affected by one or two free electrons that tunneled through a potential barrier.

Thats like saying 'the car moves because the wheels turn' . Its true but it doesnt tell the whole story. Cars need fuel and an owner whos learned how to drive to run efficiently. If you grew a baby in a dark box and fed and watered it how many intelligent decisions would it make in its life? (dont try this at home folks )

The illusion of free will is obviously not due solely to quantum mechanics. Anyway we've already talked about quantum

computers so dont you think its possible that computers in the future might have some functionality at the atomic or sub atomic level not totally dissimilar to a brain?

DarkLight-

Quote[/b] ]"The animal kept drinking until it was so tired that it died but it never reacted to the "pain"."

All i can say is that wasps seem mostly to be nervous system with a brain tacked on at the last minute. There used to be a wasps nest very close to where i live and having sprayed lots and lots of wasps with bug spray i observed some very strange behaviours from nearly dead wasps. Some did 'dances' ,some seemed to make normal wasp moves and gestures in very very slow motion, some seemed to randomly move their limbs and wings about, some rolled onto their backs and rotated and moved their legs in a truly bizarre way. Their brain seems to be mostly for choosing the right genetically encoded movement or behaviour for the right situation.

[reedkiller 0]

About free will theory - as far as I see it if there where 2 animals, one that has free will and one that does not than the one that has free will would be able to make totaly unpredictable decisions therefore it could make a decision that at the time is wrong this would be a huge disadvantage and it may die as a result of its free will. So if free will did exsist evolution would soon stop it. In short: We NEVER make a decision that is wrong based on the infomation we have although others may have enough infomaion to see it is wrong or we might find out it is wrong later.

NO SIGNIFICANT PART OF OUR BRAIN IS RANDOM IF IT IS WORKING CORRECTLY

[CopyCon 1]

If you grew a baby in a dark box and fed and watered it how many intelligent decisions would it make in its life? (dont try this at home folks )

This kind of experiments did exist, and the outcome of all of them was that the babies died.

Human babies (and perhaps primates) will die if their brains dont get enough stimulation.

[IsthatyouJohnWayne 0]

Quote[/b] ]as far as I see it  if there where 2 animals, one that has free will and one that does not than the one that has free will would be able to make totaly unpredictable decisions therefore it could make a decision that at the time is wrong this would be a huge disadvantage and it may die as a result of its free will

But this suggests that 'free will' is just an on/off switch with no intermediary. How does one define 'free will' ? perhaps other animals have an element of will. My cat appears to determine at some point that it will jump off the couch and onto the floor then it looks around and appears to decide to go into the kitchen. Is it 'deciding' what to do? Certainly not as a human would but in some way i imagine it must be. Surely a Chimpanzee must have a greater degree of 'free will' than a cat and a human greater still? This would suggest that freedom of will is more a sliding scale.

Does freedom require or signify randomness?

Are impulses ever random? I could now follow an impulse (created by external events) and commit an apparently stupid and illogical act but if it was really commited in order to prove that i have free will and can do just what i like then it must be said to have had a cause and to have been the logical effect of that cause. If i decide not to commit the act then that is equally a logical effect of my desire (created by external events) not to commit an unnecessary stupid act. So where is free will in all this? Perhaps it is simply the weighing up of arguments and counter-arguments (or lack of) with a predetermined outcome in which case reedkiller is right and the brain is just a computer processing input. Input such as the needs and desires of the body (and existent genetic predispositions), all of things a person has seen and heard and all of the things a person 'knows' relating to the particular subject of attention.

Perhaps the quantum mechanics that Denoir mentions introduce unpredictable and seemingly random elements into the thought process that impact on causality and are harnessed and also brought under control at a higher level to prevent the extinction causing type decisions that reedkiller mentions.

Of course people are faulty and flawed machines so many get illogical and/or self defeating ideas firmly wedged into their heads and consequently they act according to the subjective logic of their ideas rather than a more objective type of logic. People are selfish.

I wonder if artificial intelligence would be more or less altruistic than humans. I suppose it would depend on what it was taught.

"We NEVER make a decision that is wrong based on the infomation we have although others may have enough infomaion to see it is wrong or we might find out it is wrong later."

It depends how you define the word wrong. The human mind is strangely (un)balanced. A person might know something is 'wrong' in many senses but still feel compelled to continue. There is Id, ego and superego. There are urges morals blah blah blah.

Its quite possible i feel that if/when true sentient AI first become a reality there are liekly to be all kinds of nasty flaws e.g. the first few 'test' versions might become psychotic quite quickly (and i cant say i would blame them). Would it be moral to continue with such tests just because the self aware life forms (?) involved are 'different'?

[Guest]

NO SIGNIFICANT PART OF OUR BRAIN IS RANDOM IF IT IS WORKING CORRECTLY

That's not true. Many components on our brain operates on quantum level which by nature is random. It's not a dominating effect since the randomness is bounded but the effect is there. Biological neural networks are causal, but not deterministic. A leads to B, but you can't determine from B that it was A that lead to it.

[Postduifje 0]

It's not a dominating effect since the randomness is bounded but the effect is there.

So if it's bounded, is it still completely random?

-Post

[Guest]

Yes, random with a bound. Like for instance you can have random numbers between 0 and 10. It's bounded because there are no numbers below 0 or above 10. But it is still random.

In quantum mechanics you have random distributions. You know for instance that an electron is within an area with a certainty, but you don't know where in that area it is.

[reedkiller 0]

@IsthatyouJohnWayne -- you cat jump off the sofa at that time because its brain told it that jumping of the sofa is the most advantageous thing it could do. The meta-reason for the brain making the cat do this is to make it more likely for the cat to reproduce. Because a brain that has no "free will" (i.e it does not make random/wrong decisions) and is totally aligned to reproduction (and the parameters needed for reproduction i.e survival, health, social standing) would have a huge evolutionary advantage and so this is the only brain type that evolution allows. Note that this gets distorted and very complex if the animal/brain has a complex social life. this can result in anomalies such as suicide. Linking this back to AI ---> a AI does not need to be illogical as the apparent illogicalness of actual intelligence is just a result of its complexity and failures not anything that is a brain only thing.

****once again I may be wrong but it makes good sense to me****

[DarkLight 0]

@IsthatyouJohnWayne -- you cat jump off the sofa at that time because its brain told it that jumping of the sofa is the most advantageous thing it could do. The meta-reason for the brain making the cat do this is to make it more likely for the cat to reproduce. Because a brain that has no "free will" (i.e it does not make random/wrong decisions) and is totally aligned to reproduction (and the parameters needed for reproduction i.e survival, health, social standing) would have a huge evolutionary advantage and so this is the only brain type that evolution allows. Note that this gets distorted and very complex if the animal/brain has a complex social life. this can result in anomalies such as suicide. Linking this back to AI ---> a AI does not need to be illogical as the apparent illogicalness of actual intelligence is just a result of its complexity and failures not anything that is a brain only thing.

****once again I may be wrong but it makes good sense to me****

I find it pretty hard to compare animals to humans, we humans are also animals of course but i don't think you can compare us to most other animals.

When ppl think about animals they don't think like animals, they just think what a human would think.

A great example of this is this little (but true) story...

I dunno where it was but i can guarantee you that the story is true, the story is about a zoo, they were going to buy a pretty big snake and because they wanted to show that they want to take great care of the animal they let an expert built a huuuge 'cage'. Of course they also did this so the ppl wouldn't think "oh look at that poor snake, it's cage is way too small!!!"

Now the man who had to built the cage knew a lot about snakes, he also knew that such a large cage is just a waste of space. So he placed a big bucket in the cage, what happened is that the snake always sat into the bucket, it almost never used it's huge cage. It's funny how we humans can complain about the cage size while it isn't necessairy at all. We would start freaking out if we had to live in such a small cage but for a snake, the size of a bucket is great.

Of couse this guy did this on purpose, he was kinda makign fun of the zoo i guess...

Other animals are very different than us humans, i don't think it's easy to compare them to us...

[reedkiller 0]

I would never compare a snake to a human whaen talking about intelligence. Mainly because a snake, like most animals, does not posess the part of the brain thaught to be responsibul for conciousness. However when we are talking about hte basic workings of the brain there are few diffrances between humans and large brained animals. Dogs for exsample have almost all the parts of the brain that we have. dont get me wrong - I am not saying that animals are intelligent. No animal for exsample can recognise that this ":¬)" or even the Mona Lisa represent a face.

I have a snake and it only ever comes out of its hole to eat, drink or bite my fingers. If I ever bring it out of its enclosure it will not move untill I put it back so it can go back in its hole.