|
| Thread: Can Computer's Think? The Hard AI Theory |  This thread is pages long: 1 2 3 · «PREV |
|
Avallach
Hired Hero
Disputo ergo sum.
|
posted June 02, 2002 05:19 PM |
|
|
If I recall correctly, the Turing test is basically seeing whether a computer can be distinguished from humans purely by its responses to questions. Clearly it's beyond the ability of any computer/program currently in existence, but I'm not sure how much it would prove if a computer did pass - see the earlier Chinese Room argument. If computers ever spontaneously come up with the binary equivalent of Descartes' "cogito ergo sum" I'll be impressed, and moreso if they start writing poetry of a like to say Wordsworth's. But I'm not of the view that such spontaneous and creative 'thought' by computers would be possible.
The extreme hard-AI theory is interesting. Nonsense, I'd say, but interesting speculation. We could take it further than the thermostat example, and say that every physical particle, and even energy, that has any cause-effect relationship must likewise 'think'. Just as the thermostat supposedly has 'beliefs' about the temperature, so electrons in atoms for example could be said to have 'beliefs' about energy levels - too much energy and they reacting by jumping up a level, emitting the energy as light, and jumping down again.
But I think these hard AI theorists are not in the end trying to raise up these physical processes to what we consider human thought to be, but rather bringing human thought down to this level. In the end they're in the same camp as physical/chemical determinists who say that all of our thoughts and actions are no more than a pre-determined and unalterable outworking of physical processes. In this view, our thoughts have no more value than an object being pushed reacting to f=ma.
I think the question of whether computers will ever 'think' in the way that humans do comes down whether or not you believe that humans have Free Will. This is something that I think would be impossible to endow a computer with. But if you think that we follow a deterministic pattern anyway, then why not? If human thought is simply a mapping of inputs to outputs, however complex that mapping may be, there is no reason why it could not be reasonably duplicated in an artificial structure.
thunderknight, a very thoughtful post. Thumbs up  .
Human being... I love that name, what we call ourselves. So profound, yet only as a question - or the begging of one, at least. One of the ultimate questions. To be... what is that?
Ah well, but at least we know that '42' comes into it somewhere  .
____________
"Death slew him not, but he made death his ladder to the skies"
- Edmund Spenser, on the death of Philip Sidney
|
|
kain
Tavern Dweller
|
posted June 03, 2002 11:43 AM |
|
|
thank you avallanch
thanks AVALLANCH i thought cat (god bless her inteligence PLEASE COME BACK I WANT TO LEARN MORE) was the only one taking the conversation seriosly... sondonly people were talking of red stars (too much rpg i guess, go for the next level kids! go! go! go!).
so back to buisness:
donīt think in future past or present: filosofy has no timing (at least that is what i was taught). for ai... "put your enemy or point of study in the best of all conditions" but after that all you got (simplifing since it dosenīt matter the speed in which the ai reacts, but if it can react like a human)
Wasnīt tunning the guy who invented the computer in world war? note: this british (yes americans you didnīt invent that niether) decoded with this invention enemys messages and saved many lives and then again took many (bad guys?), but people wouldnīt accept he was gay. this took him to suicide.
However he defined computers as what it still is today "simbol manipulater". Until we convince ourselves
John Searle great thinking... but when i studied it i came to a doubt. Who said humans are not based on estimulation - response?
Are we in love or acting as we were in love? like love any thing which we think as pure human (fear, hate, glutony,... yes the bad things too not everything is perfect and nice and cute).
Computes do learn. put in new hardware or soft ware :
"windows has detacted new... wait while i actuilize.." they can make music, laugh, poetry,... randomly combine nice things toguether and tada!! ya got it!!
Free Will. how true. "how did you know i was gonna react that way? īcuase iīve known your since you were a child." if we have all the varialbes (people you met, accidents, games played, movies seen,...) Could we know waht you were to do? Could you calculate what the die will read? in the begining... wasnīt it all destined to combine? and make that unicelular being (thoughs who belive it that way)
Me? iīm still thinking. i donīt know. but in the mean time iīll try to make the world a little better than what iīve found, which is more important.
hope you all understand my english i've made an efort on understanding all your posts.
____________
Recordad que en el fondo todos absolutamente todos SOMOS AFRICANOS
|
|
kain
Tavern Dweller
|
|
posted June 03, 2002 11:50 AM |
|
|
by the way...
the guy inside the chineese room is Joe... and as cat said he doesnīt understand jack shipt of chineese...
just a comment...
____________
Recordad que en el fondo todos absolutamente todos SOMOS AFRICANOS
|
|
bort
Honorable
Supreme Hero
Discarded foreskin of morality
|
|
posted June 03, 2002 03:48 PM |
|
|
It's kind of funny, generally the definition of "intelligence" (I'm not sure if the modifier "artificial" really has any meaning with reference to intelligence) is pretty much the same as the supreme court definition of pornography. "I'll know it when I see it."
That said, personal computers aren't there yet. Anybody who's ever had a civ 3 automated worker getting stuck in a freakin' pathfinding loop knows that... There are some really fascinating attempts at computer intelligence out there. I think my favorite was something where they just stored as much data of all kinds as they could in a computer, queried "happiness" and the computer gave them back a picture of a father watching his son play (don't ask for the reference, I don't remember. You'll just have take this with the same level of credulousness as the kid in 3rd grade who claimed to know everything about sex).
Anyway, what I wonder more about is where in the animal kingdom intelligence cuts off at. In the case of the great apes its easy since we can teach them to communicate with us via sign language. That's pretty clearly a sign of at least rudimentary intelligence. Dolphins, on the other hand are clearly pretty smart, but how do we prove it if we can't talk to them? Then you can go down through the dogs and cats of the world and so on and so forth.
If you're looking for a coherent thesis in this response, it's not there since I've simply been babbling.
Of course, perhaps the way the question should be asked to begin with is - do humans think? I've known several who are regularly outsmarted by rocks...
|
|
Cat
Honorable
Supreme Hero
Gonna Get Dirrty...
|
|
posted June 03, 2002 11:04 PM |
|
|
Quote:
thanks AVALLANCH i thought cat (god bless her inteligence PLEASE COME BACK I WANT TO LEARN MORE) was the only one taking the conversation seriosly... sondonly people were talking of red stars (too much rpg i guess, go for the next level kids! go! go! go!).
so back to buisness:
Thank you, that was very sweet
I should do more philosophy threads really... This arguement is actually part of a larger school known as the school of monism, which also includes the ideas of behaviourism, Identity theroy and the private language arguement (I love those last two!).
Free will is another chestnut altogether.. the free will arguement is centuaries old and is usally associated more with theology. It argues that if everything is predestined, how can we have free will? If we tie this to the AI theory, we can see that computers do not appear to have free will (the therostat guy would probably have something to say here, but lets ignore him). If we are computers, therefore, we cannot have free will at all. this is just the bare bones as i'm kinda busy but I will go into more detail later...
____________
 Diwethaf Gloau Sylw y Gymreag
http://aozos.com/phpBB2/index.php
|
|
arachnid
Promising
Famous Hero
|
posted June 04, 2002 12:57 AM |
|
|
Quote:
So, it is a question of "WHEN" will AI and the computers be smarter then us...
raZor_X
The answer is never, how can something we have made ever be smarter than us?, we cant make something smarter than us so its not going to happen. Even if you think the AI can think for its self its "programme" will still be restricted by what we know and think and feel. The most we could make is what we the human mind is capable of and nothing superior to it.
____________
|
|
Xenophanes
Promising
Famous Hero
Chief Consul to Queen Mutare
|
posted June 04, 2002 01:09 AM |
|
|
Computers cannot think for themselves. Although to the frustrated newbie it may seem that the machine wants to make their life miserable, a computer will never decide to do what it "wants."
Furthermore, a computer only knows two things: On and Off.
____________
<Dragons rule, Titans drool!>
|
|
Ichon
Responsible
Famous Hero
|
|
posted June 04, 2002 10:53 AM |
|
|
instant recall
Computers have something humans don't have and really can't match- the ability to store practically infinte amount of knowledge. So- what they lack so far is a way to make that knowledge applicable in a coherent way.
It's just about connections and relevance. You can make a computer associate anything with anything else, but can we ever program them to associate things on their own observation without a list some human created? If it's possible then it's almost sentience. Also- AI is strictly intelligence, we aren't talking soul or anything to do with free will.
Besides- most quantum theorists today know the latest research is that on the level of the smallest chunks of stuff we can examine, what we observe is due to what we expect to observe. The movement of the smallest chunks aren't random, but they haven't discovered the pattern or what causes them to change behaviour.
So who is to say everything isn't alive? As for intelligence, operating to an agenda with more than one possability can occur would denote choice, though how much choice is neccessary to be defened as intelligence probably won't ever be solved.
AI = to humans have already been achieved. Humans limited by stunted brain growth or injury perhaps, but still human right? Unless you want to define coherent thought and ability to express as what makes a thing human. So when shall AI's equal average human levels of intelligence, or even a creative genius.
|
|
Lord_Woock
Honorable
Undefeatable Hero
Daddy Cool with a $90 smile
|
|
posted June 04, 2002 07:11 PM |
|
|
Well, I guess I've heard of a couple of machines that would pass the Turing test. Too bad they were only paper boxes with people inside lol.
____________
Yolk and God bless.
---
My buddy's doing a webcomic and would certainly appreciate it if you checked it out!
|
|
Wesley
Disgraceful
Famous Hero
banned
|
|
posted June 04, 2002 09:32 PM |
|
|
:p
Well.... is a computer/PC a girl or a boy?
Male or Female?
I think a male.. he has a HARD disk ... So I guess it's a male..
In that way the humans are better then the computers.. atleast the female humans.. 
But if there are any babe computers.. then get ready to get beaten by a PC!
____________
|
|
Avallach
Hired Hero
Disputo ergo sum.
|
|
posted June 05, 2002 05:58 AM |
|
|
KAIN
Quote:
John Searle great thinking... but when i studied it i came to a doubt. Who said humans are not based on estimulation - response?
I got to thinking much the same thing after making the earlier post. His analogy applies well enough to the operation of computers, but for his argument to hold he needs to show that the analogy cannot also be extended to humans. He may well have done so though - I'm sure it's something that's been studied in one field or another. It wouldn't be linguistics I don't think, nor epistomology... but maybe something sort of in between . Maybe Cat knows....
Quote:
Wasnīt tunning the guy who invented the computer in world war? note: this british (yes americans you didnīt invent that niether) decoded with this invention enemys messages and saved many lives and then again took many (bad guys?), but people wouldnīt accept he was gay. this took him to suicide.
However he defined computers as what it still is today "simbol manipulater".
Kind of. It's actually debatable as to who first invented the computer. It's generally attributed to the Americans who created 'ENIAC', although Turing's Colossus did precede it. The question is whether or not Colossus was truly a computer. *Shrugs* In any case, Turing was certainly one of the great pioneers, and ahead of his time in terms of computational theory. A shame that he ended his life as he did.
Cat
Quote:
I should do more philosophy threads really... This arguement is actually part of a larger school known as the school of monism, which also includes the ideas of behaviourism, Identity theroy and the private language arguement (I love those last two!).
Sounds good . I'd be interested in what you might have to say about Monism. I've read a little about it, but I'm afraid I just don't seem to get it.
Quote:
Free will is another chestnut altogether.. the free will arguement is centuaries old and is usally associated more with theology. It argues that if everything is predestined, how can we have free will?
*Nods* Well yeah, there's theological determinism such as the Calvinists have, in which people's fate is predestined by God. But I was thinking more of physical determinism, in which everything about us is an outworking of purely physical reactions. Perhaps there's a better term than free will to fit the opposing view, but it was the only one I could think of to fit.
Determinism is actually rather a funny beast. It can't be argued for without eventual self-refutation (or something close to it), and yet that doesn't preclude it's being true. A tricky one for philosophers, I would imagine.
Quote:
If we tie this to the AI theory, we can see that computers do not appear to have free will. If we are computers, therefore, we cannot have free will at all.
Hehe, well yes, this is the other way of looking at it, which shows that it isn't really an answer at all. If humans have free will, computers cannot 'think' like humans, and if computers can think like humans, then we cannot have free will. Around in circles.
I guess that what I was trying to get at though is that if someone has a pre-existing belief in physical determinism, I would not try to argue that computers could not think as humans do. With the different presuppositions, we'd merely be talking past each other. Indeed, I would argue that given that presupposition they should conclude that AI can potentially reach an equivalent to human thought.
Arachnid
Quote:
The answer is never, how can something we have made ever be smarter than us?, we cant make something smarter than us so its not going to happen. Even if you think the AI can think for its self its "programme" will still be restricted by what we know and think and feel. The most we could make is what we the human mind is capable of and nothing superior to it.
I don't think this argument is valid. Going from the presupposition of determinism, lets suppose that we did create an exact, artificial replica of the human brain, as you suggest is possible. What if we then added extra memory capacity, and made the connections faster? We would have created something 'smarter' than us.' I mean, if you believe that RM/NS can take us from pre-ape brains to human brains, why could intelligent design not achieve something greater still? No, I think the question is of a more fundamental nature than one of being able to exceed ourselves.
Ichon
Quote:
AI = to humans have already been achieved. Humans limited by stunted brain growth or injury perhaps, but still human right?
Ah, but how do you know this? How can you say what level of thought exists in brain-damaged humans, and whether there exists AI that is directly analogous to it?
Lord_Woock
Quote:
Well, I guess I've heard of a couple of machines that would pass the Turing test. Too bad they were only paper boxes with people inside lol.
At least they did pass. Reassuring, that .
____________
"Death slew him not, but he made death his ladder to the skies"
- Edmund Spenser, on the death of Philip Sidney
|
|
Ichon
Responsible
Famous Hero
|
|
posted June 05, 2002 07:21 AM |
|
|
anagolous or not
"Ah, but how do you know this? How can you say what level of thought exists in brain-damaged humans, and whether there exists AI that is directly analogous to it?"
You can say much the same thing about computers really. How much can we really understand about the process it goes thru to achive the results- even in simple mathematical functions which is the main work of computers, the process of carrying an integer, or dividing a fraction is achieved in terms translated for us, but the way in which the computer actually achieces the correct result is incomprehensible to most people. Not to mention all the left over circuits which carry little flux of energy when the processor sits idle- the neurons(anagalous neurons) are still firing away, just at a slower speed and not showing any result, but if there were a menu or a translation system for showing the functions being performed, whose to say there wouldn't be an actual result which is more than random.
As for human thoughts or performance, we aren't even anagolous to one another as a species, so it would be a fallacy to even attempt to compare another sentient construct in that way. In the same way damaged humans or ones which can't express themselves if there is something to express can't be judged, or the animals which demonstrate adaptive intelligence can't have that intelligence quantified without a means of communication.
|
|
Avallach
Hired Hero
Disputo ergo sum.
|
|
posted June 05, 2002 10:32 AM |
|
|
Quote:
As for human thoughts or performance, we aren't even anagolous to one another as a species, so it would be a fallacy to even attempt to compare another sentient construct in that way.
Hmm, well I guess it depends on the level of abstraction. At higher levels it's possible to compare them, I think. But in any case, your original quote was:
Quote:
AI = to humans have already been achieved.
What was your use of the 'equals' sign here meant to signify then? Or was this original post simply a joke?
Quote:
In the same way damaged humans or ones which can't express themselves if there is something to express can't be judged, or the animals which demonstrate adaptive intelligence can't have that intelligence quantified without a means of communication.
What, so this is how you equate them to AI? That neither are able to 'express' a certain type of intelligence, and so are equivalent? Poor humour, if that was your intention. If not, perhaps you could elaborate further... and I'll try not to be too antagonistic .
____________
"Death slew him not, but he made death his ladder to the skies"
- Edmund Spenser, on the death of Philip Sidney
|
|
Ichon
Responsible
Famous Hero
|
|
posted June 05, 2002 11:52 AM |
|
|
lol
Well, the AI = humans was clearly a joke since I had been talking about human stupidity. ;-) All the poor AI out there ya know...
I decided to see if I could defend it though in the second post just cause some people didn't get the joke in the first one-
I didn't take a bold stand defending it, but it's a pretty good defensive argument being circular. As for what I really believe, when an AI becomes self aware- which isn't neccessarily when a computer first meets the commonly established criteria of intelligence, I seriously doubt we'll discover the fact of said AI's intelligence for some time if ever.
|
|
Avallach
Hired Hero
Disputo ergo sum.
|
|
posted June 05, 2002 12:40 PM |
|
|
Point taken then .
____________
"Death slew him not, but he made death his ladder to the skies"
- Edmund Spenser, on the death of Philip Sidney
|
|
kain
Tavern Dweller
|
posted June 05, 2002 05:43 PM |
|
|
sorry to take a step back but... a guy in comma is a person? or like aristoteles said a sleeping man? at what point of pregnacy is the baby a baby? (i just gave away my opinion)
impresive how theese thoughts are related with AI. please if yaīll gonna talk about monism and other stuff i donīt undertand give me a ref. or the word for it in spainsh things are pretty tough as they are.
thanks again iīm learning alot...
your welcome my pleasure cat
____________
Recordad que en el fondo todos absolutamente todos SOMOS AFRICANOS
|
|
Fenix
Known Hero
In ranks of Nebwoocka Alliance
|
|
posted July 14, 2003 06:08 PM |
|
Edited By: Fenix on 16 Jul 2003
|
Similarity
I'm a club chessplayer, and you know what is the importance of Artifical Intelligence in chess. People calculate only several variations because they have learned during their experience what are good moves at first sight. Computers don't know this, but they still calculate all variations on system that humans use.
So, AI in the real life would have to calculate more, because humans have the feeling for actions and situations, but if AI knows the basic parameters it could actually think.
AI can think, but does it actually understand and has it conscience of what it is thinking?
|
|
Draco
Promising
Famous Hero
|
|
posted July 16, 2003 06:02 PM |
|
Edited By: Draco on 16 Jul 2003
|
interesting topic
Computers Vs. Humans
intelligent or not?
Computers and Humans are nearly the same, we take different approaches to acheive the same goals though.
Computers:
- to learn need to be fed exact information
- have unlimited memmory
- have feeling*
- will evolve at incredible speeds
- short uselfull life span (1-5 years) after that they become useless old and withered
- Eat require electricity
Humans:
- to learn need to be fed any information
- very limited memmory
- have feeling**
- evolves extreamly slowly (each generation grows an average of 1' - 2' or a couple centimetres)
- long lifespan (aprox. 50 years) after that they become useless old and withered
- Eat require food and drink
* computers have feelings similar to humans:
- when a computer feels hot it will shut down. or tell you to cool it down.
- Computers try to keep themselves from getting sick (viruscan)
- Computer learns what it can and canot do from experiance (change your resolution to 1600x1200 and if it canot do that it will never let you do it again)
** Humans have feelings similar to computers:
- When a human is hot it will tell the computer to turn on the air conditioning to cool it down.
- Humans try to avoid getting sick, they get antivirus needles
- Humans learn what it can and canot do from experiance, ever do something you think was stupid like jump off your two story house? and hurt your legs landing? touch electrical wires? youll never do that again
edit: i wasnt done but when i hit enter it posted! arghh, i guess i hit the mouse.
Computers are evolving as we speak, in my oppinion though, i believe they are much much more advanced they the technology industrie leads us to believe, computers know exactly how to be the faster most reliable they can be, and the only reason i think they are not being used is it would be the only computer you would ever have to buy, it would cost next to nothing to make, but they would have no residual income, due to the fact people would never upgrade.
i should have stopped when the computer told me to. oh well humans do things they know are stupid, and computers dont
|
|
Orion
Known Hero
Dark God of Ordered Chaos
|
|
posted July 29, 2003 06:09 AM |
|
|
humans do not need programing we adapt at the moment that is not something we can successfully program into computers
and we do not understand with any real accuracy how we think feel etc so we can not program computers to do it
the parts of a computer can only do one computation at once while humans can i am talking about indivdual parts here people so dont get confused
any way i believe it is impossible for any person group etc to create anything superiour to its self so even if we do makle ai we will not make it better than us
____________
Though I walk through the valley of the shadow of death,
I will fear no evil, for it bends to my will
|
|
Cat
Honorable
Supreme Hero
Gonna Get Dirrty...
|
|
posted July 29, 2003 10:32 AM |
|
|
I'd have to argue regarding the scope of a computer's feelings. It has been claimed that although a computer can have beliefs, and even be programmed to pass the turing test (including he more recent gender test), a computer is incapable of experiancing human emotions on a higher level.
To use a example:- a computer programme could be created to run on HC. I could do all the things members do:- spam, post insigtful topics, gain or lose QP's, even hassle the mods . Difference is it would be unable to fully apprciate what it was doing (ie, would not be angry if it had posts deleted, although it may appear so).
Same as, if an art critic programme was created, the computer would find no real joy or beauty in the art, as a person would.
____________
Diwethaf Gloau Sylw y Gymreag
http://aozos.com/phpBB2/index.php
|
| |
|
|
