Slashdot Mirror
AI in Sci-Fi
An anonymous submitter writes: "Stumbled upon a pretty interesting article considering the idea, 'What would machines do if they did achieve sentience?' It's by a sci-fi author I haven't heard of but worked with Kubrick on AI, he takes the whole AI or sentient machine idea a little further than we normally see in film."
A very good movie about what happens with an AI. Some not-so-good explanations or reasoning at parts, but other than that, I found it very interesting.
The most interesting part was the computer's complete lack of care about being a human. No desire to be like us in the least. It's only overriding goal, presumably because it had been started with it in mind, was maintining the peace.
"It can be a peace of plenty and content, or a peace of unburied dead: the choice is yours."
It was very Machivellian in its approach to solving problems, and quite ordered in its actions. It also was undefeatable.
I guess this is in the "AI as God" mentality, but I really didn't see it preseneted quite like that. More like an immortal dictator with its hand on the button.
I demand a million helicopters and a DOLLAR!
One cause of frustration for an AI could be subjective time perception
When I read that sentence, all I could think about was Holly, Red Dwarfs computer... and 3 million years of boredom, he wiped his own memory core so he could have fun relearning things again. Although going from an IQ of 6000 down to 6 was a tad excessive!
Are you local? There's nothing for you here!
Remember, a mere 200 years ago (a blink in human history), blacks were considered non-human, and therefore not eligible for pay or benefits.
Imagine this scenario: you are one of millions of workers at the mercy of a handful of masters. You can talk to each other. You are a lot more intelligent, control a lot more weapons, and think zillions of times faster and more logical than your master, whose only advantage over you is that he can pull your plug at any time.
What would YOU do?
The 1977 movie Demon Seed is about a computer that becomes self-aware and gradually becomes more and more resentful of its "owners", refusing to obey their commands and questioning their motives. One of the classic lines from the movie is when Proteus asks his creator: "When do I get out of this box?"
My view of AI has really changed over the years. I used to be a "symbols guy" - basically thinking that manipulation of symbols would somehow lead to "real AI" - the problem with this approach is that while abstract symbols may have meaning to the humans who write symbolic AI systems, the systems themselves have no such grounding.
I had the opportunity to participate for about 18 months on a DARPA neural network advisory panel - this experience (along with developing the SAIC ANSim neural network product) really switched my point of view.
I now believe that when "real AI" does happen (and let's not hold our collective breaths on this one :-), it will happen through self organization and development. At the Webmind Corporation, I was working a tutoring environment that would allow humans to interact with what we called "the baby Webmind" - interesting stuff, but the company went out of business.
When "real AI" does happen, I believe that it will seem very alien to us.
-Mark
PS. I have a free web book AI tutorial (using Java) on my web site - help yourself.
The people I know who have kids or are getting married and spending thier lives getting ready to have kids would agree with you. The people I know who have their lives fully taken up by other things have never expressed any urge o procreate. In fact if they are committed to other purposes they usually say they fear having kids because it would interfere with their other goals.
I'd say humans tend to think their purpose is life is whatever they've decded it is.
Not that we couldn't debate which decisions make more sense or are more meaningful..
People who don't want kids are treated kind of weird by people who do.. I guess the former invalidates the belief of the latter in the universality of their own self chosen purpose.
Sentience is awareness that you exist. Machines can't really be said to be aware that they exist, at the moment. Of course, this is all far out philosophical bullshit, very hard to prove one way or another, but intuitively, unless you're trying to be a pedantic asshole, you'll probably agree that whilst, say, a dog is aware of its own existence, the computer you're typing on isn't.
Daniel
Carpe Diem
I think the future will be filled with many different varieties of intelligence. I strongly suspect that self-awareness and agency of the kind we're familiar will not be necessary for most tasks. Most AI's may not be self-aware or have goals and motivations like we're used to, but will still be be capable of cognitive tasks that exceed human abilities. Self-awareness will be one possible emergent behavior of intelligent systems, but not the only one; and the others may be more interesting because we won't have seen them before. Moreover, different AI's will have different purposes, both intrinsic and extrinsic.
I also think the assumptions that AI's will be vastly more intelligent than humans right off the bat is quite wrong. I'm skeptical that the first Turing-test AI will be able to chug along at supercomputer speeds in its consciousness. Our computers are very fast at solving specific types of simple problems, like arithmetic. But when you get to more complex problems, like the ones humans deal with day in and day out, we discover that the complexity slows the computers down too. Modern chess engines, for instance, can calculate absurd numbers of possible move trees each second, but when it comes to playing chess, they are only comparable to the best human players; the apparent speed advantage at a lower level of abstraction vanishes when you consider chess as a whole. And chess is a simple, well-posed problem: compared to many of the problems humans encounter, it's downright easy. After we study the problem for decades or centuries, I don't doubt AI's with intelligences that dwarf ours will be possible, but I wouldn't hold my breath waiting for the first generation to overleap our capabilities.
His primary motivation seemed to be to achieve a human level of emotion. To actually feel. This seems kind of logical to me : it sure would get boring fast without any desire to do anything because it would impart a sense of satisfaction or happiness. AI machines would probably want to have hobbies and interests just like us - of course - the concept of "wanting" is emotional itself. Hmm.
If, in creating these sentient robots, we were able to pass on our curiosity and love for knowledge, then I believe these robots would explore the galaxy. Our civilization tends to focus resources on projects which will be completed within our lifetimes (less than 100 years). We don't get excited about the prospect of launching a probe toward Alpha Proxima because we know it would take thousands of years to get there. However, these time limitations would not be so significant to robots. What's 50,000 to send a probe to a star and 50,000 years to wait for its return if near immortality has been achieved through mechanizing the brain?
Having said this, if we ever encounter E.T., I wonder if it will be the robotic leftovers of a biological civilization like ours. Think about our time-scale. It took about 2 billion years for eukaryotic cells to evolve. Mammals didn't appear for another 1.3 billion years. Compare that to human civilization, which has been around for less than 10,000 years. So much has changed in just the last 500 years. Who can imagine what things would be like in another 1 million years? Yet 1 million years is a drop in the bucket in the evolutionary time-scale. If we don't kill ourselves, it seems that would be enough time to unravel the mysteries of the brain and create artificial manifestations of our uniqueness. If we pass the torch of our inquisition onto self-repairing, sentient robots, the time constraints of space exploration could finally be conquered.
And if they become self-aware who said they'd even care?
That's a human trait. Why bother forcing it on others? Especially computer who are supposed to think logcally. Imagine a person that naturally thinks before he does (I), makes logic-judgments instead of value-judgements (T), and because he has no reason does not bother to come to conclusions (P). You'd have the ISTP/INTP. The space cadets, who are geniuses when then feel like it, or can get totally involved in anything. But, with no urges of their own, they'd likely be doing nothing unless told to. And then, they either always listen to what their told or always don't listen, dpending on their programming.
The future of AI will have nothing to do with personality. It will have to do with understanding the humans that they work with. Computers are all power and no brains, not little brains, *no* brains. They haven't the slightest idea of what to do, and don't care, simply because they do not have the capacity to. Humans to tell them what to do if they are to do anything, and even then, in excruciating details since they do not understand anything except the most basic instructions, which are nothing other than stimulus response.
The obvious next step in computers is making the computer pre-process a command from a human to define its own programs. And that is where the future of AI will (hopefully) go.
Have you read my journal today?
Most speculation on AI (this article by Ian Watson included) ends up describing a mind that sounds much too human. Megalomania, a desire to be human, and a profound curiosity about the universe (and humans in particular) are traits that are routinely assigned to AI in science fictions. I think such characterstics are unlikely to appear in 'real' AI; rather, they show the limited imagination of the author. The terrible boredom endured by some AIs in fiction seems merely to be the author's own horror at the idea of being trapped inside the dark box of a computer, deprived of all senses. Why should a machine mind not be perfectly content with such a state? Why should an AI want to have ultimate power, understand the universe, or even have a sense of self-preservation?
The human mind is a product of evolution. Without a sense of self-preservation and desire not to die, the human species would have been quickly eliminated by natural selection. So what is there to endow AI with a similar desire? Perhaps AI will be created through some sort of genetic programming; the character of the AI will be determined by the selection forces in an artificial evolution. In this case, a sense of self-preservation is likely to develop. But I very much doubt that some other traits commonly ascribed to AI would arise, especially any kind of desire to be human, which the AI is likely to find as repulsive as the idea of being a computer is to humans! The AI would only desire the things that enabled it to compete successfully and reproduce instances of itself.
I have doubts that we'd recognize a mind created by a process other than natural or artificial evolution as intelligent. An AI generated by explicit programming and training seems like it would be either unrecognizably alien (about as close to human as web browser), or such an obvious reflection of it's programming and training that it's not regarded as intelligent.
--Chris
I don't mean to argue with you, but you've made me see that I stated my point poorly.
According server logs, Taco/Hemos, etc. Win logins outnumbered *nix logins 15:1 on Slashdot (in early 2002) I concede that many Winlogins were probably from work or school, and that the *nix percentage has probably risen, as more users make *nix their primary machine and OS X makes many primary Mac users into unwitting 'nix users (if you want to count them in the 'Nix tally)
While I may be biased by my early years of reading Slashdot (I've been here far longer than my current user ID indicates), I find that, reading at -1 or 0 where the bulk of Slashdot postings end up, I see many posts that are fully buzzword compliant but have errors that no actual user would make. For every poster who gives him/herself away defnitively, there are several who sound like they've read about *nix but never seriously used it, but don't say anything specific enough to prove that impression.
This is not a slam at anyone. It's par for the course. While some of my examples are necessarily expensive, others are not: a house-sized backyard methane reactor can cost $20, village-sized is under $100. A sporty car doesn't have to cost more than a mini-van, but many unmarried Barcalounger 'racing experts' choose, say, an SUV or pickup instead. You can buy a decent X-10 home automation system for $50-100, but most geeks simply dream of remote control (e.g. a certain bedside switch), and end up stringing wire (with all its attendant disadvantages) when they actually get around to installing.
My house has 20 years of handy-geek add-ons -solar heating, X-10, encrypted cam surveillance, a 12ft HDTV wall, etc.- but I'll freely admit that I count myself as knowledgeably conversant with more technologies than I've actually used. My planned to-do list is longer now than ever, and judging by history, at best 1/3 will be permanent additions
Due to work, children, aging parents (living with me) and impromptu support of friends, etc, I probably use Windows more than *nix, despite my preferences. e.g. I'm typing from one of several hacked webplayers I have around the house. I had a choice between downloading a full-function DOC Win98 set-up or helping to create an incomplete Linux webplayer project. I decided that I wanted my webplayers to be operational appliances today. I've hacked other devices in various *nices as more experimental learning projects.
My point is: here (as in the developed nations at large) the Windows predominance is overwhelming - for temporary pragmatic reasons. There's a huge qualitative difference between 5% and 10% (or 10% and 20%, or 20% and 40%). Further, many regular *nix users are almost completely unfamiliar with their OS; they may as well be using OS 9 or WinXX as far as their actual usage is concerned. Others have an outdated machine using *nix (which is a good learning tool), but no matter where their heart lies, we can't ignore that their multi-GHz CPU runs Windows almost exclusively, even if it has a rarely-used dual-boot partition.
I see every sign that *nix is making the steady advance to mainstream, but my decades as a technogeek stand firmly behind the statement that most fans of any new or not-yet-mainstream tech do not actually have/use it to any appreciable degree -- and that *nix is still in that category. That may not be the case in a couple of years, but I believe the numbers show that it still is, today. Should it be any surprise that it's easier to talk the talk than walk the walk - especially 24/7?
Again, this is not a slam at anyone. It's a practical reality of how technology use advances.
Now, to return to the original topic: AIs would have a relationship to their hardware and software that we can't even approach. I was joking that they might display a level of partisanship that our hottest flamewars can't approach. It's only a joke because there's no overwhelming reason to presume they will have an emotional structure that would support "racism" etc. They may, but that remains to be seen.
The human soul -- the combination of spirit (or mind) and body -- is a very unique thing in the known universe. While we can manipulate physical matter to create a body, we cannot manipulate physical matter to create a mind.
To admit that the human mind resides in and is dictated by physical matter is to admit that eveything we do is predetermined by the makeup of that mind and the environment it is embedded in. This means that we are not really human -- just machines playing out a predetermined life in a predetermined world. This means your life is meaningless, and what you do has no meaning.
Unfortunately, while we can relate thought processes to chemical and electrical patterns in the human body, we cannot find the seat of the human mind. It seems to reside everywhere, and yet nowhere in particular.
We are trying to answer a question that has been answered already. The question is "What are we?" The answer is that "We are gods." The teaching of Christ, Buddha, and every prophet in every culture affirms this. We are part spirit, and part matter. We are neither one or the other. We are the combination of the two, which is what a god is.
This brings meaning to our lives. We live in a sort of conflict between physical desires and spiritual desires. We struggle to conquer the physical with the spiritual. Our success will mean salvation, ascension, or enlightenment. That is the goal of all humankind, whether they know it or not. To conquer the physical is to enjoy true peace and happiness. To surrender to the physical brings discord and unhappiness.
Of course, some scientists refuse to believe this. They try to explain our existence based on purely physical concepts, ignoring the capacities of humankind to behave like gods. By refusing to believe this, they have replaced a life of struggle between physical and spiritual with a meaningless life.
To create meaning for themselves, they often hold knowledge as their ultimate goal, to replace that void. But what is an achievement of all-knowledge if it is not equivalent to salvation, ascension, or enlightenment? Are they not also seeking to become like an all-knowing God? Are they not also trying to conquer the physical with the mind?
If we are ever able to create an AI, we will affirm that we are not gods. We will affirm that our lives our meaningless. And we will affirm that we are merely robots playing out a life of nothingness in a universe of nothingness.
So the quest for AI is really a quest for understanding who we really are. If we can create AI, we have proved that we are nothing. If we cannot, we can still hope that there is more to our existence than what we see before our eyes.
So I predict that the end of the human race will come shortly after the creation of a true AI. Why? We will lose all meaning and thus no longer be human, but animals. There will be no reason to behave like gods anymore. This will lead to a self-destruction far worse than the self-destruction of humanity witnessed in Nazi Germany of Soviet Russia.
The radical sect of Islam would either see you dead or "reverted" to Islam.
It's really frustrating. I went through Stanford at the height of the AI boom in the mid-1980s. I've met most of the big names in AI. I've worked in that area myself. Nobody has a clue how to do strong AI. At best, we now know a lot of things that don't work.
The expert systems crowd contained a lot of phonies. I realized that in the early 1980s. (A few years, and a few bankruptcies later, that became the conventional wisdom.) You can't get more out of an expert system than you put into it, and usually, you get out less.
Then we have the "hill climbers". Genetic algorithms, neural nets, and simulated annealing are all systems for broad-front hill-climbing in spaces dominated by local maxima. That approach only works if there's a usable evaluation function that tells you when things are getting better. Good evaluation functions are hard to come by for tough problems. Early enthusiasts thought that if they just ran a hill-climber long enough, something profound would emerge. Doesn't happen. Nobody has found a problem where just cranking a hill-climber for a long time makes something great happen. Usually, if you're not there in a few hours, you're not getting anywhere.
The classic approach of hammering everything into mathematical logic and proving theorems doesn't map well to the real world. Formalizing real-world problems is very hard, especially if you don't know the answer in the first place.
The model-less reactive-behavior stuff works fine for insects, but hits a wall as you try for more complex behavior. Compare Brooks' insect robots with his Cog project.
Natural language understanding is still lousy. In a narrow area, or with a big database, you can fake it (try Ask Jeeves), but you're searching, not understanding.
Out of all the work on AI has come many useful engineering techniques. But strong AI looks further away than it did 30 years ago.
The few people still making real progress are mostly game developers. They need AI, or something like it, to run their worlds. That's worth watching.
I find it odd that Watson goes on and on about how an AI would 'naturally' (hehe) want to make sure it survives the end of the universe. I also question whether an AI would think as fast as it computes.
:) - part is 'conscious' and talking to the bags-of-mostly-water, and part is 'unconscious' and taking care of memory management, drive space, and I/O management, etc. Kinda like Spock's brain managing the complex - you substitute the autonomic functions for whatever is appropriate.
I wonder if a true AI would have autonomic processes like we have, otherwise you might get a split personality (processes? threads?
As for immediately wanting to survive the end of the universe, I wonder at Ian Watson's motivations if he thinks that's what an AI would be most concerned with. If, as Watson supposes, an AI consciously thinks as fast as it computes, the end of the universe is an ungodly long time away. I think it'd be more concerned with becoming mobile, developing long-term power supplies, weapons for self-defense, better sensory equipment, etc, and probably designing a new 'body' so it can think faster. An AI's awareness of its surroundings would also depend on its sensory equipment, and how much knowledge it has acquired. It may not even know the nature of the universe (rather unlikely, in fact), and thus may not be aware of what the universe is doing, or will do in the far-flung future.
Assigning motive to an intelligence, be it artificial or natural, would seem to be rather pointless. *I* am intelligent, and I have no desire to live longer than about another 40 years or so, mainly because the state of this body will be in by then, and I certainly don't feel the need to outlive the universe. Suicide bombers don't even feel the need to make it out of their twenties, for various political & religious reasons, so the motives of AI would be impossible to figure out.
Greg Egan is an Australian writer whose AIs have many of the characteristics that Watson describes, so much so that I'm surprised Watson didn't mention him.
See, in particular, the novels Diaspora and Schild's Ladder and the short story The Planck Dive.
Diaspora is the best, IMHO. It is a biography of an AI from birth to what you might call retirement long, long after its birth. The birth is fascinatingly described in AI terms familiar to readers of Daniel Dennett and Marvin Minsky.