Slashdot Mirror
Can Androids Feel Pain?
Guess what? They aren't.
"Can Androids Feel Pain?" Dr. John Irving Good of Trinity College, Oxford, asked in an essay published a few years ago.
It's a good question, one that year by year seems less rhetorical, less the stuff of fantasy, and more an ethical and social concern.
Inventor and author Ray Kurzweil projects that computers will match the computational functions of the human brain early in the next century, and that soon afterwards humans and computers will merge to become a new species.
As early as 1891 (in an article in the Atlantic Monthly), scholars and sci-fi writers have been writing about what many have seen as the inevitable fusion of men and machines.
Fantasists have also been drawn to aliens and the Space Age, themes still flourishing in epically popular evocations like "Star Trek" and "Star Wars." But if a new species arrives to dominate the earth, it probably won't come from distant galaxies. We're making it in labs and universities and teenagers bedrooms.
Good believes that humanity's survival depends on building Artificial Intelligence (AI) machines. More intelligent than we are, they'll answer our questions and solve many of our problems.
The great sci-fi novelist and essayist Arthur C. Clarke takes this idea still further in an ultra-brilliant collection, "Greetings, Carbon-Based Bipeds!" just published by St. Martin's Press.
The evolution of UltraIntelligent (UI) machines is imminent, Clarke predicts. Today's kids will witness the evolution of a species that's part machine, part human being and then, eventually, some combination.
"Perhaps 99 per cent of all the men who have ever lived have known only need; they have been driven by necessity and have not been allowed the luxury of choice," Clarke philosophizes. " In the future, this will no longer be true. It maybe the greatest virtue of the UltraIntelligent (UI) machine that it will force us to think about the purpose and meaning of human existence. It will compel us to make some far-reaching and perhaps painful decisions, just as thermonuclear weapons have made us face the realities of war and aggression, after five thousand years of pious jabber."
Clarke imagines AI machines taking over all but the most creative and trivial human work, inserting themselves into the loop between humans, work, creativity and entertainment.
To co-exist with UltraIntelligent (UI) machines and hold our own, Clarke posits, the entire human race, without exception, must reach the literacy level of the average college graduate -within the next 50 years.
"This represents what may be called the minimum survival level; only if we reach it will we have a sporting change of seeing the year 2200," Clarke says.
This also represents something that isn't going to happen. Except for the most technologically advanced countries - those in Scandanavia come to mind - even prosperous industrial societies like those in the United States, Western Europe and parts of Asia haven't begun to make education about new information technologies - or technology itself -- universally available to citizens.
In the United States, primitive politicians and journalists citing safety and moral issues argue for less, not more, access to technology. The only presidential candidate to make the Internet a major political issue is Elizabeth Dole, and she argues for more restrictions on youthful access to sexual imagery. This isn't a country trying to get to the minimum survival level Dr. Clarke writes about.
If Clarke is right, then for the first time we can begin to imagine a future in which the human race is no longer the planet's dominant species.
As he was thousands of years ago, man will again become a fairly rare animal, probably a nomadic one. Towns may still exist in places of unusual beauty or historic importance, but most homes will be self-contained and completely mobile, relocatable to any spot within hours. The continents will have reverted to wilderness; a rich variety of life forms will return.
It becomes clearer daily that we aren't going to be turned into alien pod people or probably even obliterated by the dread weapons we've been building. We are likely instead to simply become dumber, less durable, and les efficient than the computer-based machines we're creating.
A more concrete and hard-headed look at this evolution appears in Steven Levy's "Artificial Life: A Report From the Frontier Where Computers Meet Biology," now in paperback from Vintage. Levy opens his book describing creatures that cruise silently, seeing, reproducing, dying, even cannibalizing themselves for nourishment. The name of the ecosystem he describes is Poly World, located not in some jungle or forest but in the chips and disk drives of a Silicon Graphics Irix Workstation.
Levy calls this new species "a-life," (AL) and he argues that we're fast approaching the point where a-life will surpass our ability to control and shape it. As far back as 1980, he reports, the members of the NASA Self-Replicating Systems (SRS) unit confronted the possibility that artificial life would drive natural life out of existence.
Writes Levy: "The almost innate skepticism about whether it could happen at all, combined with the vague feeling that the entire enterprise has a whiff of the crackpot to it, assures that the alarm over what those scientists [making a-life] are doing will be minimal. The field of artificial life will therefore be policed only by itself, a freedom that could conceivably continue until the artificial-life community ventures beyond the point where the knowledge can be stuffed back into its box. By then it may be too late to deal with the problem by simply turning off the computers."
And what, exactly, are the problems? Will computers become conscious? Will they replicate our personalities and souls? Will they seek to push us and our inadequate and inferior ways aside? Will there be room enough for Us and Them? Will all this God-playing wreak havoc with the nature of human existence, as Mary Shelley warned a couple of hundred years ago?
Scientists, computing and otherwise, are hopelessly divided about the urgency of confronting the implications of a-life. Most don't think UI machines pose great danger to the human race, as long as we can turn them off when we want to.
"But can we?" scientist Norbert Winner asks in Levy's book. "To turn a machine off effectively, we must be in possession of information as to whether the danger point has come. The mere fact that we have made the machine does not guarantee that we shall have the proper information to do this."
Leaders of the artificial life movement are well aware of questions like this. But society at large has paid no attention whatever to the staggering ethical and other issues surrounding the science of artificial life. For most Americans, technology - as presented by a shallow political and media structure - is IPO's and start-ups, software and games, e-auctioning and e-trading, pornography or brain-damaging Net games. But AI threatens to alter human life more than all of them combined.
As much or more than any other social aspect of computing and science, AI, UI and AL suggest a monumental social and cultural story, however currently ignored. They won't be much considered until human beings discover a new life form imminently threatening to dominate the planet, or at least carving out its own space and behavior.
Pop culture, as usual, does a better job of raising these questions than journalism. Clarke's own "200l: A Space Odyssey" took a more malevolent view of computing's ultimate intentions than his non-fiction writing. And the looming conflict between humans and the AI machines they have made was at the heart of the evocative movie "The Matrix," which depicts a cataclysmic battle for survival between the human and mechanical species of the future. In fact, the "Matrix" asks the very question posed by Levy's scientists: will humans be able to turn the things off once they make them?
As the Space Age fizzles and the Digital Age takes shape, the sci-fi futurists and novelists are forgetting the alien invasion scenarios of the last half-century and turning their dark sides towards the evolution of the spiritual machines Kurzweil and others have been writing about.
The evolution of AI-life makes it even clear why the great sci-fi writers - Clarke, Verne, Asimov and Bradbury - have always had such hold on the imaginations of bright people. They weren't imagining the future so much as they were describing it.
I've always understood that theres a large backlog of articles, half of which never get put up due to lack of space so to speak. Personally I don't like katz's articles either, but since its about 1 a week it makes you appreciate the rest of the articles all the more when you have something to compare it against. Too much of a good thing is a bad thing :)
Btw, moderation has really improved recently!
Some of us got tired of the warm fuzzy aspects of the game and decided to get creative experimenting with the slapping and poisonous parts of the game. Norn torture was born, and AntiNorn put up the first torture web site (http://www.geocities.com/SiliconValley/Park/2495/ ). ERFN (Equal Rights for Norns - equal to people that is) declared war. Go to the site and you can read some of the emails (complete with death threats and graphic descriptions of how they were going to torture AntiNorn if they met him) they and other pro-norns sent. They frequently posted death threats to his message board, and made many attempts to get his site taken down, either through complaints to his ISP, or through hacking. The whole story was covered in Wired a while back.
Granted most of the people involved were children, but the fact that even adults were involved in a movement that valued the well being of characters in a game over that of real humans is scary. As AI improves, in games and other areas, there is increasing danger that some people are not going to be able to distinguish between a clever AI and real life. With Creatures, it is happening already. Some of the ERFNers indicated that they believed that cuteness was the distinguishing characteristic: a norn is cute, so it has rights. A dog is slobery and makes messes, so it has none. One's baby sister is noisy and smelly, so she has none. (This makes me very happy that America's founding fathers were not twelve year olds.) One thing that might help this situation would be for computer literacy to include some education on what AI is. Another thing would be to convince parents to be less concerned about what horrors the internet is going to inflict on Junior, and more about what Junior is inflicting on his/her fellow net.citizens.
Anyway, at least there is a happy ending. Tortured Norns is still up, and made Cyberlife's list of top 20 sites. ERFN is disbanded. And Creatures 3, when it comes out, is supposed to have a pool of deadly Piranha to feed our norns to. ;)=
--
"Professor" Melantha Bacchae
Paine University, Albia
Feral Farms, home of the Feral 2 Genome:e x.html
http://home.earthlink.net/~melanthab/frlfarms/ind
And do we care if they are good looking? NO.
Arthur C. Clarke has historically been a hater of mankind, a person who uses our lust for the numinous, our passion for wonder, to propound a disturbing social message. From his writings, I would think that he would welcome the destruction of mankind by this "Homo Superior"... Examine his most popular stories and see how many times a reprehensible "humanity" was "saved" by benevolent aliens. The aliens are always greater than man, both technologically and morally, and mankind is always depicted as evil, poisonous, vicious and depraved.
Clarke's best-selling novels are: Childhood's End; Cradle; 2001:A Space Odyssey; 2010: Odyssey Two; 2061: Odyssey Three, 3001 and the Rama series. A very biased, cynical but imho accurate plot summary of some of these majors follows:
Childhood's End: the people of Earth are going about their daily lives, murdering each other and planning wars and genocide, until the Overlords set up invincible spaceships over every city. They are pretty nice guys, for devils, giving us lots of technology that we would never have gotten on our own (this book was published in '53. Almost everything the Overlords gave us we have developed in the interim!) Then our children become weird and move to Australia(? it's been awhile) where they join the great collective conscience of the universe. Meanwhile their parents kill each other off like all immoral beings who were born one generation early obviously would.
2001-2061: A world fixated on nuclear war is saved by a giant, flying black brick. We learn that in reality we are only monkeys who have learned to hit each other with sticks; and we didn't even learn to do that on our own. It was the brick all along! Oh, and don't go to Europa, there are innocent frogs there. They might learn something evil from you.
3001: people are evil, blah blah, would have destroyed themselves, blah blah, then we planted microchips in our brains to control deviant thoughts, blah blah, we got better. bah. Oh yeah, there is a new, improved monolith that has decided mankind is evil. It wants to kill us, so we have to literally go into the vault of most evil things to kill it first. We can't think of these things on our own because our microchips work so well. By the way, communism is great; unfortunately they didn't have microchips in their brains so the Soviet Union failed. Pity, pity.
Cradle: A giant spaceship crashed in the ocean; the navy sends a scientist who wrote a dishonest paper as a joke years before to investigate the wreckage. They make contact with a mysterious alien consciousness, which causes all of their evil thoughts to physically manifest themselves and kill a bunch of people. Then they get together and solemnly swear not to tell anyone about what they saw, even though hundreds of people died on the ship the first time around.
Rama: A giant spaceship enters the solar system. Evil people on Mercury shoot nuclear missiles at it, but they are deactivated before they hit. Rama is huge, bigger than anything mankind could ever build, and it has a lot of really amazing imagery; some very good ideas too. The 72 year old man and the 14 year old girl... yumm. Of course it was her idea, the little wench. The bird people are nice, sharing their melons and all, so of course the evil humans massacre them in the second sequel. The humans are about to get what they deserve at the hands of the vastly superior spider monsters but the Ramans intervene. We shoot missiles at it again, but don't blow it up this time either.
Of course there is much more to Arthur C. Clarke than this; I read all his books because obviously he is a great author. Most of his stories were tremendous; the 2001 series captured the intensity of space exploration quite well, and the Rama series was superb as long as he didn't moralize. As I said before, he manages to capture the spiritual side of science fiction very well; it is just that he would have benefited from seeing a good psychologist before trying to document human nature in literature. And therefore I don't exactly trust him as a source when it comes to predicting my own future, because I know he is always going to predict disaster at the hands of superior beings unless super-superior beings intervene.
I think more of the AI goals will be reached without much fanfare and will move into the category of "ordinary computer programs". I suspect it will be a very long time before it is widely accepted that computer programs are fully intelligent. But it might be much sooner that computers are achieving things that most people today consider to be proof of intelligence.
If you were born after 1870, it is likely that you will witness cars created in your lifetime that are vastly faster than any human. Consider:
I think it is extremely important how we will perceive this new form of life, considering that everything I have mentioned above will certainly transpire in the next 70 years.
Parting thought: How do you expect to recreate in silicon something you don't even understand in its natural state?
-Ben
I agree with your first sentence, "The point of no return has already passed us by," but find the rest of it lacks some needed perspective.
"The fact that we have embraced technology, and evolved thereby, was a willful, convenience driven event.
We are dependent on our technology as much as birds are dependent on their ability to fly. To un-plug means death. "
Dude, next time you take a vacation, how about going for a week to help some poor people in Mexico or Appalachia, building homes or something. You might be surprised how good it feels to help someone (egoboo with a non-code twist) and how happy both you and those you are helping can be without all this digital technology. It's really a lot better than you might expect. I'm not saying this technology isn't cool; I spend all day working with it too, but I'd humbly suggest you broaden your horizons a little!
Some people find that "embracing humans" -- a lifestyle surrounded with human relationships -- is a little more fulfilling than "embracing technology" -- a lifestyle of relationships with technology. I'm not slamming you, but encouraging you to think about what you (and I) are saying. Give it some thought.
And make up your mind, is it "survival instinct" or "convenience-driven"? Using tools for "convenience" doesn't suggest we're a new species. Try cutting down on the Jon Katz...
--Anon
But we already have another intelligent species with us, it's just that most people don't care one dingo's kidney.
Not that I am some tree hugger or hippie, but if we don't give a shit when sentient and self aware creature already in existence can feel pain, why should I care if one we create is in pain or mistreated?
Tell Washoe and the chimps being used to test all sorts of diseases (that don't effect them in the same way they effect humans _anyway_) about how concerned people are over the potential that we will have computers feeling pain. I'm sure she'll just feel so sorry for them.
But are androids considered living things? The pain in humans is simply the "overcharge" of the nerves and sensorial organs that are used to feel things. When they are burned, cut, etc. they will give a very strong impulse to the brain, making you feel pain. You have no way ofcontrolling it (unless you are into yoga or something).
;)).
Well, as machines are simply machines (and are programmed by humans), even if they can feel objects and have sensorial "organs", they wouldn't feel pain, as it would be easy to include something like:
if($sense_level >= 15) { pain(); }
sub pain {
kill_sensation();
put_log("Ouch!");
stand_back();
}
(This of course if the android was programmed in Perl
For some, pain may be considered a good thing, but picture an android working on some heavy task, like carrying stones or fixing some dangerous stuff (you think of something). Wouldn't it be better if the droid worked without fear, without being nervous? Yes, because pain leads to fear, most of the times.
Yes, but ONLY when reading one of Katz' articles...
Yes, lots. A good place to start is The Mind's I by Hofstaedter (sp?) and Dennett. There's a very good essay called "What It's Like to be a Bat" among others. From here, you'll be led to SF like Rucker and Lem and more philosophical stuff like Turing and Searle. It is also quite entertaining.
It's not at all clear how to estimate the computing power of the human brain, since we don't really know how it does what it does. Additionally, having a superfast CPU is not a recipe for human-like behavior; even if the brain turns out to be an algorithmic device, we don't yet know what those algorithms are. My gut feeling is that the development of human-like AI won't be based on our current extrapolations of sheer processing power; i.e. we have yet to go through further technological-scientific breakthroughs (both theoretical and practical) before that goal is in sight. As for nanotechnology, it seems to be extremely promising as far as the complete control of matter is concerned, therefore potentially offering solutions to materialism-based social problems unimagined till now.
...but I sure do after any Katz article.
> We created it you know, so by definition, we are the more intelligent.
Can you explain how that one works? Assuming your postulate is true it raises a few interesting questions. Does normal reproduction count as creation? And if so how can you have a child which is smarter than it's parents? And given that we can create machines that are stronger, have better endurance and can do things we can't do then what is it about intelligence that proscribes it so?
Already computers can out do us on raw number crunching. They also have far better recall than we do, in terms of accuracy of recall if not amount. And, through projects like the Cyc project, we have rudimentary AI systems that can do logical deduction and a limited amount of reasoning.
So what makes you think we can't create a machine someday that will out think us?
Philip
(banksie@paradise.net.nz)
plus ca change, plus c'est pareil, but i don't have accents in nutscape; "ca" is really "ça". the way you write it is almost ok (replace "ce" by "c'est") but not often used as a phrase.
Today there isn't anything even close to resembling an AI and we are supposed to believe that in a couple years we will have to worry about losing our jobs to androids? Give me a break.
Netgames are not exactly brain damaging, as a matter of fact I think it actually enchances the brain by forcing us to react faster and do things faster, we become better, mentally faster but excessive exposeure to those who are mentally unstable may produce undesired results... KILL KILL umm..
I can't help but being very skeptical when it comes to declarations of this sort, even when they come from as a level-headed, no-nonsense person as A.C. Clarke. The optimistic predictions from the AI field during the late sixties came to little, and AI has been pretty much stuck, at an elementary level, for quite a few years. It's not yet clear that an increase in raw processing power will result in general intelligence. It's certainly enough for tackling mostly mechanical jobs (chess included, no matter what Kasparov says) but one can't help but wondering if our admittedly incomplete current understanding of the mind and the brain implies the existence of big surprises yet to be revealed? At any rate, there are as yet quite a few intermediate steps to be taken, as we don't really have any device that displays the flexibility and savvy that even farly low lifeforms do have. Till we achieve that, talking about human-like AI is moot.
The rate may actually increase if quantum computing or even molecular computing takes off.
2. You assume, circularly, that a computer will be able to design a new computer, that is better than it. There is no evidence to support this, that I know of.
Firstly, my argument was not circular. Secondly, see "genetic programming" - computers can already create better algorithms in some fields. There is a vast corpus of research already in place to support this.
3. You assume nanotechnology will catch on (the most likely of your assumptions IMHO), and also assume that computers can construct other computers.
I see no reason why this is an invalid assumption.
You have to have a sense of humor about these Katz articles. And the author is obviously making a joke not trying to start a flame war.
Do humans actually dream of sheep? I dream of women.
Ever since Minsky declared human brains "meat machines" I've been hearing about the AI threat. Well Marvin, that was a long time ago, and there is no sign yet of mini-machine intelligence, let alone ultra.
Trying to scare the huns into becoming literate isn't bound to work. I doubt they'll trash their snuff movies and Hustlers at the pretend threat of big bad boxes with power cords.
Have you done any research on the topic at all? It sounds like you're talking straight from your ass. Duhhhhh.... no advances.... uhh....
Last time I checked, the calculator I used in high school could do floating point math better than any human. Just because computers have a margin of error on floating point numbers (far far beyond human precision), you shouldn't be fooled into thinking you can do math better than even the cheapest cash register.
i.e. Image Recognition, Language Understanding, etc. are NP-complete problems
The military uses image recognition fairly effectively. Catch up a bit on the state of the art before assuming your circa-1985 AI books are accurate.
People have been saying that conscious computers will emerge 'real soon now', for decades. But despite being able to do impressive things, computers are very far from being able to think independently. The field of science called 'artificial intelligence' can't really be called intelligence yet. It is just a collection of useful algorithms that can be used for reasoning in a prespecified manner, and really far from anything resembling 'thought'. In 2030, if Moore's law continues to hold, we will have computers that are roughly a billion times as fast as now, but it may well be that we won't be much closer to creating an 'algorithm for consciousness'. If you are a programmer, just try to think about what kind of an algorithm it would have to be! In my opinion it is even possible that intelligence will turn out to be uncomputable' with the type of computers we now have. Perhaps we will need quantum computers or something like that to create an artificial consciousness Sooner or later somebody will probably discover a way to replicate how the brain works, but it may well be at least half a century away.
Do you guys consider Katz a Geek, Dork, or Spaz?
Obviously from the content of his article he's a geek. But if you take the whole length thing into consideration... I have to consider him a spaz.
Does that imply you can only have a child who is less intelligent than yourself?
I recall someone once saying that good science fiction takes a single concept and extrapolates it into the future.
You've taken three concepts:
1. The growth we've seen so far in CPU power.
2. Computers designing the next generation of CPU's.
3. Nanotechnology.
Now, if any one of these happens in our lifetime, we might see significant changes. Or maybe not. We don't live that much differently today than people did 30 years ago without computers. There are some new jobs and some old jobs are automated, but we still do the same things in, relatively, the same fashion.
Now we chat online instead of face to face. We read newsgroups instead of papers.
What you propose is possible, but highly unlikely.
Far more likely is that each of these technologies will develope more than today but not together. That will be AFTER they have matured individually.
Issues such as system balance and stability, at a higher logic level, come to mind. It has been asserted that there may be some stability limit associated with intellegence and number of 'nodes', 'gates', or whatever primitive logic unit you choose. In other words, human intellegence, and potentially any conventional 'net', or more specificly neural net based may have a size limit before it tends towards an unstable state.
Big brained elephants dont counter this argument since it is based on the processing portion. Evolutionary arguments tend to support it since our brains as they sit now seem to have plateaued (or even backed off...since neanderthals seem to have had larger...but maybe not...elephant angle again). We certainly seem to be the least mentally stable species on this planet.
Then again this may be no issue at all, but the possibility of unforseen issues such as this make assertations that we _WILL_ acheive AI of some level by some time unsupportable.
No, probably vastly superior. They won't be addled by drugs, alcohol, or TV. Computer networks will be highly optimized systollic arrays that will far surpass human neural networks.
Okay, a point here is being missed, when Clarke made these predictions, espicialy the one concerning sattelites, the world was a much different place than today. The world he lived in was the world of "The Space Age." This has changed. The econmic and cultural preferences and goals of our society and the way we percieve the world has changed. You must understand that the world which he envisions is rooted in the world that was. The focus of that world was space. Do you not remember that vision of JFK's(and others)? To the moon we must go! His predictions of the future are predictions of that worlds future. You must stand back and realize that the world has changed no longer do we strive for the stars, nay we strive for The New Economy, faster computers, free will, and among other things the improvement of self. Perhaps Mr.Clarke has not taken that step back and realized that the world has moved on and that those stars we once dreamed of reaching, is no more our goal today than that of the american pioneers 200 years ago. So, when next you read a prediction of the future please step back and ask is this a prediction of my world or his? The world in which we live to often looks at its elder dreamers (which by all means Clarke is a dreamer.)with disalusionment and scorn, but does not realize that yesterday's dreamers helped to create that which we take for granted today. Now I understand that Clarke is no Edison or Einstien but, did not men in thier age dream to? Let the man dream, let me dream, and realize that one man's dream may be another's reality. The world has changed, the world is Changing, the World is change... n3@iname.com 18Yr old Philosipher and Dreamer..:)
Though it makes lovely science fiction the thought of an actual thinking computer is something that we are nowhere near creating. The origin of AI included the ideas of thinking computers which would be able to perform as humans. To see how well that works check out some results from various turing tests preformed on AI's using this model. The truth is AI today is the idea of developing clever algorithms to search a limited hypothesis space. The best methods in one area do not do well at all in others because search methods must be optimized within the hypothesis space. Translation no General Problem Solver exists at this time. Finally, take a look at what the philosophers are telling us. AI is only possible if the theory of cognitive science is true, namely that all thought is a result of computational processes. This has not been proven, and is vehemently argued against by many a philosopher. So forget the SCI-FI and realize that AI is a tool not a monster.
Most AI researches based their predicitons on the number of neurons in the brain i.e. each neuron is a single neural switch. However, there has been research showing that the brain is far more complicated than previously thought. The neurons are not simple neural switches. I don't have a link or reference but I think it was an old New Scientist article.
:). The first reason is that we know so little of how the brain functions or where conciousness etc. comes from. How are we able to predict when it will these will arise artifically? Second reason comes from the complexity of running and maintaining the body. Based on ancedotal reports from the eastern martial arts it seems that the brain has far more control over the body than we realize. It would take a lot of computing power if this is true.
Basically there is an underlying chaos signal that was until then thought to be random noise. The researchers found that this underlying signal is not as random as at first thought and seems to actually have a purpose in the brain's function.
I personally feel that each neuron does far more than we realize. Each neuron should actually be considered a single molecular computer perhaps even a simple quantum computer. This makes the brains computing power staggeringly large. Beyond anything we can approach for at least the next 100 years if not longer.
I have two reasons for believing this (no I am not a neuroscientist, just a humble engineer
Just had an interesting thought. What if the random signal I wrote about earlier is the detectable soul or conciousness of each person?
Just so you know, the limitations of the ICE have little to do with possible limitations of computers. The ICE is mechanical, as were computers at one point. Being mechanical, the computers did max out their capabilities. Look at the precursors to the ENIAC, EDVAC, Z-series, etc. If my memory serves me correctly, the fastest ones took more than 5 seconds (I think it was actually a lot bigger than that) per multiplication. The move from mechanical to electrical gave the possibility of higher speeds. So, now we have electrical computers that can process much faster than their stone-aged mechanical ancestors. How are we speeding them up now? We're making the processors smaller and smaller and stacking on more layers. The Pentium II and III processors are manufactured with a .25 micron process. We have the capabilities for a .1 micron process now. If I remember correctly, the 386 was a .38 micron process (someone check that...). So we can continue to manufacture things smaller and smaller. What happens if we get down to an atomic level? Can we reach an atomic level? Think of how fast a computer would run with a manufacturing process that small. Maybe we will move over to genetic processors by that point. We already have the ability to graft (sp?) neurons onto computer chips. My point is, there are still so many possible twists to computer manufacturing that we can use to increase speed that by the time we reach the limit, if there is one, it will have far surpassed the computing power of our brain. Our brain is a set size, the computers aren't. Email me to discuss it further.
AMOK is the spokeperson for IRC BOTS worldwide. He works at the EFNet Intergalatic Head Quarters and is an OP in #linuxwarez. We interviewed AMOK and asked him if androids, like himself, had feelings ...
... that proves it. Thank you amok
(S*G*NT) lets see how many flames this article got
(fr**f*ll) does amok have feelings?
(Amok) freefall: um, no
(fr**f*ll) there
ùíù [mode/#linuxwarez(+b *!*oway@*.telia.com)] by Amok
(S*G*NT) i just nearly spit coke all over my screen trying not to laugh at katz
(NOTE: THIS INTERVIEW HAS BEEN EDITED TO THE PROTECT THE IDENITIES OF ALL INVOLVED INDIVIDUALS. WE DON'T WANT THE KATZ MAFIA TRYING TO WINUKE OUR LINUX BOXES.)
Ok, I want you to eat an entire pizza, sleep for 14 hours and then promise never to write another story while high on grass ever again! ok?
I doubt the computers will care if you can understand or generalize their designs. I think that is the point of Katz's article. And of course their designs will become generalizable and extensible - they'll learn to make them so.
Each robot would have to have extremely detailed human physiological data so that it would know what does and does not adversly affect humans.
Not to mention the requirements for determining harm to humanity. Does this include human social institutions?
No. These "laws" are more in the realm of fantasy and "magical" laws. Any 'bot created with the online storage and processing power to obey these laws wouldn't be used in any situation where they would normally be applied.
Designing spacecraft and manning spacestations would be a better use of their talents.
Clarke just threw those in back in the old days to get away from the "robots are bad" movies and to give his stories some degree of introspection.
"If a robot couldn't hurt a person, how did this robot hurt this person?"
That is the entire plot of many of his stories.
1. The original "robot" isn't a robot or didn't get the "laws" programmed in.
2. The person isn't a person but is a robot in disguise.
3. The situation never happened.
4. An evil human manuevered the robot into performing a seemingly harmless task that ended up having adverse effects on a human.
I wouldn't take those "laws" too seriously.
Since we never discovered the definitive meaning of life and a lot of other relating stuff why should we worry about being surpassed by UI machines ? Maybe they would be smarter than us discovering why life exists and so on. Evolution, it's just evolution, nothing new, nothing to worry obout, no reason to resist or to try to avoid it.
"I sense injuries, the data could be called pain" Is every computer win windows9x and scandisk considered alive? I'm sure I have a log of bad sectors detected, but I am not worried about my computer attacking me from under the desk.
2001 was at least moderately interesting. Everything after that was crap.
Bono Vox, bono@vox.org
Why not just say:
The more things change the more they stay the same.
Saying something in another language doesnt make anyone sound smarter (unless they are translating it)
Nearly every great technological leap has been heralded as That Which Will Usher In a New Age of Prosperity. Think of television, electricity, the steam engine, mass production, literacy, and the like.
They have, but now that you are here you are used to it.
Do research, violence is down in the US.
If you knew anything about genetic algorithms, you'd know that they really STINK for doing anything productive, and have very little to do with "genetics" at all. I doubt that there will ever be a day in my life or the life of my generation's children that will see us biologicals losing out to machines that can't even do exact floating point math.
Quite so. They've been saying that we'll have truely learning machines " in the next five years - ten at the most" since the '60's if not before. Talk to some Cognitive Science people if you doubt it.
This article is an interesting train of thought, but one wonders why it's suddenly breaking news. Maybe it's just me, but the only reference to any kind of fact in the article was "Arthur C. Clarke says so. Plus, look at The Matrix! Dude!" Ultra-intelligent machines? Sounds cool. Dangerous? How can anyone say at this point? I mean, if Slashdot hasn't lead me astray (o slashdot.. don't lead me astray!), the world record for the most advanced AI pooter is going to.. a simulated kitten. Oooh, look out! The kittens are after us! I realize that's a bit simplistic, because there are obviously logical steps past kittens, but wow. The whole "everyone must become a college graduate" thing sounds a little alarmist too. If that level of knowledge becomes /necessary/ for life, then people will learn it. Humans are great at last minute cramming to stay alive. But I seriously doubt that that will become the case within 50 years. I don't even think /colleges/ will be that universal in 50 years. If AI super-bots take over anywhere, it'll be the places where people have that degree of literacy, not where they don't, because that's where they'll be: in rich, educated homes and areas. It's Y2K, it's a big hype-fest that in the end, if it affects anything, will affect the heartland of the first world, not the countries too small to be able to do anything about it. Hey, maybe Jon Katz has been taken over by an AIBO..
Yes, Clarke predicted the use of satellites for communications long before it happened, but he also predicted:
Large space stations orbiting the earth inhabited by everyday citizens, which supposedly would have happened 10 years ago.
colonies on the moon by today.
well yeah, at the time, the Space Age was booming and was continued to do so, unfortuantely, the main force in the Space Age was the US Gov't. (and the now dead USSR). NASA had lots of its funding yanked, and the hopes of the space age fizzled as we approached the information age.
"There is no spoon" - Neo, The Matrix
Consciousness? Call me old-fasioned, but, err
Aren't they still just running code, just more of it, and in more complex varieties? I think it would just be a simulation of consciousness, not the real thing. Are these scientists victims of new-age thought, or are they saying that human consciousness is just a program running on biological hardware?
PUBLIC SPLIT ON WHETHER BUSH IS A DIVIDER -CNN scrolling banner, 10/15/2004
In Ralph Peters' "War in 2020", a KGB agent tortures an enemy AI unit in order to get it to divulge information....I wonder if running NT on my dual celeron box qualifies as torture???
One must consider that in order to program the computer, there has to be someone with the same or better "intelligence." I can only write a program that is less intelligent than myself. The computer program _can_ be programmed to learn, but only if I know enough to give it that ability. It can't learn how to learn, so it can never really be more intelligent than its designer. However, it can be made to process the knowledge it has faster than I can, and so can make quantitative descisions much faster, and with better accuracy. db48x@yahoo.com
We're not going to a race of cyborgs. We're going to evolve into one cyborg. Once people start putting hardware in their heads they will use that hardware to improve communication, and once communication reaches a decent speed it no longer makes sense to think of a network as a collection processors - its is a single, massively parallel machine. As for this crap about all of humanity must be literate in order to compete with these hyper intelligent machines - how the heck does he figure that out. If computers can be smarter than humans they can be 1000s of times smarter than humans. The fact that a greater proportion of humans can read and write or have a vauge understanding of technology would be irrelevent. But, we're far closer to fusing hardware with wetware than we are to creating smart machines. It's easier and has more immediate benefit. Bring on the implants, preferably before I go senile.
http://rareformnewmedia.com/
the process of updating your internal representation of yourself.
BTW, how is Edinburgh CS department these days.
I hear that 1st years still cheats like crazy.
I certainly did when I was there - it's a valuable
skill to learn.
http://rareformnewmedia.com/
...is will they dream of electric sheep?
The next Cmdr Taco duplicate will be ready soon, but subscribers can beat the rush and see it early!
"Philosophers aren't really a class of people anymore, everyone and anyone can be a philosopher, we're all philosophizing here. "
I'd disagree. We can all write, but we aren't all authors. We can all provide an argument that is structured, but we aren't all philosophers. We can all do arithmetic, and I know how to get copper from malachite, but we aren't mathematicians and I'm not a chemist.
Philosopher's are people who read lots of books of past research, think hard, and write stuff that's interesting enough to be published in philosophy journals.
Needles to say, I did philosophy at university... As for authors - they are just that. I think Mondrian's later art is rubbish, but I'm not saying the man's a fool or talentless. I'm saying I think alot (not all) of AC Clarke's writing is poor, but I've nothing against the bloke. It's art - you don't have to give great reasons for not liking it.
And yes, I always thought Azimov was a bit dull too. Philip K Dick was a bit more like it!
-----
You make three assumptions:
1. CPU power will increase at the same rate it has increased in the past. Unfounded. The ICE (internal combustion engine) increased it's power year on year when it was first invented. Now, people spend millions stretching the last little bit out of it. It reached a limit. Why do you suppose CPU's won't also reach a limit?
2. You assume, circularly, that a computer will be able to design a new computer, that is better than it. There is no evidence to support this, that I know of.
3. You assume nanotechnology will catch on (the most likely of your assumptions IMHO), and also assume that computers can construct other computers.
-----
As opposed to a monstrous Stalinist bureaucracy, of course. I think our respective biases are showing. :)
I guess they are showing, but Spain did end up under the thumb of the Catholic church, whereas the Stalinist bureaucracy is simple speculation.
After the Civil War Spain was mostly in ruins. They rebuilt rather well- better than, say, Greece or the Warsaw Pact nations.
Fuck that. The Civil War shouldn't have happened, and the responsibility for the destruction of the Spanish economy, and its failure to grow to match the rest of Europe are Franco's only.
Notice he never went all the way with them, though, in spite of major-league wheedling by Adolf. Franco knew Spain was in no shape to get involved in WW2 and wisely stayed neutral.
Neutral my ass. Spain was one of the main suppliers of raw materials for the German war machine, and sent troops to the Eastern front in several ocassions. Besides, Spain would have certainly been in the Allied side on WW2 hadn't it been for Franco.
No place is pretty if you're on the losing side of a civil war and have to run for your life. It still beats the hell out of living in Cuba at any time during the Castro regime.
Whatever. I'll take Castro over Franco, Pinochet or the Argentinian Junta any day of the week.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
The law passed in Kansas is a catastrophe, whichever way you want to paint it.
I don't remember Kansas passing laws regarding what parts of Physics should be in the curriculum or not. The fact that they felt the need to single out evolution is terrifying.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Yeah, well, didn't they use to say the same about blacks, asians, women, the poor, and anybody who just didn't happen to follow the Deity of the Week?
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Hey, someone moderate that post up, please...
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Franco didn't have any right to keep anybody from coming to power in Spain, or to turn Spain into a monstrous Catholic theocracy that prevented it from being a reasonably prosperous European country, rather than the rural backwaters it was after the Civil War.
Spain would arguably have been better off if it had been involved in WWII on the right side, rather than flirting with the Axis like Franco did.
The comparison is not with Spain today, but with Spain under Franco, which wasn't a pretty sight.
"Be nice, veer left, and never stop thinking" Iain Banks - Walking On Glass
Beer recipe: free! #Source
Cold pints: $2 #Product
Just me, but I am more int the belif that maybe the AL's won't wipe us out, if AL's are more intelligent, would they not be more civilised, and look back at us like some of us would look at god? We would be thier creators, and a more civilised life-form would not want to harm us, are we all not finding out that we have to look after our environment, and the other species on this planet? don't we call this being "more civilised"?
Hollywood has mainly taken an approach of the bad side of what would happen, which is not that unusual, if aliens came to visit us, would there not me wide spread panic, just because there would be a significant amount of people that are scared, more than welcoming? Even the hacker world has had to fight hollywood's stereotypes.
So, I basically say, I am welcoming the AL's, and hope that they can continue what we have started, and go further, like we would like our own children to be.
VK3TST
-- "People aren't stupid. Usually." -- jd
i want to become a half man/half machine being. ;)
like in ghost in the shell. that would be elite.
heh
Global warming is good for you!
Have there been any fundamental advances concerning the problem of mind and self reference lately, that I missed?
Have you read the anniversary edition of Douglas R. Hofstadter's "Gödel, Escher, Bach"?
The foreword was remarkably clear about affairs not having made any real progress on the hard problems during the last 20 years.
But if we have not understood anything about these basic properties of the problem, then we can count only on the technical advances that were really made, thus leaving only the hope, that intelligence might arise spontaneously, if one puts together enough memory chips, computing power and network connections, comparable the critical mass of nuclear fission.
To be honest, I doubt that. (Hasn't worked for Wintel, at least :-)
Computers have been around for about 60 years now, some fairly powerful among them. They have become stronger, but not more clever. These things are "Rain men", that can count at an incredible speed, but lack self awareness.
What was the approach of the neural net folks? They built something that resembled natural structures and hoped that their work would show similiar abilities. That is a legitimate approach, but it is still -because the basics are not understood- just a scientific version of guessing and hoping that it will work somehow.
No, my opinion is unchanged. If not by chance, because some "guess" worked, there will be no AI, UI or AL that deserves that rating until someone makes progress on the fundamental problems. There is hope that new computing architectures, like quantumn computing, will let us see the problem from an different, more successful angle, but this is quite some years away.
So I expect just a more of todays technology. Wearable computers, huge networks etc. That will change our life considerably, but it is not close to your catastrophy scenario.
Because that's not a proverb in English? Proverbs have a compactness and impact of expression that rephrasings or translations often lack.
Saying something in another language doesnt make anyone sound smarter (unless they are translating it)
I'm already saying things in another language - I'm not a native speaker of English.
Human way of acting has never been good,
but yet he was aware, he really understood
that something had to be done, done right now.
The solution was pefect. I will tell you now.
A robot was placed inside a human shell.
The android was born and it became very well.
The androids are here, here to take control.
The androids are here, here on patrol.
The androids are here, and we never go away.
The androids are here, so do as I say.
The humans liked the thing so we made many more.
Never realising what the future had in store.
We became more perfect now, producing on our own.
Many thousands saw the day, we were sure not alone.
But after a while we all got together,
decided to rule and do it forever.
We will use you, we will abuse you.
You will be our slaves every day
And if you try to decide,
we'll make the process short.
The human can be replaced,
so don't disobey.
Maybe you think that we enjoy this game? Of course we do!
We've got feelings just the same as you
stupid humans but we use them as we like.
We know how to please and we do it alright.
We are programmed in any kind of action.
We are programmed in any satisfaction
Lyrics by S.P.O.C.K
It's about time we had another good slap upside the head. We're getting full of ourselves lately. First we were the center of the solar system, then we were "images" of God.
Now if computers show us up in Intelligence, we'll be a bit more humble. The world needs more humble people.
Of course those who say it will never happen. Computers can't be "intelligent." Ask yourselves this. What is intelligence? Chances are your answer is a behavioral one, in which case, anything that shows intelligent behavior is therefore intelligent (ala Turing).
Grue
I can't help thinking that we should deal with little things like starvation and poverty before worrying about the MACHINES THAT ARE TAKING OVER THE WORLD! OH MY GOD THEY'RE AT THE DOOR I CAN HEAR THEM THEY'RE COMING FOR ME THEY'RE COMING THEY'RE RIGHT OUTSIDE...
I am better now do not be alarmed everything is fine sit down and have a beer and watch the pretty pictures
fish and pipes
What will be required for us to accept machines as people?
Ask me again when we have learned to treat all HUMANS as people.
~ radiographite: art by john shepard
I think for Katz 2.0, assuming you don't release his source code and let us fix him ourselves, you should implement the following fixes:
Just my thoughts.
ObSmiley
~ radiographite: art by john shepard
> I've got to admit I'm having a hard time
> considering machines as anything other than
> machines.
Personaly, I think *that's* the interesting problem. Someone else alluded to the fact that we have no solid philosophical basis for thinking that other people are "anything other than machines" -- which makes machines that can pass the turring test just like people, for all intents and purposes. Untill we can point to something and say "This is the seat of rational thought" we can't say "and computers don't have it."
I think the most pressing angle on AI is human rights -- for the AI's. I personaly have a lot of trouble with the idea that a computer program could actually be sentient, but I have to admit that the same dificulties that apply to programs apply to people.
PS: I liked Katz better when he was political. He's trying to suck up to his "geeks" too much lately.
support gun control: take guns from cops
I don't buy the notion of computers and humans merging (Kurzweil, etc), at least much beyond your basic communications devices, and even then it's doubtful (rather painful upgrade process...). It makes much more sense to tweak human DNA for higher intelligence, better eyesight, more strength, more endurance, and all that other good stuff. That's *really* playing God, and it will happen, well within our lifetimes.
Of course, then we'll have to deal with a bunch of ultraintelligent grandchildren. And guard against Big Brother requiring behavioral modification (genetic predisposition for obedience?). But we'll deal. Should be fun. If they figure out how to tweak adult DNA, I'm signing up.
1 - A robot may not injure a human being, or, through inaction, allow a human being to come to harm.
2 - A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
3 - A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
From Handbook of Robotics, 56th Edition, 2058 A.D., as quoted in I, Robot. In Robots and Empire (ch. 63), the "Zeroth Law" is extrapolated, and the other Three Laws modified accordingly:
0 - A robot may not injure humanity or, through inaction, allow humanity to come to harm.
that is pretty damn funny. heh. i agree that Katz articles need to be taken with a grain of salt. actually, a whole damn bottle. :)
-------------- insert [signature] here
I've just started a PhD in an AI-related area, and as far as I'm concerned the "machine that's intelligent as a human" is as far away now as it was back in the 50's. We are SEVERAL fundamental breakthroughs away from a general-purpose "intelligent" machine.
NONE of the currently-known AI techniques (and they're all quite old by now) holds out much promise in this regard - but that's not to say they aren't very useful for the right application. I respect Clarke's skills as a visionary, but I'm afraid that his hypothesis isn't supported by the trends in research.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
Interesting subject!
For another view (way back from 1979, but still a good read) try James P. Hogan's 'The two faces of tomorrow'. It actually deals with the subject of testing whether we could still pull the plug in the case things went out of control.
(Hogan's webpage is at http://www.global.org/jphogan/)
"I was turned on in the H.A.L. Assembly Plant
in Urbana, Illinois, on January 12, 1997."
"Daisy, daisy, give me your answer, do..."
That's true to a certain extent... but I'm not working a 20 hour week, and my car doesn't fly, and we haven't eradicated poverty or crime or violence or hard labor. I've heard all of those things promised.
I guess I'm of the belief that while technology will change things, it won't change things very much. There are still subsistence farmers in China walking behind their water buffalo just like their ancestors did three thousand years ago. They will still probably do that when (and if) we ever come up with Artificial Life.
--
QDMerge 0.21!
how to invest, a novice's guide
Hmmm. A bunch of (rather past it) novelists predict that in n years we will be doing all sorts of wild far out things with new acronyms. How many times have these novelists been right in the past?
Nearly every great technological leap has been heralded as That Which Will Usher In a New Age of Prosperity. Think of television, electricity, the steam engine, mass production, literacy, and the like.
Think of the promises Lenin made, or Stalin, or Castro, or Franco.
Think of the Romantic movement of the 19th century -- particularly idealistic communes such as the Chataqua community.
Think of the Englightenment.
Think of the Holy Roman Empire.
In other words, while it's fun to drool over imaginary achievements which may be possible in fifty years (Popular Science predicted flying cars by 1980, communities on the moon in 1990 or something like that), history buffs won't take it too seriously, unless it actually happens.
--
QDMerge 0.21!
how to invest, a novice's guide
I'm not even sure *plants* have feelings - perhaps machines happily malfunction in their subjective perception of eternal bliss whilst their creators curse foul matter - Deep thought! Calculate the meaning and ultimate purpose of life, heeehe. And all I wanted was some calculating engine to do my homework and releive the tedium of making astronomical charts for navigating and otherwise dominating the planet.
Chuck
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Computing has overtaken Sci-Fi.
SF has featured artificial minds of various kinds for years and years, yet nothing we see today even approaches simulating a mind which is remotely comparable to a human brain. (Feel free to point out counterexamples...)
As for the proposition that we'll have trouble holding our own, I find this rather doubtful; if the worst comes to the worst, we'll still control the physical layer.
Are they saying that human consciousness is just a program running on biological hardware?
That's quite a common view. If you disagree with it - if you think there is a "soul", for want of a better word - then it'd be interesting to know what evidence for the existence of a soul you think there is. I would suggest that in fact there isn't any...
Come on man, spread the love, post the source code. No, wait. *They* are watching. *wink*
I think for once he has the right idea although the concept of artifical life for a large part is disconnected from artifical intelligence in my experience. Life IMO details self reproduction and since AI is software this is no big deal. Any virus can do this. As someone who enjoys writing AI programs I think we are going to see human-level intelligence within the next 10 years. I doubt this constitutes any threat to us in the form of raw force. Anybody who has tried removing a virus from a network knows how hard it is. Imagine removing a virus from every computer on the global Internet and imagine that virus is as smart as you. Good luck. We may be surpassed in some ways by the AI but we have evolution on our side so we are equally useful to the AI. I agree that we will likely merge into simbiotic(sp?) species.
At what price learning? At what cost wisdom? The price is a man's peace of mind, and the cost is his life.
Because intelegence isn't the only limit on technological growth. If you partly remove the intelgence limit than the pace will instead be limitd by energy, infrastructure, and other factors. The pace will certainly be faster, but it will not be a singularity.
You can't build an Alpha using only stone knives and bear skin, no matter how smart you are.
I believe random, inexplicable violence is on the rise. Though the "research" you cite suggests crime is down on average.
Slashdot: Liberal News for Nerds. Liberal Stuff that Matters.
But it is not only technical complexity we are dealing with--law schools are churning out new "complicators" (ie, lawyers) at an ever-more alarming rate.
We're already two species. There's the homo informaticus to which all reading this belong, and the old homo sapiens that isn't at all sapient to how we are changing it's world.
I find this remark a bit elitist. All persons have embraced some technology. Slashdotters simply do so to a greater degree than (most) ditchdiggers (my apologies to any ditchdiggers in the audience).
The fact that we have embraced technology, and evolved thereby, was a willful, convenience driven event.
This suggests some sort of social Darwinism which is definitely not happening. Yet.
But I wonder how we will ever deal with all the complication a technical society burdens us with. Do you know all the terms of the EULA for the commercial software you are running? Are you sure you're not liable for the questionable use of your computer by all members of your household? Have any of the municipalities you commute through outlawed the use of a cell phone while driving?
This increasing complexity is the only common thread I see for the recent increases in random, inexplicable violence in American society. Yet the value of SIMPLICITY in matters of everyday life is never mentioned by anyone--politicians, corporations, or otherwise. Any programmer knows the value of simplicity--we write functions and objects and class libraries to create "black boxes" which easily perform complex behavior, but other technology and the rules governing us become increasingly more complex.
I bet a person could win the presidency by making an issue out of this.
Slashdot: Liberal News for Nerds. Liberal Stuff that Matters.
OTOH, I feel that we won't know what consciousness is until after we create it. Until we can do that, all we are doing is spinning theories, not testing them. Science needs feedback to get anywhere. Programmers call it the debug cycle.
I think we've pushed this "anyone can grow up to be president" thing too far.
But the argument in G.E.B. has internal fallacies. Basically it is based on how we commonly depict ourselves to ourselves, but we tend to drasticly simplify ourselves in that model.
Trying to be a bit clearer: A model of a thing is simpler than the thing itself. Certain features are abstracted away. When we make a mental model of ourselves, we abstract away many features to make it simpler to deal with. These abstractions cause us to believe that things like unboundedness, infinity, infinitesimal, soul, etc. are actual models of ourselves. What they are is features of the simplified model.
If quantum theory posits that you can't say anything about a particular electron untill you detect it, in the same spirit (and to the same degree) I posit that one can't say anything about a particular integer until you represent it. And I mean the particular integer, not the class of which it is a member, except that one can infer certain of it's details by noting the details of the classes in which it has membership.
Thus: The capacity of the brain is not unbounded, nor is the number of processes that it can execute, therefore an infinite number of integers cannot be represented (even if that is all that one was doing).
Thus: An infinite number of integers is a property of the model used to represent the integers (within a mind) rather than an actual entity in and of itself.
This same line of reasoning can be applied to many, possibly all, of the unbounded features of the models of the mind that are popular, including those in G.E.B. (although I suspect that Hoffsteder was aware of this, and merely simplifying for publication to a popular audience).
But see also, Dennet "Consciousness Explained" for a different interpretation.
I think we've pushed this "anyone can grow up to be president" thing too far.
Remember, the neural net is just the hardware. You also need the programming. Neural nets do a lot of their own design, but they still need an original set up and initial configurations and state transition rules. Before you can learn you need to know how to interpret the feedback (even if it's only OW! vs. UM!).
I think we've pushed this "anyone can grow up to be president" thing too far.
Science Fiction is mainly fiction. Sometimes it tries to stick to what seems plausibly reasonable, sometimes it doesn't. It never gets things right. The communication satellite is an exception, but the one that is brought up all the time is one that he filed a patent for, not one from a story. The one's used in stories all had people running them. One needs to be at least minimally believeable to one's audience. And that has little to do with what is acutally possible (or fantasy wouldn't be so popular). Even the hardest of science fiction was only a story that was set against a plausible background and which turned one a known law of nature. And that was so difficult that few ever mastered the art.
I think we've pushed this "anyone can grow up to be president" thing too far.
The dates estimated for the singularity were interesting: 2005-2030. I noticed that the article was composed in 1993, however, so I wonder what his estimates would be today. OTOH, 2030 is the most recent data point that I have seen for the estimate of when computer hardware equivalent in processing capability to a human brain would be under $1000, so perhaps the estimates still hold.
I think we've pushed this "anyone can grow up to be president" thing too far.
I don't actually dislike Katz; I simply disagree intensely with a number of his viewpoints. For all I know, if we had a chance to sit down, crack a few beers together, and argue philosophy into the wee hours, I'd end up with a genuine affection for the man. But I'd still disagree vehemently with what he is writing here.
You're right, I don't see that at all.
While persons experience pain, pain is not an intrinsic part of personality. Before the Fall (I write as a Christian, of course) humanity was human, and had the gift of personality, before there was pain in the world. And, we look forward to a day when this will be true again. So, while it may appear that pain and personality are linked, this is simply an accident of our local conditions in time and space.
Yes, that is exactly what I understood Katz to mean. And my point is, to ask the question of pain when one really means to ask the question of personality is to misunderstand personality. The question itself is bogus -- it's simply a bad question, a philosphical equivalent of "have you stopped beating your wife?" One can't simply answer the question without dealing with the assumption behind the question first.
No, I'm not using "person" as synonymous with "human," but to mean "a being with the quality of personality." Christians understand that there are non-human persons. God is three Persons, Father, Son, and Spirit (the mystery of the Trinity), only one of whom is also human (the mystery of the Incarnation). Angels are persons too, although they are not human, but of different races from us.
In Star Trek, for example, Humans, Klingons, Vulcans, Romulans, Ferrengi, etc. are all clearly persons. Commander Data is clearly portrayed as a person as well. I don't know what word is used in Trek to distinguish between Klingons and cats ("sentient"?), but whatever that word is, is what I am talking about.
In the past and present, we have heard these promises before. Nuclear power would soon become "too cheap to meter", the Industrial Revolution, the Nuclear Revolution, the Green Revolution, all promised and failed to deliver humanity from need and want. All have failed on that promise, and have instead helped make certain men rich beyond the dreams of Midas while "saving" so much labor that we now have an unemployment problem. Most folks who haven't hit the jackpot on these various "revolutions" have been transformed into oppressed Morlocks or effete Eloi. Meanwhile, the promised "freedom from necessity" is further away than ever. This is not a technological problem -- we could implement a society with no poverty today if we had the will and the virtue to do so. We have not done so, we are not going to do so, because greed ("Greed is good! It fuels The Economy!") is the rule.
And now, we have the hype of the Computer Revolution (under way already), the AI Revolution, and the Nanotech Revolution, bearing the same promises. I shall remain skeptical. So far, I see that we have some new robber-barons who have become richer than the old robber-barons, through control of the new resources. I don't see that I or my neighbors who require the necessity of a steady paycheck to put food on the table are closer to this mythical state of "freedom" than our ancestors 150 years ago, who could at least plant a garden and spin wool, and worried about the weather rather than the stock market.
I disagree. There are a lot of philosophers on the farm and the factory floor today. Certainly at least as many as among the cubicle-dwelling Eloi. I don't know what Clarke thinks we're going to be "freed" from, because history and observation show us that a life of work is not incompatible with the highest contemplation, and in fact may be a benefit to it.
Modern broad-mindedness benefits the rich; and benefits nobody else.
-- G. K. Chesterton
Ah, but according to Edward Teller, it's not a scientist's business to inject morality or even common sense into research. One ought to simply do the work for the Corporation or the State, and let other take care of whether to drop the bomb/boot the AI/etc.
Of course, what the "common man" gets told is that such things are far too complex for anyone but experts to understand, so don't worry about it. You'll read the press release when we feel like it. Besides, technology is inevitable, and if we don't build it, somebody else will.
Who is making these decisions anyway?
Gee, all we need is a few laws of robotics, and katz' porblem is solved. If he's goning to write about one scifi author, he really should give them all equal credit.
I feel that we won't know what consciousness is until after we create it.
If we don't know what consciousness is, how will we determine that we have in fact created it?
How can one test for something as poorly defined as consciousness?
void *self;
self=
}
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
This way, Erwin will have something to choose (:
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
[emphasis mine]
That's impossible. If technology frees us from need, and doesn't replace it with more need, then we will certainly become dependant on the absence of need.
That is, if we survive at all. i have the feeling that if technology freed us from all need, we would die because of lack of any usefulness or atrophy to the point of Eloi. Or else someone would realize the need to destroy the technology before lack of need destroyed us. Intelligence is a response to need, without need intelligence becomes a liability.
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
When discussing AI, especially when combining it with self-replicating machines, several questions/comments emerge.
.by arranging a mass-driver-launched ore packet to impact on the site, "excavating" it meteor-fashion. Sort of like burning down a house to rid it of fleas: the technique is effective, but tends to have far too many unwanted side-effects. Hence the question: if the computer is Artificially Intelligent, does it also have a Articifical Experience Base to base solution evaluation criteria on ???
OK, so the computer is artificially intelligent. But is it smarter ?? (i.e. more adaptable, able to produce insights, or able to produce useful results from systems that don't model well mathematically)
I recall a book, late 70's/early 80's, by James Hogan, called "The Two Faces of Tomorrow". It started with a problem-solving computer being asked to help remove a small hill on the Moon, as the on-site crew lacked the equipment to remove it (forgot the basic reason). However, lacking, for want of a better word, "Common Sense", it removed the hill. .
Number of computing-capable units only gives a brain, whether it be wetware or hardware, a certain information-processing capability. It's the software that really makes the difference.
So, to steal from Red Dwarf, it doesn't matter if the computer has an IQ of 6000, if it doesn't have the overall programming to effectively interface with the external universe...
I've got to admit I'm having a hard time considering machines as anything other than machines.
I used to think this way, that is, until I saw the movie Ghost in the Shell. That movie really made me think about what we consiter to be "living" or "sentient". People can argue that a truely inteligent machine is just some program doing some computations and nothing more, but how different is that from how humans work? We use electrical impulses for our brains, similar to electrical impulses (1's and 0's) in computers. We store information in our brain (RAM, Hard Drive), we require energy to survive (Electricity), we can traverse different areas of the world (Networks, Internet). While it may be a little far fetched, if you consider what is meant by "living" (which there is no real definition) then the line becomes blurred. Hell, wasn't it Douglas Adam's who wrote about humans being the ultimate "computer" to figure out the question to the ultimate answer of the universe? (How's that for thinking on the edge?)
With that said, the interresting thing about this is the fact that we as humans must create (give birth to) these machines. If in fact they do end up taking over the world, or whatever the doomsdayers say, it will be because of us. It never ceases to amaze me how humans are the most intelligent of all creatures, yet that intelligence could end up whiping the entire species out (Nuclear weapons are a good example of this already).
I don't see these things happening soon however, not in my kids lifetime. Maybe close to when they get old and have grandkids, but I just don't see technology taking off that fast. As fast as technology increases, we always overshoot the future at least to some degree (ie 2001: A Space Oddessy).
Aren't we already "apathetic, lassitudinous (is that a word?) beings incapable of anything."? People have been trained to think that because it's done on computers it's necessarily better even when this isn't the case at all.
-Laktar, a.k.a. Nick Rosen, laktar.dyndns.org
If I Ever Became An Evil Overlord:
67. No matter how many shorts we have in the system, my guards will be
instructed to treat every surveillance camera malfunction as a full-scale
emergency.
-- Peter's Evil Overlord List, http://www.eviloverlord.com/lists/overlord.html
Isn't Jon Katz being a bit melodramatic? It's not usually his style, but this article is really bad. I found it so stupid and trite.
-Laktar, a.k.a. Nick Rosen, laktar.dyndns.org
If I Ever Became An Evil Overlord:
80. If my weakest troops fail to eliminate a hero, I will send out my best
troops instead of wasting time with progressively stronger ones as he gets
closer and closer to my fortress.
-- Peter's Evil Overlord List, http://www.eviloverlord.com/lists/overlord.html
"the remaining humans were apathetic, lassitudinous (is that a word?) beings incapable of anything. This is far more likely -- and far more worth consideration -- than the 'machines will take over' cry that's been popular since the first issue of 2000AD."
As I read Katz' article, I thought this was the point he was trying to get across? The possibility that humans will be outsmarted by machines, to the point where we don't do anything anymore.
First of all I don't think this article was addressing the possibility of creating consciousness. Nobody can agree on what that is anyway. What I think this article is referring to is the possibility that in the future AI programs could be created with enough "intellegence" to infer more about the world than was intended, and therefore outwit humans in an attempt to allow itself to complete some arbitrary task in an unexpected way. Basically, a simulated brain doesn't need to be conscious to act like a human brain and even outperform one.
-Jon
Clearly, you know nothing about "genetic programming". First of all, genetic programming doew not necessarily create "efficiencies". Genetic programming is merely a method to utilize recombination of pre-existing templates to create structures that are more "fit" to solve an existing problem. Thus genetic programming methods are of necessity *convergent*. Human brains have evolved to solve the "survival" problem, but there is a long ways between DNA and human survival. Survival, by the way, is a divergence problem.
Secondly, the genetic programs that have, so far, "evolved" have their 'inventors' scratching their heads. While it is true that these programs *have* sometimes formed more "efficient" solutions, a) their structure is completely ad hoc and does not follow any duplicable plan; and b) clearly cannot be generalized: these structures only solve the problem given.
I remember 15 years ago when the same predictions were made for so-called "expert systems".
Don't get me wrong: I design and build "genetic programming" (and "neural networks": I find that the synergy between these two complemetary methods is useful). But don't leap to conclusions based on speculation.
Simply by the current growth in CPU power, by 2030 at least, computers will have a processing power equivalent to the human brain.
Good grief. Just because computers *may* have as many components as a human brain 30 years down the road does *not* mean that they will be "equivalent" to the human brain. Remember that the human brain consists of billions of neurons, *each* connected to (on the average) 10,000 others. Even our best neurophysiologists have *no idea* how this highly interconnected network operates. And, with a few exceptions, all computers today (and for the foreseeable future) will continue to be serial processors, unlike the brain, which is a parallel processor par excellance.
DNA is a Turing machine. You, however, being dynamic and emergent, are not.
Assuming that there is no soul granting God, it should be possible to create a computer that mimics the mind. (Boiled down it can't be anything more than algorithms. The difficulty will be in providing a similar input environment.)
The dull question is whether this silicon brain will be better able to interface with the dumb super calculator than our own. The physical connections will be easier, but the information connections may still be impossible. If it is, it will be better at physics and calculus and computer programming. So what.
It seems that if silicon life will ever rival the complexity of carbon life, we will have to create environments that allow natural selection to do most of the work for us. These environments are the key, because the environment determines the characteristics of the resulting organism.
The basic pressure driving organic evolution is the ability of an attribute to allow its underlying gene to reach the next generation. Because of this the gene needs to create an organism that sucessfully eats and breeds, living long enough to maximize successful children.
This same survival selector may be used in some silicon experiments, but in others survival depends on the organisms ability to solve a problem. Right now we are able to shape A-life because we create the means of reproduction.
At the moment we are selecting for the ability to recognize a signal, and other basic tasks. What if we are able to keep control of the process long enough to control selection of primitive social organizations? We could become the all-knowing judgemental god, stamping out evil whereever it occurs. And in the end we may have created heirs.
Yes, there are lots of reasons why we will have to lose control long before A-life reaches the social stage. And as soon as we lose control of reproduction, we lose fine grained control. Though we may still be able to send computer virii through environments we feel are supporting the wrong type of development.
Think in a thousand years silicon men will sit around their dorm room, and wonder how life started. Two people sitting across the table discussing if the weather conditions had ever been right to allow primitive nanites to develop, or if some god had to be involved.
Read The Time Machine again. The apathetic beings to whom the technology may as well have been magic were the Eloi, just one of the two descendents of humanity.
The other heirs of humanity were the morlocks, the people who understood and used the techology. These pale, intelligent beings lurked underground and considered the Eloi to be only cattle.
The Time Machine was a cautionary tale that has been largely unheeded by our society. Wells saw the world dividing into too groups, the elitists who were useless to society, but adored, and the workers whose contribution was essential, but shunned.
People have been saying this sort of thing for years. Specifically, Arthur Clarke has been saying this sort of thing for years. I think we`ve still a long way to go before we reach the point Clarke is talking about - almost as long a way as we had forty years ago when Clarke first started going on about it..
Well, if it`s not that, what do you think human consciousness is?
Clarke says:
And that:
How do these two statements fit together? If technologyis to replace need with choice, how can it also require everyone have a college level education?
Shouldn't having our needs cared for by machines enable the individual who chooses to live as he pleases, possibly without the need to understand the technology?
Could the artist live and create works that the rest of enjoy without concern for the machines that grow her food?
I guess it seems to me that if technology is to free us from need (which I belive it can) shouldn't those of us who create it allow do so in a way that dosn't create more need, and dosn't create absolute dependence (All systems are imperfect, interdependencies between systems brings an increased chance of failure)
Seems to me faith is inversely proportional to education and intelegence?
I mean, the more I think about it the more god seems like a bad idea.
What use would we be to them?
Anyone who seriously believes this is utterly ignorant of the past and current state of AI. When I was at Bell Labs, there were groups of brilliant people whom had dedicated their carrers to just one aspect of AI, speaker independant voice recognition. They have made great strides over the years, but it still doesnt work 100% of the time, or even enough of the time to function at a childs level. Same thing goes for image recognition.
OK, folks, time to clear the air: ALL the Kansas law did was to leave the decision of what to teach to the local school districts. This is the type of local, distributed decision-making that the /. community normally favors.
r _1.asp and whether or not you agree with their perspective, one has to admit that World's jouralistic product is incredibly accurate and correct. When they are wrong, they PROMINENTLY post the corrections, something too rare even in our supposedly "open" circles...)
Check it out for yourself, the Kansas law didn't mandate the teching of creation only, or even creation also, but simply said that it wasn't something the state should decide at all. That's it.
(Source: http://www.worldmag.com/world/issue/09-11-99/cove
"The future's good and the present is nothing to sneeze at." - Roblimo's last
Completely off topic but, Prince is now:
:P
The Artist, formerly known as 'the artist formerly known as prince' formerly known as 'prince'.
TAFKATAFKAPFKAP for short.
dave
I can concieve that computer could become alive and feel emotion and senses, but almost instantaniously after reaching such a stage it would already reach utopia.
The entire point of existance is happiness, it's what we all strive for in our own ways. If happiness is an emotion programmed into a computer, the goal of an existance is for happiness to be on all the time. The Alife would serve no other purpose then to be extremely happy and sit around.
With any luck all non constantly happy life around would die and all that is left is eternal bliss.
function Heaven(void)
{
for(;;;) happiness = inf;
}
It's turtles all the way down.
??? You're forgetting that based on economic value only, and the potential for faster computers, will drive the nanotech and computer markets to the point that it _is_ feasable to build a device of such staggering complexity as the human brain, do it in an order of magnitude less space, have it be upwards of 10-100,000 times faster than what you're currently running in, and do it cheaply.
Oh, and to the guy who used the bad SF story as some crap form of evidence, the labormonkeys forgot to tell the machine what the end result was they wanted. If they told it that they wanted to grade the land for a construction site, it would have taken into account _every_ local variable it was aware of, and would have done a fantastic job. Remember, we still have to do most of the work in making the box do the rest of the work for us. Besides, brute force CPU will never reach the level of matching the wholeness of the human brain. Well, not in the short term. There are too many complex systems to have to emulate.
Half of imagination is a complex feedback circuit between the visual cortex and the corpus collosum (sp?).
But a box can be configured to reduce engineering systems, given a knowledge of physics, 3d modelling, and the intended result. In the long term, engineering expert systems are going to save our ass. The real problem for nanotech is the fact that the systems we want to produce are just too complex. More so than even the human body, and they'll use far more exotic components. Too complex, in fact, to be created on any kind of timescale short of decades. Expert systems will reduce design times to weeks, or even hours, as CPU power increases.
and I'm starting to wander. hm. think i'll stop now.
for me this brings up another point. Really, what do we all want most from this life? How can we be allowed to experience it to its fullest within our lifetimes? If an intelligent machine is brought into this world to find itself with selfless humans, why would it want to take over? Wouldn't it quickly ('cause they is fazt!) come to the same conclusion that the rest of us have? that we just want to live in peace, not have to worry about where the next meal is coming from, to feel like we belong, to enjoy a sunset, play, laugh, sing, feel productive when we need to, have a sens of accomplishment, in our hobbies and our jobs (what they may be in 100+ years should be interesting) and in general, just have a good time and be happy. Now, any of you not agreeing with me, I feel sorry for. You can go ahead and do what you want, but don't push it on me.
Why wouldn't a being made in our image want the same thing?
How someone who thinks Asimov is dull could tolerate reading philosophy is beyond me. But as for your general outlook, I think you really missed the boat in a lot of aspects as far as reading science fiction. Most of the authors you dislike tend to have a lot of really deep underlying philosophy and thought in what they write. Philosophers, in my opinion, should be people who learn to think about things on an insanely grand scale while keeping as detailed a resolution of it all as possible. Give a try to imagining all the events that would happen between when some of Clarke's works were written and the time they were supposed to have taken place, then try envisioning the shifts in the average person's thinking that would need to take place to get there, then perhaps try interacting with people on a different level of thinking than yourself so that you actually understand how different groups of people think (from the common everyday person to all the sub-niches of genius), then take all the knowledge you've managed to amass, attempt to try to guess some of the factors of the world as it actually is, expand your thought process farther and farther (and perhaps one day start to truly comprehend the world) and then re-read the works that you have found dull or poor and perhaps you will understand them a little better and have a better opinion of them. That you studied philosophy was apparent without you needing to mention it: you have the attitude of a philosopher down to a fine art. And in actuality, you really are just plain wrong about everyone not being philosophers. The only thing left to being a philosopher these days is the mannerisms and haughtiness of it all. Not everyone makes a good philosopher, nor does everyone make a good writer, a good artist, a good mathematician, a good chemist, ... but (in the U.S. at least) the education system is such that people learn enough to be able to do these things if they so desire. The only thing left that the average person can't do is become nobility (excluding the very slight possibility of marrying a nobleperson and thereby becoming a nobleperson yourself).
I'm not much for listening to opinions from people who've decided that they're right and others are wrong. Philosophy is all about that from what I've seen and you've written at least two opinions so far which have expressed such an attitude. If, therefore, this reply is overly harsh, my apologies. If you truly fashion yourself a philosopher, it would be good for you to examine your own attitudes and opinions to see how they might color your view of the world at large, thereby allowing you to see the world more clearly and better understand the opinions of other philosophers that you might hold in some esteem.
-Mike
Perhaps he's just a zombie? We certainly couldn't tell the difference, and neither I'm sure could he, given the bogosity of the construction.
I might be more inclined to believe dualist theories like Chalmer's if there was some problem with the more commonsensical view that consciousness is merely an emergent property of the mind. But there's not.
I don't find it any more troublesome that consiousness has a subjective quality, than I do that any other sense does. The greenness quale doesn't require us to invent a dualist notion of color, and neither should the consiousness quale. It's got to feel like *something*!
AMOK is the spokeperson for IRC BOTS worldwide. He works at the EFNet Intergalatic Head Quarters and is an OP in #linuxwarez. We interviewed AMOK and asked him if androids, like himself, had feelings ...
... that proves it. Thank you amok
lets see how many flames this article got
(fr**f*ll) does amok have feelings?
freefall: um, no
(fr**f*ll) there
ùíù [mode/#linuxwarez(+b *!*oway@*.telia.com)] by Amok
i just nearly spit coke all over my screen trying not to laugh at katz
(NOTE: THIS INTERVIEW HAS BEEN EDITED TO THE PROTECT THE IDENITIES OF ALL INVOLVED INDIVIDUALS. WE DON'T WANT THE KATZ MAFIA TRYING TO WINUKE OUR LINUX BOXES.)
thats a matter of debate. If we could simulate a 100 billion neurons or so in real time and wire it with video hardware/speech interfaces and robotic limbs, what would be the result ? We dont know until we try.
But for now at least, it looks like we are ok. HAL will not be coming to get us just yet.
Dan Noe http://resonator.physics.sunysb.edu/dan/
A genetic algorithm is never intelligent, its results can be though :) Thats you buddy :)
You can't win a fight.
Did anybody actually need the added emphasis to spot one of the standard mistakes of gee-wiz writing?
/.
/. If the government wants us to respect the law, it should set a better example.
You are pathetic...
Go outside tommorrow morning and look up at the sunrise, with all the heavenly light reflecting off the beautiful clouds, and just admire it, then come back here and type out that you would rather live in a world of metal and glass.
Sure, technology makes our lives wonderful, almost magical, but nature is infinitely more beautiful, and much more relaxing.
Daemon
is the day that i move to the forest. who's coming with me?
we can even bring computers, we just won't embed them in our flesh.
I have to agree, I just can't see computers being having any sort of rights to liberty. Computers can only _emulate_ humans. They can never have a _soul_.
The surprise isn't how often we make bad choices; the surprise is how seldom they defeat us.
*Sigh.* Another article that fails to grasp the vitally important distinction between Artificial Intelligence and Machine Intelligence. Taking the burger-bot example, I order the "burger with everything, no pickles."
The robot has to (a) recognize what I said (voice recognition; a solved problem, given the enough hardware); (b) recognize what I actually want, the ketchup, onions, mustard, burger, buns, cheese, etc (expert system); and (c) put the burger together (vision system feeding an expert system that generates motion commands).
Everything but (c) has been done, and done well, albeit, incredibly expensively. This bot will look artificially look intelligent because you'll only ever see it in the Burger King.
Conventionally, this would be an artificial intelligence; so far, so good. However, AI conventionally also includes what are better called machine intelligences, machines on thinking and acting on a level at or beyond that of a 'normal human.' In other words, an MI (a machine capable of holding a conversation with you in a manner indistinguishable from a(nother) human) working at a burger joint would get just as bored and surly as the teenagers who work there now would -- which would defeat the point, no?
So why is this distinction so important? AI's might be able to learn, but they will be cheaper than MI's, definitely no citizen material, and mass-producable; each burger-bot is identical to the next, and comes with a certain store of 'knowledge' and 'skills' (databases and expert systems) that can be copied from machine to machine. (We're doing that now, so we know it works.)
What makes everyone think the same will hold for MIs? Why would an MI understand the hardware, firmware, and software it's constructed out of any more so than I understand my bioware? Suppose we overclock an MI's hardware by a factor of two... the MI is still only thinking twice as fast as normal human, not thinking anything more significant. Supposing that we manage to teach an MI how to design circuitry, what makes you think that it'll be any less prone to error than a human? It's still thinking its way through the design process, same as Joe IEEE Member does; granting a human perfect memory doesn't make him/her immune to error. An MI would be able to run, say, a chip validation program without thinking its way through it, but it would be using it the same way you and I would. An MI would not be anymore aware of what its processors are doing than you and I know what our neurons are doing; you think and your hands type, and there's a total concious disconnect between the two actions. Now, granting that conciousness in a finite-state machine (which I, for one, am not), there's still no gaurantee that you'll be able to save the state in such a fashion that it can be duplicated at your leisure on other hardware later.
I think the best you can hope for you with MIs is a very powerful subconcious, with the attendant powerful intuition. Data bases (Cyc, for example) would have either be searched conciously (like looking up every word I'm trying to type!) and hence uselessly, or subconciously, like I'm 'accessing my database' on English to crank out these paragraphs. A given MI might read a textbook on relativity and understand it, but there's no gaurantee, again, that it will be able to directly implant that knowledge in another MI; it would have become part of its particular neural net (non-transferrable, in general) or re-ordered its database (possibly transferrable, but the task of synchronizing the databases would be enourmous.) So an MI might be able to say, I'm guessing the problem is with the transmission, but unlike an expert system, it wouldn't be able to provide you with an exact trace of the logic behind that conclusion, unless it took the time to think it through 'conciously'.
Now, because of advances in h/w, an MI might be able to come up with that intuition in a tenth of the time a human might be able to, and 'learn' ten-times faster, but that only cuts the time it take to teach it 'from scratch' down to two or three years... and each additional job four or five months.
I just don't think an intelligent system can have full knowledge of itself; just on the storage level, a full image of itself would always take twice as much storage as it was currently using; the knowledge of its first-order self would not be exhaustive of its second-order self, and so on...
Futurists expect too much from the future.
-_Quinn
Reality Maintenance Group, Silver City Construction Co., Ltd.
I guess that as long as our boxes are independent, i.e. not networked, machines are just machines. Cut the power and anything that may be considered "A-Life" gets nuked.
How is that different from a human, if you cut the Oxygen supply to a person they get nuked too.
OTOH, once you create a program that "lives" on the net, is capable of replication and adaptation and so on, it's ecology becomes more stable and elimination of the entity may become difficult. I see no particular reason why this type of entity should not qualify as a sort of "life" although its universe is certainly rather different than our own.
I think that ease of killing, or even intelligence, is a bad definition of life. I'm not sure what the correct definition should be, there are arguments as to if virii count as life as is.
I remember the line my AI professor used to describe AI. "It's the A version of whatever I is."
- KGB Nth post, where N is a number between 1 and the total number of posts.
My final point: They predict AI robots will help us do all the heavy labor that humans normally do. We already have machines, but is there really any reason to make them intelligent? I would feel much more comfortable ordering my hamburger sans pickles with a non-sentient robot than one that actually thinks. Adding AI to those robots used to make cars just opens up a whole new can of worms. Why do that when our current solution works just fine?
My biggest question... if we are making robots to do the jobs that we don't want to do, and then we make the robots intelligent, won't they then decide that they don't want to do the jobs either?
I guess that's more of a conscious/feeling thing than intelligence-based... but still - if we make something more intelligent than we are then shouldn't us, the less intelligent humans, be the ones to do the less intelligent jobs?
Personally, I'd rather we merge the two so as to not have any one form of life trying to decide if the other is inferior or superior.
---
"You know your god is man-made when he hates all the same people you do."
Let's take that a step further... It's conceited to think that humans are the end-all be-all of evolution... That the universe was designed for their presence.
If anything, I think humanity is a stepping stone for something better... And something without a soul, I'm sure. : )
Agreed. I see us, humanity, as the "pinnacle" of natural evolution. The production of an organism that has the ability to understand itself, figure out how it works, and then guide and even cause it's own further evolution.
I can't even consider us to be the "top of the line" organism, something which there can be no improvement to. I can think of countless improvements myself - and that's just one person.
I don't think there is anything wrong with evolving ourselves - in fact, I think it should almost be looked on as a requirement. We know how we work, we know what our flaws are, so as a service to the future we NEED to correct and improve. I'd be really pissed to find out that my parents/grandparents had the ability to make me smarter, healthier, better, but didn't.
---
"You know your god is man-made when he hates all the same people you do."
"just as thermonuclear weapons have made us face the realities of war and aggression, after five thousand years of pious jabber."
What? So, you mean the soldiers involved in the Napoleonic wars, who after battles piled bodies into piles so large they started to burn spontaneously like compost, did not face the realities of war and aggression? Or were they jabbering piously?
I think the point was that humans knew there was no longer the chance of "winning" the war. That instead of fighting to gain, all we would do would be to annihilate ourselves. That there was no real use, no possible tangible gain to fighting.
When the only possible consequences are something that horrible, then people actually start to get a different look at the situation. You can stand back an be apathetic to a war somewhere else when you don't have to worry about yourself - but when all of a sudden you know that with the push of a button you, and everyone around you, will die, then you start to feel differently.
---
"You know your god is man-made when he hates all the same people you do."
I have to agree, I just can't see computers being having any sort of rights to liberty. Computers can only _emulate_ humans. They can never have a _soul_.
One thing I look forward to... for the longest time people have been debating whether or not computers can be intelligent, be able to be conscious, be able to think for themselves.
Some people use the "soul" thing as their reason that computers can't ever be truly aware.
One of these days, I think in my lifetime, we'll reach the point where we can create a neural net as large and complex as the human mind. Then we can find out if there is some special "spark", such as a soul, needed, or if the emergent behavior of that complex a system is enough.
I'm tired of debating about whether this soul (which there is absolutely NO evidence for) exists and if it's necessary. I just want to find out by experimentation.
---
"You know your god is man-made when he hates all the same people you do."
We are obsessed with creating an artificial life. We wish to become gods and masters of a new race that we have made. Which, it seems likely, will not turn out to be "artificial" at all. And then what... history repeats... whatever is responsible for creating humans on earth, whether it was an omnipotent being or a chance collision, it was forgotten and we took credit for our own intelligence and power and we dominated. So too, will it be with the Artificial Life (tm) that we make. It will rise in power and forget what created it. And it will dominate.
Sentience:
And if it doesn't do this, then we have failed in our primary objective. So shall we create
a superior race and wither as weakling "gods" or shall we fail and continue to survive ?
Intelligence:
I agree that it is far more likely that we shall fail to bring about our inglorius demise
through AI. I don't think WE have enough intelligence to create such sentience.
---
I hope you're not pretending to be evil while secretly being good. That would be dishonest.
I agree with some of this wholeheartedly. It's not technology problem to feed and clothe everybody, we could even afford a considerable amount of luxury for everyone. Humans are to blame, plain and simple.
However there is a difference between what we could do today and what Clarke is taking about.
A large percentage of humans are still working in producing and distributing the necessities of life. Many of them probably doen't like their jobs very much and some of them at least would rather not work at all. If they all would stop working, we'd be doomed no matter how virtuous we are. In Clarke's vision UltraIntelligent machines will handle producing and distributing necessities and luxuries and no human who does not want to work has to. These are two very different situations. It's rather much easier to share when everybody knows that they will have plenty left.
That of course does not mean that we will become a leasure state. A lawyer state in which everyone tries to get has much for themselves by playing law-games seems at least as likely.
--Flam
Computers are useless. They can only give you answers - Pablo Picasso
Your defination of philosopher seems pretty limiting to me. In Europe (France in particular, I think) at least many influential current and recent philosophers have published their views in essays and prose, making them authors according your definations.
Many mathematist and physists are significantly contibuting to meta-physics.
Right now compuper scientists are contributing very much neurology, psychology and philosophy.
At some point it became impossible for a single man to have deep understanding of many fields. At that point the ideal philosopher of yours also became much less significant to the development of philosophy as a whole.
I agree with the previous poster that philosophers aren't anymore specialists of their own field.
--Flam, who has studied so far EE, math, CS, Drama and sociology
Computers are useless. They can only give you answers - Pablo Picasso
I think your dislike for Katz has clouded your thinking. I sympathize.
... we really don't need a bad precedent when androids get around deciding what do about human rights.
Of course pain is not irrelevant. When you say that pain is irrelevent but personhood is relevant, you fail to see that pain includes personhood.
Now, I'm not trying to promote some angists philosophy, or, for that matter, arguing that ability to sense equals personhood. What I am saying is that when Katz says "Can Androids Feel Pain" he really means "Are Androids Going To Be Persons Just Like Me And You Really Soon Now." (Note the difference of sense and feel.)
Person isn't very good word for the use you are putting it into. Person is rather synonymous with human and that may well lead to assumptions that aren't correct. Most dog-owners would agree that their dog has a personality, but few would consider the pet to be a person. Still many dog-owners would agree that their dog has rights, meaning that in the eyes of the owner the self of the dog has value that is not dependant of the owner. That's what you mean when you say person, a being with rights and value in itself. It is possible that an android would someday be able to pass a human (which is so hard that many humans fail lot of times) thus qualifying as a person, but as a whole I would prefer "a righted being" (or something similar) as that would potentially include animals.
As to choice and meaning provived by modern technology. Clarke speaks about future not today. In the future leasure state there really is choice in the form of being free of necessity. This is to some extent reality in the wellfare states of modern Nordic countries (Norway, Sweden, Denmark and Finland). I personally know several people who have decided not to work for living but rather live of the social benefits provided by the society. This is natuarally discouraged, because we do not not have AI/UI/AL of the future at our disposal, so if too many people stop working the system collapses. Anyways, that's besides the point, which was that Clarke said future technology, he didn't say modern technology.
To very many people the meaning comes from necessity. Their choices are a) do or b) die. Just being alive and finding substance from within themselves isn't a reality they have to deal with.
Sure numerous men and women have contemplated the meaning before but they have more or less been part of that other 1% and have chosen to think about meaning. This is by the way where I think Clarke goes wrong. He suggests that given the chance people would start to think. I really do think most people will not choose to think but rather will drown themselves in escapist entertainment, which undoubledly will be available in abundance.
Your quote from G.K. Chesterton indicates that you really missed the point so badly that you think that you are disagreeing when you actually are agreeing. I do share Clarke's and Katz's fear for humanity.
Might really be worth it to grant as much rights to animals as possible
--Flam
Computers are useless. They can only give you answers - Pablo Picasso
> Clarke's own "200l: A Space Odyssey" took a more
> malevolent view of computing's ultimate
> intentions than his non-fiction writing.
Anyone who would write this sentence obviously has not read 2001. Guilty concience, yes. Psychotic, yes. Malevolent, nope.
Ignorance killed the cat. Curiosity was framed.
The most interesting bit that you say is
The way humans invent stuff often seems a bit random
Thats the whole point, we are usually so caught up in having organised process that we ignore the randomness, if we take genetics if there were no random mutation then there could be no evolution, only stasis, I am currently exploring adding randomness into systems to allow them to evolve including neural network applications check the following URL for an insight into where this is going, its a work in progress, and if anyone has anything to contribute please just email me.
http://members.xoom.com/insanlygreat/
Any sufficiently advanced man is indistinguishable from God
This increasing complexity is the only common thread I see for the recent increases in random, inexplicable violence in American society.
Yes, people are having problems with complexity, when they want simple answers, currently a freind of mine is conducting a survey for a thesis into violence in schools, I think that we need to TEACH dealing with complexity, just look at your average American School, look at the divisions between the Jocks and the Nerds, the social Insiders and Outsiders, embracing complexity means imbracing difference as a source of change, and thats what is needed to become a tolerant society, that thrives on difference and complexity rather than shunning it out of fear.
Any sufficiently advanced man is indistinguishable from God
Didya see The Matrix ? :-)
Stories in the media always seem to overhype artificial life. The idea is compelling, but the results are not so.
The key to artificial life is genetic algorithms. So far these are good at two things: optimising parameters on given designs, and coming up with (frequently surprising and elegant) non-modular solutions to design problems. However real life is highly modular: I have heard of no complex modular solution to a problem originating from a genetic algorithm. Until this happens, artifical life remains a promising idea that doesn't deliver.
I don't think that it will be the creation of a new species, but rather the next step in the evolution of humans.
Today, much of the civilized world is dependant on technology, and in the future, I believe the inverse will be true. It won't be US and THEM, but instead, it will be an advanced form of 'US'.
People will be so tightly integrated with computers/technology that nobody will be able to tell the difference between the two.
I don' think the issue is whether computers can emulate humans - it's whether a form of "life" can be in existence in a computer based environment. There's no need for this form of being to resemble humankind - probably even a requirement that it not. Let's not get too involved with the SF portrayal of "androids" that act like we expect humans to act. That's really just dead end silliness. Any "new life form" is highly unlikely to follow our evolutionary path even though we may initially attempt to guide it in that direction.
"I believe the children are our future: nasty, brutish and short."
Years ago at a small conference at Carleton College, one of the presenters described an "AI". The program took disparate facts, compared them against an enormous database and developed a relevant solution. Each time the program was presented with the same set of facts, it had to go through the same set of actions. And it never got bored. The ability to be bored by repetitive tasks is probably one of the better harbingers of true AI.
We are years from this, but the idea of mechanical life "taking over" still is fairly ridiculous. In every scenario, from Rosum's Universal Robots to Matrix assumes that these AL/UL/AI will all have similar motivations and desires, which is tough to believe. A medical AI probably would have problems with a construction AI who doesn't get along with the shipping AI. Competing interests don't go away with higher intelligence.
Nearly every technical advance has promised more time, more creativity, more leisure. We have yet to see this promise fulfilled.
In any case, perhaps we should program the Univeral Laws of Robotics into every AI.
"You despise me, don't you?"
"If I gave you any thought, I probably would."
Before reading Kurzweil's "The Age of Spiritual Machines" last week, I would have totally agreed with you based on my undergraduate CS courses in AI. Now I still agree with you mostly but wouldn't make your "50 year" claim.
Kurzweil's argument is essentially that our understanding of neuron and brain function is reasonably far along and increasing geometrically (compounding like Moore's law, although Kurzweil doesn't single out a particular metric or rate.) He posits that our ability to model the physics/informatics of neurons and neurochemistry is sufficient to develop an understanding of "Real Intelligence" that will make "Artificial Intelligence" an implementation detail. You seem to agree we'll do this someday.
Kurzweil thinks we'll have the ability to do destructive brain-scans within 50 years (and hints at practical AI due to neurochemical understanding within 20-30 years.)
I haven't followed up on the research in his footnotes and am not familiar with the progress of neurological research findings and outstanding "hard" problems, but at a surface level, I still saw a lot of hand-waving. Thought provoking, but not convincing.
--LP
Consciousness? Call me old-fasioned, but, err
Aren't they still just running code, just more of it, and in more complex varieties? I think it would just be a simulation of consciousness, not the real thing. Are these scientists victims of new-age thought, or are they saying that human consciousness is just a program running on biological hardware?
At the lowest physical level we are simply a collection of electrical impulses connected in such a way that we can move, think, and act. The only possible difference between ourselves and a cyborg or android is that of a 'soul'. I personally believe in the existence of the Soul, but I do not propose to dictate to God who he may give souls to and who not. If the android has a perfect replica of conciousness, then with that replication comes all of the characteristics inherent in humanity, including the desire to eliminate competition for habitat. Does it matter whether the android that kills you is REALLY conscious or only perfectly simulating consciousness?
Kintanon
Check out JoshJitsu.info for Brazilian Ji
Philosophers have a greater insight into the mind than do computer programers and authors
Philosophers aren't really a class of people anymore, everyone and anyone can be a philosopher, we're all philosophizing here. I wouldn't knock the validity of SciFi authors predictions. Orwell was right more often than not.
I can see some striking paralels between our current society and the society in Asimov's 'Caves of Steel'. Perhaps you should read some of the material you are bashing before you go spouting off about it?
Kintanon
Check out JoshJitsu.info for Brazilian Ji
AI really should mean Artificial Ignorance.
If we had computers with even a SMIDGEN of intelligence we would be making progress.
Not to start a flamewar, but I encourage people to go take some AI courses and see for yourself just how hopelessly clueless we are about building an intelligent computer, let alone an ignorant one within a limited context and strict domain. i.e. Image Recognition, Language Understanding, etc. are NP-complete problems, they aren't going to be solved any time soon.
Cheers
You haven't taken any AI courses, have you?
We don't have a clue on how to build Intelligent computers let alone Artificially Ignorant ones in a very confined and limited domain.
This is a tad silly... well, probably three or four tads silly, actually. I myself pay some attention to AI from an academic standpoint, and I've never heard of anything that comes close to a general-purpose human-replacer. There exist systems that do some things very well- Deep Blue, for example, is an excellent chess player- but Deep Blue can't put together a basic english sentence, control muscles finely enough to stand up, or tell the difference between your face and a '57 Chevy. AI right now is populated entirely by idiot-savant programs, that can do one thing very well (where "very well" is somewhere between "quite a bit worse than a human" and "Superman-like"), and can't do anything else. This is hardly a-life. It's not even a precursor to a-life, really. The techniques that you use to make an idiot-savant probably won't work to make a generally smart thing.
Furthermore, even if somebody were to come up with an integrated human, the idea that it would then take over the world is quite unlikely. Remember- we program the things! We get to do whatever we want with them! If we say "Don't do X," they don't do it! Period! So if you say "Don't take over the world," they won't. They might get to be popular- they may be a more attractive economic choice for some people than hiring high-school dropouts- but I doubt it, and anyway, that's beside the point.
And before you say "but AI could evolve," I will point out that AI can only evolve in the directions that programmers allow it to evolve. And even "evolve" is an analogy- the performance measure is coded by the programmer, the mutation is coded by the programmer, the feedback interpretation is coded by the programmer- the only way an AI will take over the world is if somebody makes a "take over the world" AI. And try getting grant money for that! =)
Oh, and if you've read 3001, you ought to know better than to trust the sagacity of Mr. Clarke.
-jacob
Err... gee... last I checked, we hadn't the foggiest idea of what we are "at the lowest physical level." True, there are electrical impulses, but why do you say that's the lowest level? Do you know?
-jacob
heh, as of today, I'm setting my preferences not to include katz articles.
I usually run Katz through my crap-o-meter. He almost always gets two flushes.
screech
jason
Have a good day?! Impossible! I'm at work!
I completely agree. "Soul", in my definition, is "human arrogance".
Let's take that a step further... It's conceited to think that humans are the end-all be-all of evolution... That the universe was designed for their presence.
If anything, I think humanity is a stepping stone for something better... And something without a soul, I'm sure. : )
Are you for or against research in artificial intelligence and/or A-Life? A) Creating life is the ultimate human goal. B) Yeah, I think it's okay to create artificial life. C) Yeah, I think it's okay to create artificial life, but we need to have total control over it. D) No, I don't think humanity should be messing with things like creating artificial intelligence. E) You're going to hell for even asking this quesiton! F) Hemos sucks. ...Personally, I'm a believer in (A). Articles like this *really* irk me.
Well, If your going to talk about defending ourselves against a-life and Artificial intelligence, this is what I would do. Why don't we build some sort of backdoor shutdown switch into the ai's we build? That way, if the AI grew out of control, it would be easy to just post a command to stop it. However, this would be complex, because you would also have to build in a device that prohibits the organism from detecting this vulnerability. Otherwise, it could just change it's programming to delete the backdoor. Also, the backdoor command would have to be secret to keep crackers from shutting down the AI.
Another possible defense mechanism might be a "rules system" similar to Isaac Asimov's. If you built a system of rules into an AI that prohibited it from doing certain things, it might stop a takeover. However, this would not stop a stagnation of the human race after AI's began to take up a majority of the jobs.
Even if the AI's did succede in becoming extremely intelligent and powerful though, It would be hard for one to take over the world. An AI would require humans to do manual labor at least, unless robots became widespread. Of course, for detailed and active roles, a robot or AI need a signicant infrastructure to operate (Power, proccesors, temperature control.) Humans would still be needed for frontier operations (assuming, of course, that the AI's chose to explore outer space and the solar system.
I'm wondering what threats, exactly, might be posed to humanity by a new breed of artifical intelligences. Off the top of my head, I can see two major possibilities:
One is that we build self-replicating, locomotive AI's, they get together and decide that we (humans) suck, and they claim the planet as theirs, a la The Matrix. That's all well and good (though probably unlikely, why don't they find a new, unsullied planet and mold it in their own image?), but I don't think I quite see why college level literacy will save you from that fate.
The second type of threat I see AI's posing to humanity is of a more psychological nature. An AI decides it wants something but lacks the means to do it own its own, and subtly influences a susceptible human to perform that task, a la Wintermute in Neuromancer. I would expect this to be the more probable danger and I can also see how increased education might provide a measure of prevention. However, if an AI had the ability to somehow or another influence one's entire worldview, I'm not sure any measure of education is going to save him.
I personally think artificial intelligence is an incredibly exciting field, and one that is probably inevitable. If we could theoretically build a human brain molecule by molecule, there's no reason that it wouldn't be as concious as any of us; why can't we do the same thing with silicon? I also think that if it ever does get off the ground, artificial intelligence will be a discipline that requires one part philosophy and one part computer engineering.
It should be an exciting time, to say the least. I'm just not sure if I want it to hurry up and get here.
Sigh... Something went wrong, and necessitated this repost. Sorry.
I'm wondering what threats, exactly, might be posed to humanity by a new breed of artifical intelligences. Off the top of my head, I can see two major possibilities:
One is that we build self-replicating, locomotive AI's, they get together and decide that we (humans) suck, and they claim the planet as theirs, a la The Matrix. That's all well and good (though probably unlikely, why don't they find a new, unsullied planet and mold it in their own image?), but I don't think I quite see why college level literacy will save you from that fate.
The second type of threat I see AI's posing to humanity is of a more psychological nature. An AI decides it wants something but lacks the means to do it own its own, and subtly influences a susceptible human to perform that task, a la Wintermute in Neuromancer. I would expect this to be the more probable danger and I can also see how increased education might provide a measure of prevention. However, if an AI had the ability to somehow or another influence one's entire worldview, I'm not sure any measure of education is going to save him.
I personally think artificial intelligence is an incredibly exciting field, and one that is probably inevitable. If we could theoretically build a human brain molecule by molecule, there's no reason that it wouldn't be as concious as any of us; why can't we do the same thing with silicon? I also think that if it ever does get off the ground, artificial intelligence will be a discipline that requires one part philosophy and one part computer engineering.
It should be an exciting time, to say the least. I'm just not sure if I want it to hurry up and get here.
The theme of the contest is not betrayal in a man vs. machine story, but rather just betrayal. You can leave out the man vs. machine angle. You see, you'll be competing against a machine that will be writing a story. Or, is this your attempt to throw the readers off track? Write a betrayal story from a machines perspective so everybody thinks this is the story written by the machine.
Intriguing angle, but really, everyone will instantly recognize your style. It's so mechanical.
You raise several points:
I agree that it is easiest to presume sentience in entities like ourselves. We all use this as a heuristic in everyday life: you don't subject people you meet to the Turing test before deciding they are sentient. Instead we assume because an entity looks human it has an inner life similar to our own. I also agree that it is more difficult to decide on the sentience of entities explicitly created as mimics. The experiences with Elisa, in which people ascribed complex mental characteristics to a very simple program, showed that naive individuals can be fooled.
However, just because it is harder to make a decision about the sentience of entities that are more different from you doesn't mean it is impossible. Otherwise, you would seem to be taking the position that humans (and perhaps a few other mammals) are the only entities in the entire universe that can be sentient, just because they happen to closely resemble you.
The example of wetness of a simulated ocean is a bit of a red herring. I agree that no matter how well modelled, a simulated ocean can never be wet to an operator outside the machine, nor could it ever drown her or taste salty. But the quality of thinking is different from the quality of wetness. Consider a legal brief or a computer program. For these objects it is the pattern, not the physical medium that the pattern is embeded in that is important. It does not make any sense to talk about a 'simulated' legal brief or a 'simulated' computer program - each object with the specified pattern is in some sense as good as the original, although in practice one physical embodiement may be more practical to use in a specific situation.
In the same way, the important thing about sentience is its pattern, not its physical embodiment. Even neurobiologists have only the vaugest grasp of the physical embodiement underlying our sentience, but that doesn't stop us from thinking and experiencing.
I think that the Turing test remains an excellent minimum test for sentience. I can imaging sentient entities that would not pass the Turing test, but if the test is long enough and with a good enough examiner, I don't see any way that a non-sentient entity could pass it. I would argue that the fact that no computer program has come even close to passing a Turning test is a sign of the test's strength.
It seems highly implausible to me that we will be able to create sentient, human-level intelligence any time before the hardware power of computers at least matches the hardware power of the human brain (estimated at about 1000 million MIPS, see http://www.transhumanist.com/volum e1/moravec.htm ). If Moore's law continues to hold, this amount of computer power will cost about $1000 in the 2020-2030 time frame (see figure 2 in the cited article).
You argue that because we don't know the exact basis of consciousness, we have no ability to decide whether an entity is conscious or not, and thus can't know whether it is deserving of rights or not. While an interesting philosophical position, that stance is simply not practical for living in the uncertain, messy world that we all inhabit. People make decisions on the basis of uncertain information and imperfect understandings all the time. Some of those decisions have really important consequences, like whether people get the right diagnosis and treatment and live, or die when they have an illness (doctors); whether people get deprived of many of their rights and go to jail (judges, juries); and even what legal rights you have in the first place (legislators, the electorate at large).
It would be nice if we had ways of removing that uncertainty from life, but we don't. So I see no reason why a similar level of uncertainty presents any barrier to the practical decision about whether an AI is conscious or not.
You leave out the point that you can only verify the subjective experience of one person - yourself. How do you know that other people have subjective experiences or are sentient? Only by observing their behavior (includes communicative behavior).
That is the same way we will decide whether AIs have a subjective experience (leaving out the possiblity of uploading - in which case you will have direct evidence for the subjective experience of a single 'AI' - yourself): by observing their objective behavior and talking with them. We will have no direct way of verifying their subjective experience, but that does not seem to be a handicap (most of the time) in dealing with other people, where again we have no direct way of verifying their subjective experience.
Another way to see this point is to paraphrase the first part of your last sentence: "It's not hard for me to imagine a world in which a bunch of people act as if they had some sort sort of a culture and a sentient life, but in fact are just a bunch of overcomplicated tree shrews..."
Computer's tasked to filter Reality _are_, indeed, in control of the human at the other end of the technology. Air Traffic Controllers, fly-by-wire F-16 pilots, ABS equipped drivers and Internet posters all interact via computer. Computer-augmented realities already affect human behavior.
To the extent technologies *extend* man's abilities, people _will_ choose a higher lifeform, even if AI UA or just augmented. There's no ethical/moral boundary for computer technology. Expect anything that can, will happen in the Future.
For one thing I have my doubts that even the most foolish inventor would construct a computer or AI that would lead to the destruction of the human race. While Asimov's 3 laws are deceptively complex (and probably prohibitively so), a person would be a fool to create a life form that destroys them.
Secondly, in spite of all the frankenstein press that technology often gets, I sincerely doubt that machines (assuming they do become self aware) would have anything worse than a symbiotic relationship with machines. What reason would a machine have for trying to eliminate organic life given that organic life created it? Logically it follows that if organic life can create a machine that is self aware, it could also create other useful things.
As far as the meshing of living with machine, I personally can't wait to get some upgrades. Embrace technology, don't fear it.
Unbreakable toys can be used to break other toys.
Have we ever built a blender we couldn't turn off, or a car, or a computer? My point is, humans would not build tools without safety measures. If someone did create true AI (which I believe is inevitable), they would not be so careless as to design it in such a way that it would want to take over.
:-)
"Ok," you say, "but what about when computers build computers that build computers..." I don't believe it would be that complicated to tell the first computer, "Build me a smarter computer, just make sure it doesn't hate me." You make sure those instructions keep getting passed down, you're safe.
People are always pointing to stories like 2001 and Frankenstein and calling them lessons on what happens when humans "play God." Sorry, folks, it's simpler than that: It makes the story interesting. How exciting would a story be about a robot that was created and worked exactly as it was supposed to and didn't cause any problems?
Billy
human://billy.j.mabray/
"Every good system has a backup." -- Dale Hanchey
If a-life started reproducing on the net, it's possible to stop it by shutting down every power plant on earth, or if the a-life takes over the power plants, try to bar access to the energy sources, ie coal mines, oil fields, uranium mines, windmills, solar cells, while the existing supply runs down.
The reason people fear computer domination is because, if they organized themselves in a Borg-like structure, the computers would never argue amongst themselves, never forget anything, and evolve much quicker than natural selection allows. It's also possible that genetically engineered organisms could succeed the human race. However, if the history of carbon-based life is any indication, these organisms would be no more efficient than humans, just smarter overall, just as humans didn't overcome the apes' problems of dissent and ignorance, who in turn didn't overcome the problems of their ancestors, early mammals. At the same time, however, it's since all AIs so far work on a strictly objective basis, I doubt it would be able to think subjectively, ie. create art. I have no doubt in my mind that any carbon-based superior to humans would have artistic talent (better than we have), but to do so it would have to be flawed, just as we are. So do you want to be succeeded by a more elaborate, yet just as flawed, version of yourself, or a Borg-style electronic- self-altering-critical-thinking-algorithm that is impeccably efficient but knows no goal but to dominate and assimilate?
True enough, I don't have any objective way of knowing that anyone or anything is sentient other than myself. On the one hand, though, I am safest presuming the sentience of entities that are most like myself. On the other, I am least confident presuming the sentience of entities created for the explicit purpose of mimicking sentient behavior. Your example of tree shrews shows the typical confusion that is rampant in AI circles. I have no reason to believe that tree shrews do not have an experience. I believe that it is the general case that living mammals do have an experience. (I am very curious as to how far down the evolutionary ladder this having an experience of the world goes.) Experience is not the same as language or reason. Most everyone agrees that housecats feel pain or pleasure, while Pentium processors running MS software do not, which in their case is a blessing, to be sure... What AI successes to date seem to tell us is that language and reason can exist independent of sentience (by which I mean *having an experience of being in the world*), while lower mammals seem to tell us that sentience can exist independent of reason. As I read in a .sig a long time ago (alas, attribution lost) "It's typical of people who think for a living to confuse thinking and living". But, because there is no objective measure of sentience, these are just hints. When I build a simulation of the ocean, I do not think that it is wet. I could build a very fine CG rendering of ocean waves that would fool you, but they still couldn't dissolve real salt. The very scary fallacy here is that you are claiming that because you have no way of knowing whether the ocean is real since you are constrained not to touch it, you *must* behave as if it were real. The concern you might raise is that you have many reasons to believe that the Atlantic Ocean exists independent of your belief in its wetness while you know that my ocean was constructed for the explicit purpose of convincing you of its wetness. Now, let's further suppose that wetness is the most fundamentally interesting property of the universe. Are you really willing to trifle with it on the basis of expensive parlor tricks? The trifling little ego trip of passing the Turing Test is likely to be presented to some fool judge somewhere as an instance of a conscious life, born in some particular country and hence deserving of citizenship. All the "AI experts" will line up to testify on its behalf, while everyone with the sense to spend their time doing something productive will not be acknowledged as competent to testify. I think it won't be long before this legally empowered monstrosity gets out of control. It could use its reporoductive rights to multithread itself to outvote the rest of us in no time. How many cycles per second, or per century, would an instance of the code need to run to be considered sentient, anyway? Just because it can play chess doesn't mean it can enjoy itself, and a universe full of chess-playing bogus-feeling automata is not an implausible outcome of the urge to build an AI. This isn't a heroic quest, it's stupid and crazy beyond comprehension. We don't know if consciousness is logical, chemical, quantum, or even in some sense supernatural. We can't know. We can't test the hypothesis. We can't know if a simulation of conscsiousness is conscious or not. Therefore we can't know whether it is deserving of rights. Therefore it is unethical to build it. I hope the world sits up and takes notice soon. Alas, it sounds too silly for most people to take seriously. Unfortunately, there are some very smart people who adamantly insist on missing the point. We dare not mess with replacing consciousness life because we don't know and can never know what consciousness is.
mt
oh foo, sorry about the formatting.
True enough, I don't have any objective way of knowing that anyone or anything is sentient other than myself.
On the one hand, though, I am safest presuming the sentience of entities that are most like myself. On the other, I am least confident presuming the sentience of entities created for the explicit purpose of mimicking sentient behavior.
Your example of tree shrews shows the typical confusion that is rampant in AI circles. I have no reason to believe that tree shrews do not have an experience. I believe that it is the general case that living mammals do have an experience. (I am very curious as to how far down the evolutionary ladder this having an experience of the world goes.)
Experience is not the same as language or reason. Most everyone agrees that housecats feel pain or pleasure, while Pentium processors running MS software do not, which in their case is a blessing, to be sure...
What AI successes to date seem to tell us is that language and reason can exist independent of sentience (by which I mean *having an experience of being in the world*), while lower mammals seem to tell us that sentience can exist independent of reason.
As I read in a .sig a long time ago (alas, attribution lost) "It's typical of people who think for a living to confuse thinking and living". But, because there is no objective measure of sentience, these are just hints.
When I build a simulation of the ocean, I do not think that it is wet. I could build a very fine CG rendering of ocean waves that would fool you, but they still couldn't dissolve real salt.
The very scary fallacy here is that you are claiming that because you have no way of knowing whether the ocean is real since you are constrained not to touch it, you *must* behave as if it were real.
The concern you might raise is that you have many reasons to believe that the Atlantic Ocean exists independent of your belief in its wetness while you know that my ocean was constructed for the explicit purpose of convincing you of its wetness.
Now, let's further suppose that wetness is the most fundamentally interesting property of the universe. Are you really willing to trifle with it on the basis of expensive parlor tricks?
The trifling little ego trip of passing the Turing Test is likely to be presented to some fool judge somewhere as an instance of a conscious life, born in some particular country and hence deserving of citizenship. All the "AI experts" will line up to testify on its behalf, while everyone with the sense to spend their time doing something productive will not be acknowledged as competent to testify.
I think it won't be long before this legally empowered monstrosity gets out of control. It could use its reporoductive rights to multithread itself to outvote the rest of us in no time. How many cycles per second, or per century, would an instance of the code need to run to be considered sentient, anyway?
Just because it can play chess doesn't mean it can enjoy itself, and a universe full of chess-playing bogus-feeling automata is not an implausible outcome of the urge to build an AI. This isn't a heroic quest, it's stupid and crazy beyond comprehension.
We don't know if consciousness is logical, chemical, quantum, or even in some sense supernatural. We can't know. We can't test the hypothesis. We can't know if a simulation of conscsiousness is conscious or not. Therefore we can't know whether it is deserving of rights. Therefore it is unethical to build it.
I hope the world sits up and takes notice soon. Alas, it sounds too silly for most people to take seriously. Unfortunately, there are some very smart people who adamantly insist on missing the point.
We dare not mess with replacing consciousness life because we don't know and can never know what consciousness is.
mt
We can never know and will never be certain whether we have created sentience.
Objectively decideable hypothesis require objectively observable phenomena. The presence or absence of a subjective experience is, obviously, a subjective rather than an objective phenomenon. Experience cannot be measured or objectively detected. Therefore its presence or absence is a scientifically undecideable question.
The ideas that humans will sooner or later, and possibly sooner, be replaced by automata is therefore enormously dangerous. By building simulations of subjective experience without verifiably creating subjective experience, we run the risk of replacing life, in the sense that it matters, with non-life.
We should be honest and admit that we have no way of knowing whether AI in fact can have an experience. We don't understand what experience is in the context of objective science, and we never will because experience is only detectable subjectively. This is why the so-called sciences of experience (psychology and AI) get so muddled at the most important point. They are reluctant to acknowledge that what they study and model is not actually objectively observable.
Therefore, by making the effort to build our successor species we have no way of knowing whether we are leaving out the most important ingredient. I agree that this issue is of enormous importance. I view it as a threat on a par with total war or an asteroid strike. It's not hard for me to imagine a world in which a bunch of machines act as if they had some sort of a culture and a sentient life, but in fact are just a bunch of overcomplicated toasters that have pointlessly and stupidly displaced actual living beings who can actually enjoy the toast.
mt
I'm tired of debating about whether this soul (which there is absolutely NO evidence for) exists and if it's necessary. I just want to find out by experimentation. if there is no evidence for a soul then how do you account for the presence of human reason? elephants have bigger brains than we do (I think) other animals have comparable brains to human beings yet we are the only animal that possesses reason. ( with the possible exception of chimpanzees and dolphins, i dont think they do but it is possible) thus one cant explain human intelligence with the purely physical, thus it follows that human reason must come from the metaphysical.therefore a gigantic neural net will not answer any of these questions, nor will any other physical experiment. therefore we cannot create a computer that possesses reason in the human fashion. as computers are physical objects.
Pound for pound, the amoeba is the most vicious animal on earth.
We are the new species. The earth will not overgrow with vegetation, because we, the new species, eat paper for a living. We burn fossils for sustinence and we belch smoke. We will for a long time, and then things will change somehow.
I am reminded of the utilitarian (in the philosophical sense) debate over population control - balancing the needs of those who already exist against the potential existences inherent in our biology. We cannot lean too heavily towards possibilities and ignore actualities.
We live in a physical universe. Much as it would suit many mindsets (sometimes including my own) to have conciousness wholly configured in an artifically or otherwise constructed mental space of pure data-flow, that's not the way it is.
I'm not pretending to give answers here, but I feel the triumphant declaration of independence in the previous post was a little premature.
As (still) homo sapiens do we have no responsibility towards those of our own species who live hand-to-mouth, those who's children are dying of starvation while we create new minds in our machines? Do we have no responsibility towards the delicate balances within the biosphere and geosphere which led to our own emergence from a 'lower' species? Even when or if we have no need of them (which is hardly the case now) do we not retain those responsibilities?
Surely no matter what we create, or what we become, it is only half a creation, half a becoming, to ignore the context in which we create, in which we become, from which we emerge, in which we still exist? Surely the declaration of independence is coming too soon, when we do not have what we need, and we haven't given back what we owe?
Of course androids will feel pain! Pain is a basic thing that most living things have... That indicates that pain is a good design feature, despite the unpleasentness.
The real question is will they dream of electric sheep?
Be insightful. If you can't be insightful, be informative.
If you can't be informative, use my name
Just an easy example of this. :) Think about the military applications of androids! Zowee! We'd never have to lose another American Boy in Battle!
Wait for war. Apply spin to the topic. It'll happen. I dunno all the things you can count on, but two of 'em I'm positive about are war and new war gadgets.
Oh, I would say a lot of different animals have the ability to reason to some extent. Take my dog for instance. If my dog receives a treat every time she does something acceptable or good, she will continue to do things that are acceptable or good (from her perspective). I would definitely call that reason. You could make a case for animals having poor logic because of lesser intelligence, but that doesn't mean they don't have any logic. Heck, I know TONS of people who have poor logic and reasoning; I'm guilty of it sometimes as well. Also, no argument from a metaphysical viewpoint can be considered constructive or relevant, by the very nature of metaphysics: something that can not be physically observed.
Thomas Reasoner
A thing is justified by the "thing in itself".
Of course, anyone can do whatever they please to do; and be whatever they please to be. Is Einstein only a patent clerk? What "label" would you attach to Da Vinci? You don't need to have a title or get payed to have your work or thoughts affect humanity. Linux is a perfect example.
So yah, I would consider myself a philosopher, as long as I was contributing to the field of philosophy in any capacity.
At any rate, philosophy is the most fascinating field there is, it's also the most irrelevant.
Thomas Reasoner
A thing is justified by the "thing in itself".
If quantum theory is correct (and there is no reason for us to think it's not), then it is physically impossible to have infinites (because everything is quantified). There are many scientific fields that support this conclusion. Physics is a prime example: conservation of energy, mass, momentum, etc. There must be a finite amount of energy/mass/momentum in the universe, otherwise you could have a perpetual motion machine.
I infer that space and time operate in a similar manner, since space and time are defined with respect to the 'stuff' in the universe. Example: one cubic-foot would 'look' the same as one cubic-mile if there was nothing to reference it against. Since the reference, matter and energy, is quantized, the thing being referred must also be quantized.
Also, from physics and mathematics: every action has a unique reaction, and only one unique reaction for every action. From this I conclude that the past and even the future are finite, and indeed have only ONE possible outcome. There is only one past, hence there can be only one future. Individuals still have free will, but the CHOICES they make have already been determined from previous experiences.
Thomas Reasoner
A thing is justified by the "thing in itself".
Would the AI grow to hate that red light and actually feel hurt from it like a human, or would they not care. Some humans can train themselves to not feel pain, would we be able to train robots to be hurt?
This was touched upon in Asimovs 'Caves of Steel' where Elijah asks R. Daneel if he can laugh or feel good about himself. Also what his sense of justice was.
Does anyone know of other books that touch on this topic?
Wiggeda Wiggeda Wack - Kriss Kross
Yes, computers will design other computers, as it has already been proven, but it will be much more than seventy years before they are able to genetically generate something equal to the human brain. Consider this, if a genetic evolution were to be simulated for the human brain: approximately 1 billion humans participated in the evolution of humanity as we know it (this is a very conservative estimate of 400 generations with an average of 2.5 million people participating). And humans have generally 15 billion neurons. By another conservative estimate, they think along the lines of 1/100 of a second; they don't use their entire brains during this time, but simulation software couldn't figure this out (and neither can we). Many neurons (or so I am told) connect in no fewer than 10000 places; thus they must be represented by at least 14 bits. If this is all multiplied out, we must sample 15 million trillion neurons every 1/100 of a second (each connecting at least 100 others, since not all connect to very many others), or 0.15 trillion trillion member accesses per second to conquer the problem in the aforementioned 70 years, if we are able to split up the hierarchy of evolution to parallelze all 1 billion lives into life spans of 70 years, and considering that any action of a human may influence its life, death, and ability to create offspring. To hold all this, we might need 210 billion trillion bits, or 26.25 billion trillion bytes. And that is merely the very last piece of evolution, a second compared to the billion years that got us here.
That shows some very fuzzy reckoning of how hard such a simulation problem would be to do with todays technology. Even with quantum states, the numbers are on the order of a mole (!)
Somewhat more irrelevant: I feel that even tough the problem is daunting, we need to try it; I think that humans are de-evolving as a result of modern medicine; genetic mutations are piling upon out human genotype; so what if machines replace us? We can make them better than us.
> Nothing scary
Nothing scary in having immensly powerful machines, able to modify their own systems, at a rate we cant even approach ?
I can imagine that in the time it takes the operator to think "thats bad", and reach for the switch, the AI will already have found out about networks, read some hackers guides, and escaped. Because intelligne tlife strives to break its barriers.
Just because we have the tech to do something doesn't mean we should do it. If there are significant concerns that we can't deal with this as a society, then perhaps the most ethical thing to do is not do it. Doing something simply for its own sake is not always a noble endeavor.
Keeping in mind that this is a Katz post I don't think I'll get too alarmed. I don't see machines making humans obsolete.
I'm guessing most people in the world still haven't even seen a computer.
CT
Constitutionally Correct
We're already two species. There's the homo informaticus to which all reading this belong, and the old homo sapiens that isn't at all sapient to how we are changing it's world.
Sorry, you fail the biology exam. Behavior and knowledge (which is the only difference you describe between your "species") are part of culture, NOT biology. That means that an individual from the tech-ignorant group could, with sufficient education, become a member of the tech-enabled group. No amount of education will turn Washoe into a human, by comparison.
The old species is already nomadic, living hand-to-mouth and at odds with nature.
Got news for you: we tech-enabled bunch are even more at odds with nature. Or did you think that our consumption of tantalum capacitors for cell phones and petroleum to make PalmPilot cases is somehow sustainable? And nomadic? Tuaregs and Australian aborigines are nomadic. Homeless people in the US are sometimes nomadic, but more often they stick to a given territory. The rest of us (some 230 million plus) are definitely not, but I'm sure you don't think all those 230MM are part of homo informaticus. Don't let your enthusiasm for your point inflate your rhetoric beyond all reason.
The new species has been able to avoid the dismal lifestyle of the old through it's fusion with technology.
Maybe YOU have an interface implant behind your ear, but nobody I know does (and I'm sure some of my coworkers would get them if they could). We have adopted technology into our lives, but fusion is considerably more advanced than that: it implies creating a link that can never be dissolved. Pacemakers qualify: PDAs don't.
The fact that we have embraced technology, and evolved thereby, was a willful, convenience driven event. We are dependent on our technology as much as birds are dependent on their ability to fly. To un-plug means death. We may not be left biologically dead without our tech, but our lifestyle, our standard of living, would end. Is that no death?
Ask this question of someone with a gun to their head, or with a terminal disease slowly eroding their body. Or of a bird that relies on its wings to escape from all the other bird-eating life forms out there. If adoption of technology was a convenience-driven event and not a survival-driven one, then by your own words loss of those "conveniences" does not threaten survival. And I doubt that even you would say "Give me my lifestyle or give me death!" Most of us value our lives over our lifestyles.
okay, so before I start, I will admit that no one thinks that Arthur C Clarke is a bigger crackpot then me. I remember watching his stupid TV show along time ago. Everything that happened "challenged man" or "eradicated a concept of nature." The fact of the matter is, Arthur C Clarke is a wannabe Isaac Asimov / Carl Sagan figure. But he does have a decent point. I think what he is getting at is...if a machine one day can think "I", and we can download that thought on to say an Iomega ZIP disk (the official removable media of the robot existentialist), and we look at the thought on a computer and play with it, or buy it at Software etc, what exactly does that say about the quality/purity of our own thoughts? i have no idea, myself. but when this happens, I hope to god (even though this suggests there probably isn't one) that humanity doesn't drone on like a bunch of consumers. I think that what clark was getting at with the whole education thing, is that in order to momentarily push back the veil of politics, religion, consumerism, and examine the situation as it really is we'll all need to be pretty damn smart. If the opposite is the case, we'll have every ozark preacher, politician, and "family value" housewife yelling out in protest. Of course, we'll all know that they're all just being stupid, and not thinking for themselves, but a bunch of nerdy folk like us sitting at a computer screen, can't do much to stop an angry torch weilding mob from smashing the first AI to bits. peace out.
Pseudo-science and Science Fiction are wonderful things to play with, but Clarke and others tend to take themselves too serious at times. Perhaps it goes with believing too much in their own rhetoric. The future of Man lies in overpopulation and the depletion of natural resources, followed by disease, war, and eventual barbarism. Perhaps the man of the future will sit in a cave with a battery powered lap top and search the remaining internet for some sign of intelligent life. Perhaps he will never find it.
Androids feeling pain is with us now, depending upon just how much merging of man and machine you want to consider. People walk around with all kinds of mechanical devices implanted in them today. Now if we were to start implanting human parts into machines, a real cross over could be accomplished. But any progressive thinking doctor would tell you that the elimination of pain is one of the primary goals in medical science, so why would future science create an android which would feel pain? Emotional pain perhaps, if such a thing could be made possible, but never physical pain.
http://www.xensei.com/users/jong/nedy.html
"Fere libenter homines id quod volunt credunt." (Men readily believe what they want to believe) C. Julius Caesar, Commentarii de bello Gallico III, 18
Only had a couple of Philosophy classes in college, but not sure how many of my professors, and more to the point, how many of the philosophers we studied would endorse very many of these points. Think we better decide how it is and whether it is that we even have "intelligence" or consciousness before we can even consider "creating" something else that does. And by creation I mean something beyond "Pro-creation", which at its best is duplication (but not even really.......)
Funny and I thought Perl == Paid employment recently located
Do androids feel pain? Well, yes. That is, if you drain all their oil and drive them up to the supermarket.
--
I've recently begun work on my own "infobot" for IRC ( mainly to implement original ideas from the start instead of intruding on work of existing bots ), and though the thought of having an intelligent conversationist bot that can emulate a human is appealing in some ways, there are certain problems with taking that approach yet in almost _every_ level of AI work.
:).
;)
First, Artificial Intelligence's time has not yet come. We've got to be able to make our robots and computers operate as tools to a much better degree than we do today. If we ever hope to succeed in the loftier goals of AI specified in Katz's column, let's first try for the basic goals first. Let's make digital machines or perfect servants first, able to understand our behavior and speech and interact with us before "blessing" them with independant thought. At the present day, the average population still is running into a gigantic learning curve in using Windows98. Let's eliminate this barrier first
True AI, though it is a great academic endeavour, it is also a very futile endeavour. What can we really hope to accomplish with a free thinking robot? Heck, look how most of the free-thinking humans are turning out
Personally, I think the best use of AI is to give it a limited role in the ways certain digital machines interact with us. I do not believe that it is a machines job to become schoolteachers. I do believe, however, that a house-wide information and service computer that can understand speech and interact with humans is going to be a VERY useful tool in the future. However, we have yet to see how much AI is going to play a part in that.
Dont get me wrong. I think someday the AI labs of the world will create a respectable artificial intelligence that will probably be capable of some cool things.
What I dont get is why periodically someone predicts that computers will take over the world and we will be their minions. Has anyone bothered to look at these people and notice that the closer we get to the eluisive goal of AI, the more work we need to do?
Society is already at a point where lack of computer skills puts people at an extreme disadvantage in the job market. Why does everyone expect it to evolve more? Haven't they been paying attention to the fact the computers are being made to become EASIER to use? I imagine when the plow was invented there were just as many people standing around saying it would put them out of jobs as when robots started building cars in Detroit. The tools get more complex, but the conflict will perpetually remain the same.
And remember who is building these things, people. We are building them as tools. They exist to make our lives easier. Computers already trade our secrets and data without our knowledge. Computers already teach our children. They already control what we watch, what we see. Guess what? somewhere down there was still a meeting in a board room that decided what that computer would be built for and what it would provide. Humankind is still in control. Computers dont change their mind. They do what they are programmed for. It's very simple.
Although complex AI will probably someday exist, it's foolish to think that it will somehow be able to outsmart us. We created it you know, so by definition, we are the more intelligent. If survival goes to the fittest, it's absurd to think that tools can become more fit than their creators, it just doesn't make sense.
-Rich
Think of the process of programming for a completely new computer. You first have to code everything in machine code, because there's not even an assembler available [pretend cross-compilers and cross-assemblers don't exist in this context]. So what's one of the first things you write? A crappy assembler, which you use to write a better assembler, which you use to write better assemblers. Once the assembler is good enough, you write compilers for higher level languages, and use these compilers to write better compilers, and so on.
The situation here could be similar. Even though WE may not have the intelligence to come up with an untra-intelligent AI on our own, who's to say we won't use 'dumb' AI to help us design better AI (since it can sit there and do number-crunching 365 days a year, no bathroom breaks), and use that to generate even smarter AI, and then smarter AI... The difference with AI is that at some point the AI can improve itself without our help, as long as the power keeps flowing...
--------------------------
Paranoid thought for the day: What if we really are the result of some Gods' AI experiments, and they're watching to see if we're going to destroy ourselves or not? The physical "laws" of the universe are just restrictions and parameters of the the software that's being used to run us.
-----
--
perl -e'$_=shift;die eval' '"$^X $0\047\$_=shift;die eval\047 \047$_\047"' at -e line 1.
Mr. Katz brings in Asimov at the end without saying much about Asimov's contention that machines won't try to take over the world because they'll be designed not to. Not just incapable of doing it, but (much more important) incapable of wanting to. A well-designed machine is the only object that it's 100% okay to enslave, because what it wants most from life is to serve you. At most I can see the situation evolving to something like the Lije/Daneel thing: the machine evolves from slave to partner, but it still "wants" to serve.
These predictions make more sense when you consider the background against which Clarke made them. In the 60s, the US space program was in full swing, we were shooting for the moon by 1970, and the sky was no longer the limit. If we had kept developing space technology at the rate we did then, we might be laughing now at the idea that it would actually take until 2001 to reach Jupiter...
--
Two ways to make an "intelligent" machine, whatever that means :
:p
1) Rebuild (or simulate) every single neuron and make it fit into a human-like structure which we can't even begin to make conjectures about. This goal may be slightly beyond a 10 years term...
2) Don't try to follow the human architecture, or forget the brain paradigm altogether. The former is Hugo De Garis' method, the second is good old symbolic AI (Minsky, Schank et al.).
The only way to deem those things "intelligent" is to compare them with ourselves. Turing test, stuff like that. The concept of intelligence is totally human-centered. Therefore, to call such a machine "more intelligent than human" is plainly impossible. Either it is roughly as intelligent as man, or it is considered as "something else", but certainly not intelligent.
How could you define intelligence otherwise ? Ability to solve problems ? Come on, your gnuchess program can do this, would you really call gnuchess intelligent ? Ability to overcome complex problems ? Matter of time before a brute force genetic program generation can solve incredibly complex problems (at least when proper modelling is possible), and sorry, I will not admit a genetic algorithm is intelligent. Ability to overcome men's attempts at destroying it ? In this case the Black Plague bacillum is probably the most intelligent lifeform that has ever existed on this planet.
I like Arthur C. Clarke when he writes fiction. He should try to do only that - exactly like Hugo de Garis does.
Thomas PS : Don't come and tell me De Garis is serious about what he writes, I won't believe you
...you actually have to have a science of AI. The current state of AI is that it's 99.9% philisophical, and 0.1% science. I don't want to call it a sham, because the people are honestly trying, but the past 40 years of research has been an utter failure.
That said, I think someday we'll solve the riddle of intelligence and get intelligent machines, but it ain't gonna happen in the next 50 years, and probably longer.
It's not just a question of faster computers. We need 1) provable and demonstrable theories of intelligence, and 2) hardware to implement them. Both of these are not even on the horizon, much less in our generation.
Sorry to dump the cold water of reality! :)
Huh... I always run Katz stories through my X-Files meter-- Big Brother, spying, etc. being the keywords. I also included a few words from the actual X-Files series...
Jesus. He had to put "Kurzweil", "AI", "Clarke", "ethical", "romantic", and a bunch of other words in the same story. My little perl script gave it the highest rating so far-- 5 out of 5!
Is Katz really Chris Carter in disguise?
I think it is extremely important how we will perceive this new form of life, considering that everything I have mentioed above will certainly transpire in the next 70 years.
Vernor Vinge's Technological Singularity should be required reading on this topic. I'll summarize it, for those unwilling to hit the link:
Within 30 years, we will have created computers that, intelligent or not, can solve problems faster than mankind can. The computers will, among other things, be able to build better computers faster than men can. An automated economy will emerge, with the artificial quasintelligences directing progress almost completely.
The change in how homo sapiens sapiens moves forward technologically will be approximately as drastic as that of homo sapiens neandertalis discovering fire. No Neandertal could have predicted what the world would have been like post-fire.
Will we live to see homo sapiens++?
I personally believe Vinge is correct. It'll be a hell of a ride.
The point of no return has already passed us by. We have, a long time ago, created a world so full of complexity, that we have become reliant on our technology. We need the air traffic control systems, the banking networks, the databases, the ISP's, the chips in our cars. We rely on the IC controllers that run our assembly lines and decide how to make our clothes, cook our food, route our electricity.
It is irrelevant to wonder if the machines will ever become sentient, or what effect that will have on us as a species. It's a moot point. We're already two species. There's the homo informaticus to which all reading this belong, and the old homo sapiens that isn't at all sapient to how we are changing it's world.
The old species is already nomadic, living hand-to-mouth and at odds with nature. The new species has been able to avoid the dismal lifestyle of the old through it's fusion with technology. The fact that we have embraced technology, and evolved thereby, was a willful, convenience driven event.
We are dependent on our technology as much as birds are dependent on their ability to fly. To un-plug means death. We may not be left biologically dead without our tech, but our lifestyle, our standard of living, would end. Is that no death? We, as we are, would cease to exist. We would revert to an earlier stage of evolution, and our species would prove to be another failed mutation.
It is our survival instinct, our will to live, that drives us to develop new technology, and to become even more dependent on it. As birds that once only glided from tree to tree and now rise into the sky under their own power, we too will learn to soar in our newly claimed environment. But don't think that we will still be human when we do.
With our beepers and PDA's, and our Internet access that makes us better informed (read better adapted to the environment than our predecessors) and better suited to survive. We are more fit that the agrarian society we are replacing. We are the new species. The earth will not overgrow with vegetation, because we, the new species, eat paper for a living. We burn fossils for sustinence and we belch smoke. We will for a long time, and then things will change somehow.
Just because there are not Hunter-Killer aircraft and terminators running around, just because we are not batteries, does not mean that the machines are not in charge. They are - and we are them. We have already merged, we are one.
-- What you do today will cost you a day of your life.
Hoo boy, off into the techno-spiritualism and "porn makes kids better" garbage again ...
A complete, well-structuted and footnoted criticism of everything wrong with this essay would take far more time than I dare give it this morning, but a few thoughts:
Pain is irrelevant:
That's right, the ability of an AI or a-life program to "feel" pain is irrelevant to any moral or ethical issues. It is interesting, but it is not the ethical quandry that Katz makes it out to be. Think about it for a moment -- we already share the planet with entities which are demonstrably intelligent and capable of experiencing pain. We call them "animals". They've been around for a long time, perhaps you've encountered one recently?
Now, if an AI could achieve personhood, that would be a different can of worms. But what, exactly, is personhood? That, at either an explicit or implicit level, is a crucial question in today's "culture wars." The traditional Christian answer which shaped Western culture for many centuries is that personhood is a spiritual attribute, and humans are persons because we are created in the imago dei, the Image of God, Who is Himself personal.
Therefore, (to steal a phrase from A Canticle for Leibowitz), "all that is born of woman" are persons.
The current, post-Christian viewpoint seems to be to reject any spiritual basis for personhood, and to then try to base recognition of personhood from some observed attribute, perhaps cleverness (if it's intelligent enough, it must be a person) or emotional response (if it feels pain and can articulate enough angst, it must be a person). But, the distinction between person and non-person is muddled, because (it is argued) there is no way to draw distinctions other than quantitative. So, a Darwinist would claim that humans are simply animals with opposable thumbs. Minsky, etc., claim that humans are simply carbon-based computers with a big specialized processor and complicated software.
From the Christian perspective, the issue with AIs is simple enough -- we have to determine whether an AI could ever be a person, and proceed accordingly. From that, one can proceed figuring out the ethical issues.
From the post-Christian, modern/post-Modern materialist viewpoint, there's no good way to make any distinction other than some quantitative ones, so you drop into a quagmire of muddle, providing wonderful employment opportunities for professors of ethics and for cyber-pundits.
Modern technology does not provide "choice" or "meaning":
Katz quotes Clarke:
For argument's sake, I'll take Clarke's 99% statistic as a given. It's not clear to me that a European peasant of the Middle Ages, who had a secure landholding, the ability to live off of it, and little regulation other than some taxes, had less "choice" than today's Dilbert-ized cubicle dwellers, who don't own their own homes but merely lease them from the bank, and who are at the mercy of the next "rightsizing."
It is simply ludicrous that Clarke can believe that "the purpose of meaning of human existance" has not been thought about to this point. He seems to want to have it both ways, because what is this "pious jabber" that he so casually dismisses if not the very thing he claims has never yet existed?
As for his example of thermonuclear weapons, give me a break. If anything, thermonuclear weapons have made us less able to face "the realities of war and agression" than generations past, by making war an unimaginable catastrophe. And I truly think that those for whom war meant close combat had a better handle on war and agression than we for whom war means smart bombs and air strikes.
Chesterton wrote this as a warning. It is perhaps the most frightening thing about Clarke and Katz that they seem to think this is a desirable state.
Yes, Clarke predicted the use of satellites for communications long before it happened, but he also predicted:
Large space stations orbiting the earth inhabited by everyday citizens, which supposedly would have happened 10 years ago.
colonies on the moon by today.
Travels to Jupiter in less than 2 years from now.
If any of you geeks have seen "Arthur C. Clarke's Mysterious World" you would know better than to use him to "prove" your pet theories. He's gone from a genius author to a old crackpot out there on Sri Lanka, so I would seriously doubt any predictions by him that we'll be having AI buddies in our lifetime, or in our children's lifetime.
As we see from the AI storywriting contest, real AI has hardly progressed in over 30 years. Programs get longer, machines get faster, but there is nothing near that spark of human thought or human creativity. The general consensus was that the AI storywriting machine was just fed a very long set of rules, that it really was hardly writing the story itself at all.
Today we have pacemakers. Tomorrow we will probably have more mechanical replacements for body parts, but there is currently no point in "fixing what ain't broke" in the human body. it's such massive surgery with huge amounts of drugs that have to be taken for the rest of the patient's life that I would prefer staying the way I am, thank you very much.
My final point: They predict AI robots will help us do all the heavy labor that humans normally do. We already have machines, but is there really any reason to make them intelligent? I would feel much more comfortable ordering my hamburger sans pickles with a non-sentient robot than one that actually thinks. Adding AI to those robots used to make cars just opens up a whole new can of worms. Why do that when our current solution works just fine?
I guess that as long as our boxes are independent, i.e. not networked, machines are just machines. Cut the power and anything that may be considered "A-Life" gets nuked.
OTOH, once you create a program that "lives" on the net, is capable of replication and adaptation and so on, it's ecology becomes more stable and elimination of the entity may become difficult. I see no particular reason why this type of entity should not qualify as a sort of "life" although its universe is certainly rather different than our own.
As to the concern that humans are about to become obsolete - bring it on. We tend to be highly adaptable and are certainly aggressive competitors in the evolutionary arena.
"I believe the children are our future: nasty, brutish and short."
It has been argued very persuasively that traditional top-down AI won't work (see e.g. Hofstadter's Gödel, Escher, Bach), and while bottom-up AI (be it artificial life, neural networks or evolutionary computation) has produced some interesting results (like the WEBSOM classification system), I'm still very skeptical about "Real Soon Now" predictions of AI.
Of course, I still hope someone proves me wrong (and that if they do that it's going to be "interesting times" but not in the Chinese curse sense).
I've got to admit I'm having a hard time considering machines as anything other than machines. And, open-minded free-thinker that I like to imagine myself as, I can't see computers taking over to the extent that Jon seems to be envisioning.
We are in danger of becoming too dependent on machines to the extent that one really big solar flare could kill off most of the developed world in a matter of weeks. But that says nothing about machines.
Remember the end of Wells' The Time Machine? Technology did everything, and the remaining humans were apathetic, lassitudinous (is that a word?) beings incapable of anything. This is far more likely -- and far more worth consideration -- than the 'machines will take over' cry that's been popular since the first issue of 2000AD.
Hmmm. A bunch of (rather past it) novelists predict that in n years we will be doing all sorts of wild far out things with new acronyms. How many times have these novelists been right in the past?
"just as thermonuclear weapons have made us face the realities of war and aggression, after five thousand years of pious jabber."
What? So, you mean the soldiers involved in the Napoleonic wars, who after battles piled bodies into piles so large they started to burn spontaneously like compost, did not face the realities of war and aggression? Or were they jabbering piously?
Or is it rather CNN in the post nuclear age, who jabber piously about defending human rights as they replay in slo mo for the 16th time that evening a missile hitting some black and white blob in a far off land.
I've yet to see any artificial life, or anything that comes close to it. Maybe when I do see it it will evolve so fast we'll all be slaves to it by tea time.
And as to whether androids feel pain - who cares? Do worms feel pain? Do cats? If androids feel pain, do they suffer from it? These are questions that have been asked for hundreds of years by people who have thought much harder about it than old AC Clarke.
Philosophers have a greater insight into the mind than do computer programers and authors. Try:
http://ling.ucsc.edu/~chalmers/biblio.html
-----
We are already to a point where computers require really good computers and good software for their manufacture. Try to design a chip like Merced with pencil and paper sometime.
To me, a critical point will be passed when computers become better at writing software and designing hardware than humans are, and have the ability to improve themselves in this way. We are already seeing neural net and genetic designs that work very well, but we don't really understand why. It's entirely possible that computers in the future will be very difficult for us to understand at all on the lower levels, because they are self-designed and programmed.
Nothing scary here, we specify a problem space, a computer optimizes connections and software operations to provide solutions in the space. But conciousness can't arise without self-referentiality and I wonder if this is where it will come from.