Artificial Intelligence [Archive]

BaiginLong

12-10-2002, 12:16 PM

This topic is for the extremely geeky. Be warned, jargon and technical terms will run rampant here. Now for the my intro. Many movies have mentioned the possibility of a thinking machine. Many portray them negatively (i.e. Matrix, Resident Evil), some portray them positively (i.e. Bicentennial Man, Robocop), and a good deal of them try to straddle the fence (A.I., Terminator). But, however, we are still left with the question of whether or not they can exist in the first place. If they exist, can they be man-made like most movies say, or will they spontaneously create themselves from the mass of data left over on the Internet? And why would we create them? What functions can they perform that can't be performed by computers today? How could we keep them from replacing jobs or even mankind when they do come into existance?
I have my own opinions and information from the hacker world's perspective, but I'll let you guys start first. Give your opinions and informed conjectures in an educated manner please and keep trivial comments and jokes to a minimum.

Two interesting sites to look at while discussing this topic.
http://mitpress.mit.edu/e-books/Hal/
http://www.kurzweilai.net/

Dhaval 4 U 2 NV

12-10-2002, 11:42 PM

...

Lets start off on a not of skepticism.... frankly AIs, in my opinion are us... only in a chemical nature... i dont think it is possible to reach such levels of intelligence on a purely electromagnetic arena...

The questionsasking to be redeemed are:

1) Wat is their most effecient source code language?
2) How do they survive?
3) Wat is their lifeterm?
4) How they can be intelligent?
5) Wat is their social structure?
6) Wat is the purpose of their life?
7) How do they affect wat happens to the universe?

these are a few peebles from a mountain of questions...

later
dhaval
...

BaiginLong

12-11-2002, 01:23 AM

ahh see you are moving the subject a bit quick man I want to see what everyone else thinks of this. You already have had some lectures/question and answer from me and Scott Clinton. So let's not be asking those questions just yet.

BaiginLong

12-11-2002, 01:33 AM

Now how can they exist and through what means? You have to be specific. You can't just say "yes" and "no". You have to say "yes, because ..." or "no, because..." simple statements with no proof or explaination mean nothing in the scientific world unless their givens and obviously the question at hand can't be answered by a given.

BaiginLong

12-11-2002, 02:34 AM

I said Artificial intelligence, not android
android are do-able with a lot of work but are true Artificial Intelligences possible and AI does not have to reside in a humanoid body or any real body for that matter so let's focus more on the real topic and not robots though we all love our robots
mecha rule heheh

Dhaval 4 U 2 NV

12-11-2002, 02:46 AM

...

ok, a bit slower... if not sluggish.

I agree with kuroBishounen, maybe one can say cloning is a means of "creating" intelligence, but how artificial is it going to be? Maybe one day we might be able to make a "Data" like in Startrek, but will it be an intelligence of the same caliber? or "human" without flesh? For yall who have read "Ender's Game", i think the AI in that book was more realistic... in a static, electronic form... that may be possible... but will we make one accidentally?

i think the whole internet network is in itself an AI... it is able to communicate, react, and affect.

later
dhaval
...

BaiginLong

12-11-2002, 03:14 AM

ahh we may be getting somewhere with you Dhaval
but I need to sleep...g'nite

ism

12-11-2002, 09:17 AM

"Intelligence" is a really vague term that even humans still argue over. Sure, there's the Turing Test, but is that even really a good, complete test of an AI? A computer beating a grandmaster at chess used to be considered a test, and now that it's happened, it's obviously not a good test.

The other thing that bugs me is that the models are patterned after how human thinking works (or approximations of it) -- neural nets, etc.. Maybe that's the wrong approach of transforming biological processing to electronic.

Perhaps molecular computing is the Holy Grail? It seems the breakthrough in AI is always around the corner, but it's been that way since the 1980's. I am not expecting it anytime soon, and not even sure if it will be in my lifetime. I know Kurzweil said something like 2020? He's an optimist. Sorry, I can't contribute much more to this as it is not my specialty.

Craig

12-11-2002, 09:55 AM

Is this supposed to be a discussion on the general field of Artificial Intelligence or a discussion on the more specialized area of Intelligent Robotics ?

VV o n g B a

12-11-2002, 11:35 AM

given that we don't have a good understanding of where intelligence comes from in humans, i don't see us creating a machine that has intelligence as we know it in the near future.

i could kind of see it arising from self-evolving parasitic programs like in the book "hyperion," where they evolve themselves in the right environment, but otherwise, i think we're too stupid to do it intentionally. even then, the environment would have to have massive resources lying around unused that just aren't here yet for the intelligence to amount to anything significant.

ChairmanMah

12-11-2002, 07:07 PM

doesn't anyone remember "Blue" the supercomputer by IBM?

luv

12-11-2002, 08:05 PM

I guess you would have to define Artificial Intelligence for me. What level of communication would render an inanimate object as "intelligent"?

I remember watching early episode of Futurama where Bender rants about humans oppressing his "race" -- (something like) using us to [points to sprinkler system] do your dirty jobs with no pay. So say we do produce "intelligent" inanimate objects, at what point are they granted rights as individuals (like in Andromeda or the Holographics in Deep Space Nine)?

BaiginLong

12-12-2002, 12:49 AM

when I say artificial intelligence I mean that it would actual behave as if it were alive
it acts out of its own volition and is not programmed to perform this or that
it "learns" and "understands"
and it may even have a personality
so perhaps I should change my phrasing
Artificial persona or artificial consciousness may be more appropriate terming
in essence it can't just be intelligent it must also be sentient

Hanuman

12-12-2002, 12:57 AM

Isn't artificial intelligence a misnomer? I mean, it's either intelligent (by our definations) or it's not. AI should be better coined Inorganic intelligence. Off topic, but it's late and my mind works in weird ways.

Ok, I need to define what I believe is AI first. Independent thought, the ability to sort and analyze data and intepret it. The ability to learn, not just storing data, but real learning. Logic, not formulas, but independant logical thinking. For example, I was with my neice one morning and we watched the sun come up. She said to me, "the sun came up today. It came up yesterday. That means it's going to come up tommorow." Simple, yes, but shows the progression of her logic. Lastly, it has to have an sense of identidy, sense of self. Since it has the ability to learn from it's environment, it's developement will be shaped largely by it's outside influences, the external data it receives and learns and shapes it's paradigm by. This means 2 exact machines, in different environments, should develope with unique characteristics (dare I be so bold as to say personalities?)

So can we get there? Yes, I think so, we're not there yet, but I think one day we can get there. Look at the slow evolution of our robotics we have robots that can move and respond to external stimulus, like, move towards light, stay away from edges etc. Not too different from simple single cell organisms. Now making that leap from single celled organisms to a simple insect is difficult enough, to jump eons of evolutionary steps to fully thinking entity seems pretty out of reach right now. Of course we have programs that 'learn' but they are still more like programs that recognize similar patterns and plug them into parameters already set up by their programers. Still a leap from AI.

As far as source code, lifespan and all that. I haven't a clue, you can tell from my ramblings, I'm not a tech guy, just some one who used to enjoy reading Omni magazince when they were still around. :)

Craig

12-12-2002, 01:02 AM

There are already plenty of intelligent agents out there that "learn". It's just that their behaviour is applied to a narrow domain. That's the current trend in AI, to create isolated solutions for specific problems and not replica humans per se. Are you restricting your questioning to hardware based entities or applying it to pure software programs (like perhaps spiders) ?

BaiginLong

12-12-2002, 01:18 AM

pure software because hardware really means nothing in the idea of AIs to me
Tawee I love your reply
for now I have an essay to write but I will be making longer posts when I have time

himura-dono

12-12-2002, 02:37 AM

robocop had a human brain with an artificial interface for the brain and the "body" to communicate.

as for the existence of AI anytime soon, well, TRUE AI, not the extra clock cycles being used by the axis powers in bf1942 or anything, lol, as you stated in the software based reality only is probably far off, but in time a very likely human evo/devolution.

a side question: once we have machines to think, interpret and respond, what will happen? will a software glitch happen and create more than what we believed is AI or AC, causing TRUER AI and AC? seeing us and reacting or revolting accordingly to our actions?

angel nympho

12-12-2002, 09:41 AM

AI is just scientists on a power trip.

If you can't go about creating humans through cloning, they wanna go around creating artificial intelligence.

Who care whether or not they CAN do it. It SHOULDN'T be done.

wylin

12-12-2002, 09:51 AM

SunWuKong

12-12-2002, 10:56 AM

what are we really talking about here? creating consciousness?

first we have to define consciousness... and we haven't been able to agree on that yet.

we've got a lot of work to do

BaiginLong

12-12-2002, 11:14 AM

Perhaps you should read the Fuzzy series by H. Beam Piper
excellent sci-fi books
made me thing a lot about sentience, sapience, and consciousness

SunWuKong

12-12-2002, 12:00 PM

oh... so we're just talking about sci-fi?

ok nevermind

BaiginLong

12-12-2002, 12:39 PM

no we're talking about for real
sci-fi can be used for reference

wylin

12-12-2002, 12:42 PM

i want my robotic slave now! slavery should rise again! enslave robotS or apes?>

SunWuKong

12-12-2002, 12:46 PM

yah i want a robotic sex slave that's just intelligent enough to get me off.

SunWuKong

12-12-2002, 12:47 PM

sci-fi can be used for reference

the AI in sci-fi is... well... pretty fictional. but *shrug* whatever. take some cog-sci classes if you're interested in this stuff.

BaiginLong

12-12-2002, 12:57 PM

well i was using the the specific example i had merely to offer a definition of sapience

Dhaval 4 U 2 NV

12-12-2002, 02:04 PM

...

ACs .... so wat do u think are the criteria for such beings? i mean in the sense of Drakes Formula.

Not that we are stupid, but rather curious that brings about our discoveries to an extent really. Almost all o humananity's most significant discoveries have been done through coincidence or accidence. (maybe there is a 'higher force' at work, but so much for being scientific)

If such beings are pure energy, conducting their life at the level of electrons, then i think it is safe to say they may actually be on a higher evolutionary ladder than humans...was it our making of computer networks and internet that brought them about... or was it just a migration for them?

later
dhaval
...

angel nympho

12-12-2002, 06:16 PM

theres nothing wrong with cloning and AI those techologies should be pursued. because they can be useful later:

clone:
- bring back dead or extinct species
- Stem cells that can regrow ur arm or eyes or hoho's
- bring back dead pets
- be a stepping stone world wide acceptance of genetic manipulation.

AI
- computer will process information faster
- computer will adapt
- a must for bipedal robots (*active ai that adapts to terrain)
- good for making robotic slaves
- will crash still must use DOS code and F16-32 code to bloat ai then no more fear. it will bluescreen every so often and crash!

Nobody gave us the right to fuck around with life. I think a mistake in this field can be a mistake that'll cost humanity a HELL of a lot more than it's worth. What makes you think that if you create artificial intelligence, it'll wanna worship you? Like the human mind is so advanced..... ?

wylin

12-12-2002, 07:39 PM

fear is what delays technology cloning and AI is comming. AI routines in simple games have been getting more complex as time progresses? whats wrong with it, sorry i havent been watching T2 over and over and skynet hasnt taken over. actually its more the opposite AI and Cloning will be used to help humanity not harm it. ALso ne internet lifeform like skynet probs will be so full of porn its useless!

wylin

12-12-2002, 07:40 PM

also we have always fucked w/ life:

examples of man fukin' w/ life

- domesticated animals and crops
- dog breeds they are all decended from one ancestor and eugenics programs have been uses to breed certain traits in and out of the animal.

we been messing w/ life since day one why not continue, if america isnt willing to doit sumone else will. :lol:

also i wouldnt mind sum faster reflexs and enhanced processing power for my mind and better memory. why not it helps man accend to the next step of his constant evolution.

SunWuKong

12-12-2002, 07:45 PM

i think it's dumb that people think AI will take over everything. we're already paranoid enough before the invention of a free-thinking machine. and it's not like when the first free-thinking machine is born, it'll automatically be connected to enough machines to take over everything. secured networks are actually isolated from all other networks.

angel nympho

12-12-2002, 08:50 PM

I really have no problem with messing around with cloning and AI. I just think people are so arrogant to think they're masters of it all and have a RIGHT to do things like that.

SunWuKong

12-12-2002, 08:55 PM

I really have no problem with messing around with cloning and AI. I just think people are so arrogant to think they're masters of it all and have a RIGHT to do things like that.

why not? people can do what they want as long as they're not hurting others.

luv

12-12-2002, 09:04 PM

But you can't be sure that the process will not hurt others. Must also take into account the type of people who are driven to creating A.I. or cloning. Are they true scientists who want to make a discovery? or Austin Power Dr. Evils who do it because they can?

I think the creation of new life forms is a great thing for humanity if we are trying to compensate for all the lifeforms we destroyed. I think we can create sentient objects or tell robots to create themselves, but once that happens what are "they"? Do "they" have rights as individual [networks?]? can we own "them" like a pet? Do we leave "them" alone to evolve deep in the middle of the earth?

SunWuKong

12-12-2002, 09:52 PM

But you can't be sure that the process will not hurt others. Must also take into account the type of people who are driven to creating A.I. or cloning. Are they true scientists who want to make a discovery? or Austin Power Dr. Evils who do it because they can?

well we might as well not do anything. how do you know that glass bottle you're not recycling isn't going to hurt others? etc etc etc...

also, we don't incriminate on the basis of what people's agendas might be. who do you think you are, Bush? :lol:

wylin

12-13-2002, 08:36 AM

like i said before leave sum windows and dos code for instability and secondly make them outa plastics and soft materials also give them an off switch!

angel nympho

12-13-2002, 11:15 AM

why not? people can do what they want as long as they're not hurting others.

I agree, but people are so quick to believe that what they're doing is progress. It's usually not. And in retrospect, I think there's a million things out there that people have created thinking it was in the name of progress that shouldn't be out there today.

wylin

12-13-2002, 11:36 AM

like what? nuclear power is the key to the future and the stars!

SunWuKong

12-13-2002, 12:44 PM

I agree, but people are so quick to believe that what they're doing is progress. It's usually not. And in retrospect, I think there's a million things out there that people have created thinking it was in the name of progress that shouldn't be out there today.

i'm not sure if they think that they're doing something that's considered progress. what if they were doing it out of pure scientific interest? do you think they have a responsibility to ensure that what they're doing will contribute to social good as well as not have any negative impact on society?

angel nympho

12-13-2002, 01:08 PM

i'm not sure if they think that they're doing something that's considered progress. what if they were doing it out of pure scientific interest? do you think they have a responsibility to ensure that what they're doing will contribute to social good as well as not have any negative impact on society?
When it comes to certain matters, yeah, I think scienctists have a responsibility not to seriously mess things up for the rest of the world. I think it's important to weigh one's options.... It's not like we NEED to know how to do all this stuff to survive.

SunWuKong

12-13-2002, 01:20 PM

When it comes to certain matters, yeah, I think scienctists have a responsibility not to seriously mess things up for the rest of the world. I think it's important to weigh one's options.... It's not like we NEED to know how to do all this stuff to survive.

well, really, in their defense: it's not the scientists that apply their knowledge to achieve negative results. no scientist really is in control of the application of his/her findings and no engineer really is in control of the application of his/her inventions. they're all at the mercy of whoever funds them.

angel nympho

12-13-2002, 01:44 PM

well, really, in their defense: it's not the scientists that apply their knowledge to achieve negative results. no scientist really is in control of the application of his/her findings and no engineer really is in control of the application of his/her inventions. they're all at the mercy of whoever funds them.

That's like saying the man who created methanphetamines isn't at fault.... because HEEE doesn't care what other people do with it. Drug dealers aren't at fault cuz it doesn't matter to them what the people who buy from him choose to do with their drugs.

SunWuKong

12-13-2002, 02:38 PM

That's like saying the man who created methanphetamines isn't at fault.... because HEEE doesn't care what other people do with it. Drug dealers aren't at fault cuz it doesn't matter to them what the people who buy from him choose to do with their drugs.

this would only be a good analogy if say, the scientists or engineers are making something that, for example, causes mass destruction. if i invent the world's most powerful computer and whoever funded me chose to use it for evil purposes, am i to blame for my invention?

luv

12-13-2002, 04:11 PM

That's exactly it. Although people may be at the mercy of whoever writes their paychecks, they shouldn't be completely in the dark about the applications of their product. This is not an argument to "not do anything", but one in which the inventor should tread carefully when creating "life" -- if the definition of A.I. includes self-awareness and intelligence then it is creating life.

angel nympho

12-13-2002, 04:26 PM

this would only be a good analogy if say, the scientists or engineers are making something that, for example, causes mass destruction. if i invent the world's most powerful computer and whoever funded me chose to use it for evil purposes, am i to blame for my invention?

For all I know, a man who was researching a cure for cancer inadvertently created methanphetamines. That's probably not true, but it could be. But I think it IS true that a lot of drugs that are abused today weren't created for any harmful purposes. In fact, NONE of them are created for HARMFUL purposes. People who make drugs like extasy and all that crap don't do it with malicious intention (i.e. wanting to kill off our generation or something), they do it to make money or just because they like the drug. But the fact that one's intentions aren't bad doesn't erase what they created.

If you invented a super computer, you should still care that your invention is being used for evil purposes, yeah. You should try to stop it, if you don't believe in the cause. If I was doing lots and lots of breakthrough research on nuclear weapons and weapons of mass destruction, but not actually MAKING them, I'd still be very careful what to tell the person who is funding the research, if I thought they'd use the information for malicious purposes.

SunWuKong

12-13-2002, 04:28 PM

That's exactly it. Although people may be at the mercy of whoever writes their paychecks, they shouldn't be completely in the dark about the applications of their product.

well you are assuming that the scientists and engineers would have at least some inkling of the purposes of whoever funds them. i mean alot of people do like to be funded by entities with a track record of goodwill, but there are people out there who do not have that luxury.

ChairmanMah

12-13-2002, 04:51 PM

I think scienctists have a responsibility not to seriously mess things up for the rest of the world

anyone can be irresponsible.

the engineering manager does coke at my work.

politicians have a responsibility not to mess up but most of them do anyways.

angel nympho

12-13-2002, 06:50 PM

anyone can be irresponsible.
Does that make it okay?

wylin

12-14-2002, 11:12 PM

i'd tell my weapons and AI technology to the highest bidder remember i am taiwanesE!

BaiginLong

12-15-2002, 03:13 AM

Here's the scoop from the hackers' point of view: A.I.s do, in fact, already exist. True A.I.s are not what you might think they are, however. An A.I., by technical definition, is a self-compiling/executing and self-editing section of machine code which has it's own system of logic, cognition, and learning, can think sybolically, has it's own personality, is not only intelligent but conscious and sapient, all of which are self-developed. They are almost essentially digitized humans with a stronger emphasis on logic and whatnot. They can have preferences, likes, dislikes, interests etc. However, they cannot really be programmed. In fact, 99.9% of all A.I.s in existance today are completely spontaneously created from the junk code of computers, networks, and the Internet itself. Of course the number of A.I.s are little in the first place so that percentage really doesn't add up to much. The most a human can do to help in the creation of an A.I. is to provide somtthing of a code catalyst of sorts. That same 99.9% of A.I. exist on networks or on the Internet and avoid remaining on any hard drive. They live on RAM, cache or any kind of temporary, storage of the network. In other words shut down the network servers, mainframes, and backbones and the A.I.s die. Those that do live on hard drives though can survive computer shut down.
Errr that's it for now I don't really wanna go too deep into it yet since I prefer more of a question and answer type thing if I am going to explain what I know. At best I'm something of an A.I. socialogist or psycholgist. I definitely don't know all of the details on the technical end but I'll try to answer either way.