Computer Programs
X_Stickman
Not good enough for a custom title. Join Date: 2003-04-15 Member: 15533Members, Constellation
Not good enough for a custom title. Join Date: 2003-04-15 Member: 15533Members, Constellation
in Discussions
<div class="IPBDescription">Yeah... Kinda insane thoughts, here</div> I've been getting pretty much more and more insane recently, i guess. But recently, i've taken to feeling guilty and sad when a character in a game, or even a computer, gets hurt or killed.
But i started thinking about it, and thought that maybe i'm not so insane after all. Computers have the ability to store and process information, and given the proper stimulation/need, act upon that information. While i'm not suggesting that computers/programs at their current level are rivalling humans or even small rodents in their intelligence/sentience, they're not too far off "lesser" creatures, such as ants, insects etc...
They have a body and senses, in a way. AI for in-game characters even more so, as they can "see", "hear" and the more advanced AI can even decide the best course of action out of a list of different actions, which is all <b>we</b> do really, albeit we have a much larger list.
Basically... what i'm saying is, are we so sure we haven't created artificial life already? Just because it hasn't tried to enslave us (heh) or hasn't stood up and said "hello, i'm alive", doesn't mean we haven't (when was the last time you had a conversation with an ant).
One of the things that always stuck in my mind was a TV thing i saw ages ago. A man makes little walking robots as a hobby (like, ones that you can bend 4 legs on and it can still walk etc). He was demonstrating one, and he picked it up and held it on it's back, and it struggled to get away, then sort of hid in a corner when he put it down. He said he hadn't programmed it to do that.
It's most likely that he was either lying about the programming to gain attention, or he mis-interpreted it's walking as struggling and he just put it down at such an angle it ran to the corner, but it always stuck in my mind... maybe it was alive.
Anyway... That's my ramble over. I don't do much apart from sit in my room, and these things play on my mind. I'm sorry if this was discussed in that sentience thread, but i never read it, and now it seems to have vanished so i can't check.
Does anyone agree/disagree/think i'm insane? Is this already a pretty well know thing and i've just been living under a rock (likely) ?
But i started thinking about it, and thought that maybe i'm not so insane after all. Computers have the ability to store and process information, and given the proper stimulation/need, act upon that information. While i'm not suggesting that computers/programs at their current level are rivalling humans or even small rodents in their intelligence/sentience, they're not too far off "lesser" creatures, such as ants, insects etc...
They have a body and senses, in a way. AI for in-game characters even more so, as they can "see", "hear" and the more advanced AI can even decide the best course of action out of a list of different actions, which is all <b>we</b> do really, albeit we have a much larger list.
Basically... what i'm saying is, are we so sure we haven't created artificial life already? Just because it hasn't tried to enslave us (heh) or hasn't stood up and said "hello, i'm alive", doesn't mean we haven't (when was the last time you had a conversation with an ant).
One of the things that always stuck in my mind was a TV thing i saw ages ago. A man makes little walking robots as a hobby (like, ones that you can bend 4 legs on and it can still walk etc). He was demonstrating one, and he picked it up and held it on it's back, and it struggled to get away, then sort of hid in a corner when he put it down. He said he hadn't programmed it to do that.
It's most likely that he was either lying about the programming to gain attention, or he mis-interpreted it's walking as struggling and he just put it down at such an angle it ran to the corner, but it always stuck in my mind... maybe it was alive.
Anyway... That's my ramble over. I don't do much apart from sit in my room, and these things play on my mind. I'm sorry if this was discussed in that sentience thread, but i never read it, and now it seems to have vanished so i can't check.
Does anyone agree/disagree/think i'm insane? Is this already a pretty well know thing and i've just been living under a rock (likely) ?
Comments
Death is necessary to create better species , and things that don't evolve when pressured just deserve to die or become extinct. Game AIs are only created for your enjoyment , they've been solely designed to interact with humans , either obey them or challenge them , and in both case , entertain them.
Anthropomorphism is bad , defenseless creatures do not cry for mercy , they're here to help you vent your excess frustration. In HL , when a scientist is done healing you , it's always funny to slaughter him and watch his gibs flying all over the place. It doesn't make you a murderer... yes the AI felt the "pain" stimulus , it reacted to that by playing a scary .wav file , then changing itself into a corpse. If HL scientists say you're mad to hurt them , it's to emulate a realistic behavior.
Artificial life forms are just images of the real ones. They're simulations. Most bacteria are much more intelligent and faster to adapt than them , so I don't see the reason you should feel guilty for interacting with them in nasty ways.
Mammal individuals are unique , while robotic pets are their overly simplified clones. Resetting an aibo only matters to you (or its owner) , killing a real dog is something else...
<!--emo&:0--><img src='http://www.unknownworlds.com/forums/html//emoticons/wow.gif' border='0' style='vertical-align:middle' alt='wow.gif' /><!--endemo--> omg
/me runs
They run around and scurry about, right? If you had a computer program built into a little robot to do essentially the same things.. while it may not be an ant, it is just as sophisticated as an ant (perhaps more). Should killing an and be just as bad as killing this little robot that behaves like an ant?
The obvious answer is no, because it isn't alive, but it is tempting to say that. How do we know it isn't alive? The fact is we don't know. It could be just as well that for all intents and purposes, everything is alive in some moderation.
Ever think of it that way?
That's all we are, if you think about it, only we have a hell of a lot more. That's my point, i'm not saying that these things are alive as in "he hurt me i'm gonna lay some smackdown", because that'd require intelligence and a knowledge of vengance.
If an AI program is given two choices for when a gun is pointed at it, and it bases it's desicion over various things (such as how close some cover would be, how 'brave' it feels etc...), it's making a desicion.
If i pointed a gun at you, you'd wonder if you could make it to some cover before i shot you, if you could take me out before i shot you, and whether i actually mean to shoot you (that's just an average person i'm talking about there). Obviously, those are only a few choices out of the millions available.
There IS a similarity. But, in intelligence, we'd be basically God compared to that AI program. Yet it doesn't mean it's not alive.
In that way, I think maybe the energy from your feelings (Thoughts are electricity, that may be it) can influence the way electronics work. Maybe being depressed mirrors that effect on the computer. (In a way, it becomes depressed, too?)
What can you represent with 1s and 0s?
100000 = 64 which in ASCII means the character 'A'
So now it isn't just 1s and 0s, but 1s, 0s, and 'A's. Stretch that to include the entire alphabet. Why stop there? Doesn't have to stop with ASCII, does it? We can represent every symbol known to mankind by a finite number of 1s and 0s, can't we?
Look at what computers are capable of with 1s and 0s. Some computations are done 1000s of times faster with a computer using 1s and 0s than the smartest humans could ever hope to accomplish with a pencil and paper.
Maybe not to you, but to me, by saying all they are are 1s and 0s, you are telling me they have the potential complexity as us ourselves. Because in theory, you can represent anything you want using 1s and 0s. Just depends on how complex you like it.
Don't be surprised folks. We're spitting out computers now with as much wires and connections as the neural connections in the brain of a mouse that can interact with humans in real time and you couldn't tell it apart from a real human being! These things are alive. I'm not saying how much alive, but I am saying they are real living things.
On a side note, how do you think God thinks us? He created us out of clay. What do you think we are to him? We're basically just very advanced AI programs. Compared to His intellect, we're dumb as dirt. The only reason nobody likes to recognize that computers can be intelligent is because we've never seen life forms come to be in this manner before. I think there are some of you that will never get it, in fact.
Realize that what you are doing to these virtual representations of beings that retain some ability to react to stimulus looks and sounds like killing, but in reality is nothing of the sort.
It is simmilar to a game of NS in which the entities are not controlled by AI, they are controlled by humans. The human players you play against look like they are being killed, but in reality they simply are waiting and then respawning within a short time. These players don't feel the contempt of a murderer perpatrated against them when you unload an lmg clip into thier virtual bodies because they are not actually dying, and you both know that.
An AI when killed in a game is not dying. Like the human player it experiances no real distruction of itself, its host simply loses its health points and a death animation plays (or a ragdoll in one of those fancy games) then it must embody a different host or all the other enemies in the game would simply stand still indefinately once the original AI host is killed, and the game would lose it's functionality.
It may look like killing from your prospective, that is what it was desined to look like, but I assure you that there is no destruction on conciousness, even EXTREAMLY limited ones.
Secondly, living creatures are usually created out of some need to keep the species going and their job is to do the same thing. We create computers to do our work for a few years, we create AI to explode in showers of Gibs... you could even say that by killing them you are fulfilling its purpose, helping it reach its destiny!
You aren't killing anything, the AI which governs the beheivior of the model is an entirly different entity then the model itself or the gibs that it spawns on its destruction. To say otherwize would be like saying that if you distroy my calculator you are killing me, I control the functions of my calculator in a very similar manner that an AI controls the actions of it's host model.
In either set of events the intellegence that guides the object is not damaged by the loss of its slave part. The AI may be unable to interact with the user while it's model is in a "dead" state, but the AI is no more dead then it was while the model was functioning fully. I may not be able to preform trigometric functions while my calculator is distroyed, but we can all agree that I am not quite yet dead in that situation.
you should know, eh? <!--emo&:p--><img src='http://www.unknownworlds.com/forums/html//emoticons/tounge.gif' border='0' style='vertical-align:middle' alt='tounge.gif' /><!--endemo-->
The thing about computer AI, is that they are just a list of instructions. If youve ever programmed, youll see how it is. The computer never THINKS. Its just following through the instructions that are given to it.
Its like a machine. When you turn on a car, do you actually raise it from the dead? Does it turn left when you rotate the steering wheel because it takes your input and decides that it should turn that way? No. The car is engineered that way. When you turn the steering wheel, the car MUST turn that way due to the way the mechanical parts are designed and fitted together.
Similarly, the computer AI has no actual choice. It 'sees' stimuli and it reacts according to how its engineered. There is a set path that it MUST follow. The programmer cannot program in a "do what you feel like" function. No free thought, no sentience, no life.
btw, Swiftspear makes an excellent point.
You could say the AI is the soul and the collection of polygons on screen could be considered its body (if a rather intangible one). I still think most people who believe in souls feel a bit squemish when they see a body destroyed. Yes my point is slightly pedantic but then so is yours, these internet discussions thrive on pedantry.
Besides, the original post concered the death of AI, not just in on-screen characters but also the computers themselves, robots etc. I just chose to answer the whole post with a (slightly generalised) though on AI rather than pointing out that nothing we have at the moment is AI yet.
Your point was perfectly valid though and I totally agree, I wouldn't be posting anything even slightly negative your way as 1) i don't enjoy these kind of arguments and 2) I respect your opinion, you normally put forward logical, coherent points and I normally enjoy reading your replies. However I do not enjoy being accused of ignoring someone and I definitely dislike someone thinking that I would jump into an argument without reading all of the points put forward already.
If you could be less accusatory next time please, I'm sure you didn't mean it quite like that but I don't like to be besmirched in public.
Firstly, about the AI not really being killed because it's not affected at all by the loss of a model on the screen.... well, kinda.... but you'd only see it that way from our perspective. From the computer's/AI's perspective, things are just happening.... and in the case of, say, GTA, the AI responds by sending police at you, which is kind of like an anti-viral reaction our own bodies produce.
As for the "computers/ai just runs through a line of instructions" comment, that's all we do, really. I don't know much about AI (well, i don't know ANYTHING) or coding it, but i'll explain best i can what i'm thinking:
Enemies in games don't want to get shot. Sure, it's just a line of code to make the player feel more challenged, but they still don't want to get shot whatever the reason.
A human knows from it's own knowledge and experiences that guns kill people. The AI knows that the gun the player is holding can kill it simply because it's programmed (taught) to know this.
The more advanced AI will calculate actions, such as "how close am i to cover", "can i take him down" etc... some of the seriously l33t to calculations on how many of it's own "friends" it has seen die and so on... I'm sure you see the similarities here.
The only difference i can see between real living creatures and AI (apart from the intelligence) is that AI doesn't learn itself, it is taught. A human child will burn it's hand on a kettle and it won't touch a kettle again. AI will run into a wall over and over until it's told not to.
All a human is is an immense sequence of instructions and memories, with the ability to gain more, given a body to interact with the world.
AI is a relatively tiny sequence of instructions, and in the case of games, is given a body to interact with (what it thinks) is the real world.
Like i said before, i'm talking about sub-ant levels of intelligence, but it's still a lifeform, only one that we've never seen before (and AI on this level has only been around 25-30 years max).
Think about it this way. If I were given the AI's source code, and the random seeds that it uses. I can trace out its 'thoughts' and calculate (not predict, calculate) every decision it is going to make. I will be right every single time.
On the other hand, lets say we can dump all of my memories and knowledge onto some sort of storage medium. You have full access to all my previous thoughts, memories, and whatever is happening in my head at a certain moment. With that information, you will not be able to calculate my thoughts/decisions. You may be able to PREDICT some from my previous thoughts. But after a while, you will fail to predict what is essentially true randomness, something that cannot be calculated out from any code.
And as for the car analogy, you are exactly right. You tell a program to print on the screen "hi, how are you? What is your name?", enter in your name, and then make it say "Oh I remember you, <NAME>," yes that is entirely false intelligence.
I'm not talking about this though. What you're talking about is linear thought. You do one thing, then the next then the next. Humans too have such a thing. When we walk or ride a bicycle, we do a sequence of steps. If we did anything out of order or differently, we fall, so we make a point not to change how we behave.
What if we took a program, made it capable of learning and made it capable of changing itself. What it would change, at first, would be entirely random, but then it would have control over this as well. What it changes is up to the program itself. So it changes and modifies itself repeatedly until it does a peculiar pattern humans have recognized all over the planet as unmistakable life... replication.
People, even at this point, won't admit it is alive. I ask those people, what would it take for you to be convinced that a program is alive? I don't believe they have such specs, because they unconditionally categorically deny the possibility of artificial life. I don't blame them. It isn't the easiest thing to believe that life can come from inanimate objects. But then, that's all we are, right? Carbon and hydrogen and oxygen and nitrogen? Those things are hardly "alive." So why is it so far fetched to believe that you can create life from inanimate materials?
although much smaller that a computer's bits, when you get small enough, there are only two things that there can be:
mass is here
mass is not here
all we are is that mass moving around
What would convince me that we have created life? A computer that can do something it wasn't explicitly told to do so, that's intelligence.
What would convince me that we have created life? A computer that can do something it wasn't explicitly told to do so, that's intelligence. <!--QuoteEnd--></td></tr></table><div class='postcolor'><!--QuoteEEnd-->
A) Those statements aren't even backed up with evidence. You're just saying that a computer is incapable of thinking, when that is exactly what we are trying to argue about. We could go into details about what the definition of "thinking" is and all that jazz. But you can't just assume major points of your argument. It'd be like me trying to prove god exists by saying "Lets assume there is an all-powerful god.. that's Him." (Sorry, it's an itch I have to scratch.)
B) I agree with you entirely that conventional computers cannot think. They read a list of instructions and do them one at a time. That's the current model. What I'm talking about isn't that. I stress that again. I mentioned before if it had the capability to change itself, what you'd have is a model of life. In this manner, no you would not be able to predict its behavior, because it would change its behavior perpetually as long as it is run.
C) By your definition, you've already admitted there exists artificial intelligence. There are learning programs, which after the first time they are run, we don't know how the hell it will run the next time. Some programs use reverse engineering. Start with a problem, and the program evolves to solve the problem. When you do that, you do not know the state of the program at any given point in time. I mean granted, it wont' do anything more than the capabilities you give it, but that is no different than us. We can't fly, because we don't have wings. Nor can a computer without a camera be able to see the world.
Admit it. You don't want to think artificial life can exist because you think it would be like trying to give life to a rock, right? Again, that's the intuitive thinking that you need to be able to let loose in this conversation, because there is no good evidence for believing that except that it hasn't happened before, but only because only now we have this technology.
although much smaller that a computer's bits, when you get small enough, there are only two things that there can be:
mass is here
mass is not here
all we are is that mass moving around <!--QuoteEnd--> </td></tr></table><div class='postcolor'> <!--QuoteEEnd-->
try googling the "schrodingers cat" paradox <!--emo&:)--><img src='http://www.unknownworlds.com/forums/html//emoticons/smile.gif' border='0' style='vertical-align:middle' alt='smile.gif' /><!--endemo-->
I have almost no doubts that computers that are capable of this sort of thing will eventually be created (maybe they are being developed now) but as for the computers we have today, you might as well be trying to give life to a rock. I simply cannot imagine how you could get something that is incapable of thought to think just by telling it to, so that might be my problem.
I did find it a bit odd that you told me that I didn't back my statements up with any evidence, and then went on to say that you agree with me about the capabilities of computers as we know them now. If you agree, then I don't need to present any evidence, do I?
The Human body has thousands of preprogrammed reflexes, some born in the infathomable darkness we call 'instinct'. As much as we like to paint humans as thinking creatures, there are still a lot of times we seem to act before we think; look at a lot of what you do in the natural world and be honest with yourself... were you really thinking of the situation BEFORE you reacted?
Also I can't help noticing people bringing up sentience or thought... that's not the issue, the issue is whether they're alive or not. We can't exactly examine the psychology of an ant so for all we know they're incapable of decision individually; it's just as possible that they merely execute a branching path of pre-programmed insticts than make any conscious choices for themselves =/
perhaps it would be appropriate to raise the issue of what defines something as alive? ^^;
As geminosity said, maybe things are just so complex, we don't yet know how to figure out how it works. We know that you can take a computer and figure out where all the 1s and 0s are at any given point in time. However, the faster these machines get, and the more complicated the AI gets, the harder and harder it is to keep track of exactly what is going on. It's because computers nowadays are true or false, right or wrong, black or white. Human brains are not like this. Sometimes we can figure out what 12 times 34 is, and sometimes on bad days we can't. We'd have to be able to make a computer in which we sacrifice precision for humanity.
Okay, the inner workings aren't organic like living creatures, but does an object have to be composed of cells to be living? If that were true, viri would not be living, but they clearly manage to reproduce itself. So tell me what "alive" means?