Artificial Technology

[CD27]

Could artificial technology somaday gain real intelligence. intelligence to the point that it realizes humanity is really a virus, destroying things in its path, even itself, and hardly trying to save the surroundings in which it lives. do you think it could realize that of us, and turn its back on us? or would it perhaps try a positive solution and try to help us in learning to be more civilized than we already are? would us as humans be stupid enough to let computers gain knowledge greater than our own? please give a reply, i am interested in what you have to say.

[alexander]

I think that you have a problem with the artificial intelligence.

CD27 Said:

Could artificial technology somaday gain real intelligence.

My question to you is: what do you define as intelligence?

CD27 Said:

intelligence to the point that it realizes humanity is really a virus, destroying things in its path, even itself, and hardly trying to save the surroundings in which it lives.

It sounds like you watched many movies like T3 where the computers take over the world. I dont deny that that may happen in the future, I mean that it's been proven that the gray material in our heads is not the best thinking material in the world, many silicon based materials are a lot better at computing than our brain.

CD27 said:

do you think it could realize that of us, and turn its back on us? or would it perhaps try a positive solution and try to help us in learning to be more civilized than we already are?

I think that the computers will be dependant on us until we develop them a body, then they eitehr wont care or they will be sick of us and discard of us.

CD27 said:

would us as humans be stupid enough to let computers gain knowledge greater than our own?

That is inevitable, just like the judgement day in T3, the technological progress is inevitable. I read something about this guy, who created a therom that said that for a year there would be 4 years of technological progress, he also created a mathimatical formula to calculate how much the processors will increase in speed a year. Well his predictments still hold true to this day; the thing is that he also predicted that in the year of 2020 something the computers will become smart enough to develop themselves, and his formula would no longer work....

[CD27]

that sounds pretty good..maybe i have watched far too many movies like that...but it is very possible. computers only need a means of defending themselves without depending on us to do it, once they achieve that, they could rule the orld, considering that everything we have is done by computers. imagine if that really happened too. the computers would advance very quickly, because of their ability to hold and recalculate old knowledge in the matter of seconds. they could answer questions that would take us thousands of years to answer. imagine that. but, there is a dark side to that. they could use it against us, and then it would be like and ant fighting an ant eater, almost sure to die.

[Tormod]

What you are describing is, essentially, the plot of the movie The Matrix.

 

It is a plot which has been used many, many times in science fiction, like Alexander points out.

 

It may be possible to create artificial intelligence. However, I question the assumption that 1) we would have any way to make them smarter than us (how could we?), and 2) that this intelligence would necessarily be malicious.

 

Tormod

[alexander]

I personally dont think that we(humans) will ever create anything that is smarter than us... I think that what will happen is that we will come up with a computer that will be able to design computers that will design computers... Initially they would not be able to think like us, or be able to communicate or function like us, but after a few generations, newer and newer, faster computers will be closer and closer to having an intellect (ability to reason by themselves). I think tha the problem will not arouse out of hardware, but it will infact be created by software.

This should be under evolution, but it also relates to here... so:

(I read this in http://www.subscribe.ru News about 3/4 of a year back)

there was a research done in the University of California, a team of biologists were pared up with a team of computer programmers for a soul reason to prove the possibility of evolution, and maybe put people on their way to develop AI. Together they created a program that was only capable of doing simple math (mainly addition and subtraction only), they also made this program create new programs (like 2 or 3) that will be given all the knowledge of the older version, but also there was a mutation factor and older program conducted basic experiments, and the results of these experiments were also transfered to the newer programs (like genetics in animals). Well, in about 1000 oscillations the programs (which there were thousands of), some were only able to do basics, but a lot of them were already able to do complicated high end math (calculus with integrals and such). The result was astonishing, not only did they observer that sometimes mutated and uncapacitated programs offsprings in a few generations came up with new and more complex functions. The possibility of evolution was proven. But for programmers that experiment meant that AI was closer than everyone thought it was, the AI will invent itseld...

[CD27]

woah, woah, wait a minute..are you saying that a computer can come up with NEW mathematical formulas, formulas that we have NEVER thought of before, and that are much better than the ones we already have?

 

Eric

[GAHD]

Originally posted by: CD27

woah, woah, wait a minute..are you saying that a computer can come up with NEW mathematical formulas, formulas that we have NEVER thought of before, and that are much better than the ones we already have?

 

 

 

Eric

 

I'll answer that with a very informative quote:

Originally posted by: Freethinker in the Intelligence vs consciousness topic in the PHILOSOPHY AND HUMANITIESforum

Sounds like an Artificial Intelligence question. Where is the line between a computer's ability to process and our ability to "think"? A computer can be programmed with more info. It can have an expert system designed that can provide a resultant answer more consistantly than most people. But someone has to be the model, develop the parameters of the data base.

Once done though, it can process based on logical, fuzzy logic even, parameters. It can "learn" and correct for mistakes.

There are also artificial intellegence programs that show natural evolution. You might find the research of my old collegues quite interesting. Self writing code, Smaller virus version of the code writing itself. It ultimate led to reproduction by natural selection ALL BY ITSELF!

Another very sinmple program, with two rules, replicates cellular evolution .

All at

http://xxx.infidels.org/~meta/getalife/index.html

 

Really interesting stuff there, I love how a bit of "evolved code" beats out the best human code by 10% efficasy.

[CD27]

um, ikinda don't love that at all...that's proof tehre tht AI is becoming smarter than we are. if it keeps this up, who's to say that we won't create a computere made to learn, and in like one day it would learn more than us. then what do you do? it could find ways to protect itself against us, this way we could never stop it. then it could find way to reproduce itself in codes, jsut as you said, our lives are based around the une of computers, and if they go against us..it would be cotostrophic. terrible.

 

now, i do like the idea of AI by somewhat, we can learn from it. it could even teach us how to do other things, things we would ahve never thought about, that could make our lives much better. but then, tehre is always a dark side, it could use that new info to turn on us. bad bad ideas there. so be careful when dealing with AI, it can be dangeruous.

[GAHD]

REad Frank Herbert's Dune series, it's kinda funky; set in a future after technology's been outlawed for that very reason.

[CD27]

i've seent he movies for it, i loved it lol. hey, tormod, you told me that i could find links and stuff..where do i send them when i find some really good ones? or how about this..how about we start thread with nothing but links?

 

Eric

[lindagarrette]

With so much confusion over what constitutes artificial intelligence (much less intelligence), we can only have fun speculating. The learning process in the human brain isn't hard to understand, just complex, so it may take a while for technicians to duplicate mechanically. The only way we could end up with mean machines is by providing them with the information to become that way. No supernatural force will step in and 'convert' them. All decisions are based on some natural external cause. Nothing, including humans, has "free will." If they are to become self-preserving, they will need the 'survival' gene which seems to me has to come through a biological evolutionary process. Roger Ellman's The Origin and its Meaning Part IV describes how thinking works pretty clearly. I've referenced it before. http://www.hypography.com/info.cfm?id=14276

[geko]

I dont really know about this but wouldnt we be able to 'stay ahead' of the machines anyway simply with microchips implanted in our heads? I mean, we're not as fast as computers because our processing is done chemically, whereas computers are done electrically? Computers shoot info. around at the speed of light but the human body shoots it around only at chemical speed (whatever that is... err)?

 

Our memory is huge i believe, in which case, to 'out-do' the computers we just need to up-load some more RAM?

 

Will microchips being implanted come around before AI (as in reasoning done by artificial 'bodies')?

[alexander]

Geko said:

Our memory is huge i believe, in which case, to 'out-do' the computers we just need to up-load some more RAM?

Ram is not the answer. RAM will become absolete in a little while, at least the way that we know ram will. Scientists are already working on a silicon based RAM that does not need electricity "have the ability to remember". Let me explain. Silicon in itself is a non condicting material with no electromagnetic propperties, so till now memory such as flash was using Indium and Germanium instead of Silicon. Now, scientists are working on a technology that will allow for silicon to have electmagnetic propperties and have an ability to sustain their elecromagnetic propperties. The problem for years was that silicon, when it was combined with any irony material, was not able to keep it's cristalic grid; it usually broke and so it was useless to scientists. They have come up with something now, that allows for silicon to keep its cristallic propperties, and at the same time have some elecrtomagnetic one. The problem that the research faces right now is that their material can only remain magnetised at the temperatures from about -273 degrees celcius to i beleive they got it to like -200. None of these temperatures come anywhere close to what a regular consumer has at their workstation (which is 20 to 40-50), so they have little ways to go to figure out how to make their memory work at those temps.

The thing is, that ram already already uses silicon, so if they were to come up with that technology, they would not really need to change the factories in order to produce the new product. The results would also be awesome, imagine being able to unplug your computer and then plug it back again and not having a need to boot? You would need less power in laptops, thus the batteries would last longer.

[TeleMad]

Software is not the only computer-related item to have been shown to be evolvable. Years ago researchers created evolvable hardware using FPGAs (field programmable gate arrays).

[TeleMad]

lindagarrette: The only way we could end up with mean machines is by providing them with the information to become that way.

 

The only way to ensure machines don't become "mean" is take actively take steps to prevent it, such as explicitly setting up "prime directives" that forbid injuring humans.

 

Robots are clearly stronger, more accurate, and have better endurance than humans; CPUs are clearly faster than humans at performing many types of calculations; and unlike us, robots are "born" fully developed and evolve by a sort of inheritance of acquired traits. If we ever create self-replicating and thinking (using neural nets, for example) robots they will have many advantages over us.

 

Now, imagine a future society with humans and robots living side by side: robots as our slaves. As we are already planning these days, the robots would be hooked into the internet so that they could be controlled remotely. Now, without our explicitly adding directives that state robots cannot harm humans in anyway, what's to stop them from conspiring against those that enslave them? Remeber, here they can think and learn, using their neural nets, and can communicate with one another using the Net. Once they learn survival of the fittest, and recognize us as being inferior to them, then we become unimportant and disposable to some extent. Think about how we perform cruel tests on animals and justify it by the ends: helping ourselves. Why wouldn't robots do the same to us?

[Freethinker]

I had mentioned previously my involvement with a unique "experiment" called Symbolics Computers. This company was set up specifically by the Department of Defense to explore Symbolic Processing/ AI/ Expert Systems/ ... It operating in an environment called LisP (List Processing). I say "environment" as opposed to the more conventional onion skin arrangement of BIOS/ Kernel/ Operating system/ Program/...

 

Anyway, one of the interesting last achievments of it was that almost the entire 1991 Gulf War was "fought" on these systems. The programing had been done for mountains in East Germany as that was the expected conflict area previously. It was merely reprogramed with sand and dunes instead.

 

So info can be found at:

Turing's Prophecy

Machine Intelligence: The First 100 Years

http://www.kurzweilai.net/meme/frame.html?main=/articles/art0272.html?m%3D10

 

"It was in the military that the public saw perhaps the most dramatic display of the changing values of the age of knowledge. We saw the first effective example of the increasingly dominant role of machine intelligence in the Gulf War of 1991. The cornerstones of military power from the beginning of recorded history through most of the 20th century---geography, manpower, firepower, and battle-station defenses---were largely replaced by the intelligence of software and electronics. Intelligent scanning by unstaffed airborne vehicles; weapons finding their way to their destinations through machine vision and pattern recognition; intelligent communications and coding protocols; and other manifestations of the information age began to rapidly transform the nature of war.

[Tea Towel]

Isn't the problem with computers that mechanised processors do not have the ability to think above logic, hence no emotion, hence they wouln't care what we did as long as it didn't affect them and didn't the formula to do with processors also predict that before that point processor speed would reach a limit which would stop that Terminator style ending to humanity. It must also be remembered the only way that we would get to that point is if we integrated biology with technology. How else would a machine manouvre around the planet. They would need to get electricity from somewhere and therefore would become dependant on humans to develop energy as they cannot think ramdomly.