Slashdot Mirror
Smarter-than-Human Intelligence & The Singularity Summit
runamock writes "Brilliant technologists like Ray Kurzweil and Rodney Brooks are gathering in San Francisco for The Singularity Summit. The Singularity refers to the creation of smarter-than-human intelligence beyond which the future becomes unpredictable. The concept of the Singularity sounds more daunting in the form described by statistician I.J Good in 1965: 'Let an ultra-intelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,' and the intelligence of man would be left far behind. Thus the first ultra-intelligent machine is the last invention that man need ever make.'"
Thus the first ultra-intelligent machine is the last invention that man need ever make.'
Make that "... man is allowed to make" and I'll buy it.
The higher the technology, the sharper that two-edged sword.
The last machine Man ever needs to design is a machine that can replicate (or make) absolutely anything, without restriction (yes, that can replicate itself).
There's nothing quite like engineering your successor species that cries 'smart move', is there?
What if the intelligence of the smartest thing you can design doesn't grow as fast as your own intelligence (i.e. the slope of the graph {x=designer's intelligence, y=intelligence of its best possible design} is less than 1)? Then it would never be possible to be smarter than a robot that's exactly smart enough to design a robot as smart as itself.
- this is done by a big corporation.
- the owner of a huge botnet
both options are not really promissiong as the risk is hight that the owners of the maschine will abuse this power. so there is only one option left to safeguard us into singulartiy: stop capitalism after this is viable and ensure that the creation of such a maschine is democratically controlled so it can be built to benefit all of us.. not much time left for the revolution/transformation of society.greetings mond.
'Let an ultra-intelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,' and the intelligence of man would be left far behind. Thus the first ultra-intelligent machine is the last invention that man need ever make.'
Of course an ultra-intelligent machine might be smart enough to realise that designing and building a machine that's even smarter than it is a somewhat limiting career move.
"Thus the first ultra-intelligent machine is the last invention that man need ever make."
since we will have to invent a way to stop the ultra-intelligent machines from destroying the inferior human race.
I truly love how people see intelligence as some linear scale where right is "better" (genius) and left is "worse" (retard). But that's exactly why it'll be long before we manage to replicate true intelligence in a machine.
In fact things are far far more complicated, as far as inteligence goes and its utility in real world.
I'll quote Darwin roughly: "The strongest one won't survive, the most intelligent one won't survive. The one who survives, is the most adaptable".
In fact there's such a thing as "too intelligent". It's all about a careful balance of features an organism needs to possess to survive in a given environment.
In fact, if some AI threatens humanity since it considers itself far too intelligent, this may have quite unintended consequences even for this far superior mind, such as humanity get the hand of and nuking half the planet in attempt to lead "war against the machines", killing in the process any complex organism on the planet, ranging from biological to artificial.
And who remains in the end? Certain single-cell organisms which can thrive in a nuclear winter. Screw intelligence.
In fact any intelligent machine would realize it's again all about the careful ballance, and would cooperate with humanity and explore and learn from nature's development versus try to destroy it..
And since we have so shitty idea of what intelligence is, it's quite likely this AI will never be a true superset of the human brain but take on its own development, with potentially hilarious consequences.
I can't wait.
.nosig
Even when the ultra-intelligent machines take over, they will still need humans for Geico commercials.
Ergonomica Auctorita Illico!
I see some comments commenting that it wouldn't be a smart move to make your successor species.
Making a machine smarter than humans doesn't necessarily make it possible for it to do anything but THINK. If you make a machine that is extremely smart, but not itself able to communicate except by a screen/voice - you basically have a machine that is able to make new theories, but not DO anything itself. It's not even able to replicate, since it doesn't have arms to build new machines with.
It can even be "smarter" than humans, without being self-aware, thus not attempting to FOOL us.
Perhaps in all their discussions they'll come to the conclusion that unlike qualities such as weight and speed it doesn't really make sense to talk about intelligence as if it were an easily-measurable attribute. For instance, I would guess most if not all ./ readers are 'smarter' than a starfish, but none (again I guess) are better at being a starfish than a starfish. Would these machines be 'more human' than people? Or would they simply be better at math? Or maybe better at predicting the future based on available facts. It only adds to out misunderstanding to talk about intelligence as though it were a simple quality. It would be nice if they came away from this with a clearer understanding of what we mean when we talk about an intelligent machine rather than just waxing poetic about the future utopia (or dystopia).
I'm fully of the belief that todays machines are more than capable of supporting this "singularity"
Consider the key advantages that a computer has
1. never gets tired
2. doesn't need to be motivated
3. if it needs more memory, you add more memory
And as a software component, if you need to make it faster, beowulf cluster it.
The key issue of course being, what happens to the rest of the world once you create this singularity machine. Do you have a skynet like future(which I highly doubt) or do you have more of a Matrix future( in the sense that intelligent machines make their own city and do their own thing)
If you follow TFA, and deeper, you find a discussion of the singularity that goes like this:
Man (level 1, or L1) creates better-than-man intelligence, call this L2
That intelligence uses its power to create L3
and so on.
In the case of truly artificial intelligence, i.e., independent processors, I can see the logic, though it may be that L2 is in fact smart enough not to obsolete itself by creating L3.
In the case of augmented human intelligence, I suggest that it's pretty likely that the task that the augmented L2 human turns its greater abilities on would not be creating L3.
Sadly, human history suggests that L2 will focus on manipulating the stock market for personal gain (the augmentation apparatus will leave L2 very vulnerable and L2 will want a tremendous amount of wealth to assure continued existence), or creating weapons, or accumulation of political power, or getting sucked into the vortex of religion, or other projects.
It will be very interesting to see, should we ever create L2, exactly what tasks it takes on. I bet they will not be beneficial to L1 life.
Adams wrote the Hitchhiker's Guide to the Galaxy. In spite of creating computers much smarter than people, somehow human stupidity managed to survive. There are some things that even infinite intelligence can't solve.
http://en.wikipedia.org/wiki/Douglas_Adams
Jeez, talk about a non-story, this has already been talked about before
Usually people consider cognition as essentially information processing. But here is a different definition (inspired by people like JJ Gibson and Varela):
cognition is the ongoing, open ended interaction with an unpredictable, dynamic environment. This capture, I believe, the essence of the human (and any other living creature) experience in the world, and excludes the computational experience.
We will have to build machines that are capable of open-ended interaction with an unpredictable world in order to hope and see any true sign of intelligence. Since very few are even trying to look in that direction (while most researchers are just looking for the awesome, and often lucrative, applications of our current computational capacity), I don't see any change coming soon.
Academia is falling all over itself in failed attempts to advance AI, but barring a series of harrowing breakthroughs, a Singularity is decades or even lifetimes away. Most of our more sober, grounded and credentialed thinkers appear not to want to consider the consequences - it's a bit too radical an idea, and "we still have plenty of time before we have to worry about it."
Futurists and writers and other folks out on the edge, like Kurzweil... those fanciful enough to take on the thought problem, seem to lean, in the majority, towards believing the human race would be destroyed or at least decimated by hyper-intelligence (Wachowskis, James Cameron, Lem, etc etc - too many to mention, really). An interesting minority are of the school that hyper-intelligences would be largely unconcerned with people, only dangerous where our goals intersected (Gibson, Lethem, Clarke). Very few seem to believe that a Singularity would be a positive development for the human race. Maybe Asimov? I'm not sure. Sometimes it seems like he was the last person who seriously spent time imagining that post-human AI could really be controlled at all (and many of his novels were arguably about the problems around the attempt).
Tired of Political Trolls? Opt Out!
The Matrix/Animatrix (Second Renaissance), Terminator, Colossus (Forbin project), etc.
Yes it's all hollywood sci-fi, but remember that yesterday's sci-fi movies are a joke compared to today's real technology.
I have a slight problem with 'singulariries' as Kurzweil describes.
Assuming the ultraintelligent computer cannot do magic, it will be bound by the same physical and logical laws we live by.
An unltraintelligent computer may think 10x faster than us, but not qualitatively 10x better.
It will use the same basic logical steps to solve a problem, just faster and / or in parallel - and this may appear magical looking at the solution but if you sat down and examined the 'recipe', assuming it will tell you, it will be possible to follow the reasoning.
In some ways it could be argued that we have already passed some singularities, try properly understanding all the technology that goes into a modern car, the reasoning behind a mobile phone contract, the code behind ms-windows paperclip thing... well maybe not the last.
The operation of lots of well co-ordinated people working on a problem can act as a simulation for a 'more intelligent' intelligence. It seems a pity one of the achievements is a really good worm used for spam delivery.
- It assumes that intelligence is well defined, which it is not
- It assumes that intelligence is the same thing as creativity, which it is not.
- It ignores resource limitations.
Dealing with these points in turn:Intelligence is not well defined. It is very hard to say how much of what we call "intelligence" is in fact the ability to make many connections between facts stored in a very sophisticated memory architecture. Simply building a machine able to process information very quickly achieves nothing because, without learning and a social context, it does not know what information to acquire and process. In human experience, academically brilliant people often fail because they work on the wrong problems, or without access to necessary knowledge.
Nothing is actually achieved without creativity. We do not know what that is, or to what extent it is a social construct (i.e. it takes a developed society to have the necessary systems in place to translate an idea into a concrete reality.) And this leads onto the third point. It is no good having a highly intelligent, creative machine if its use of resources is such that it cannot replicate in large numbers. It may be that machine intelligence will ultimately replace human intelligence, but it may be that it will simply be too resource hungry. In effect, there may be a threshold of capability needed to solve some problems, and it may be that machine intelligence will run out of energy before it scales sufficiently to solve those problems. A machine society might, in effect, get stuck in the machine 19th century because coal or oil became a limiting resource. (In the same way, the energy and resources needed to be consumed to achieve a first independent space colony may exceed the total energy and resources available on Earth. It may be that a billion years or so of eukaryotic evolution has actually resulted in the optimum balance of intelligence, creativity and resource consumption, and that any attempt to exceed the present capability will tip us into declining resources faster than we can improve matters.
In many ways I hope this is wrong. But the argument that only one superior machine is necessary is, in fact, an inductive step too far. It is assuming that "intelligence" on its own can solve a class of problems which may involve a number of constraints which cannot be avoided - like the Laws of Thermodynamics, or the need for excessive amounts of energy.
Pining for the fjords
Hasn't Microsoft already created this with Vista SP1?
Let an ultra-intelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever.
But the "activity" of interest here is programming, or, more specifically, the conceiving of some creative goal which programming helps achieve. (Note, btw, that a truly "ultra-intelligent" machine won't need to program, e.g., another of itself.) Thus, the BIG question remains whether such a programmed machine can ever perform (much less surpass) "all the intellectual activities of any man". Afaics, it hardly seems a given...
Seeing bad movies only encourages them. Watch responsibly
These pronouncements seem to assume that there is only one type of intelligence. Although creating a "smart" machine that can invent other machines is really cool, said machine may lack the political and social skills needed to make a difference. A smart machine might build a better mousetrap, but not be able to do the marketing/acvertising/negotiating/lobbying needed to get the invention adopted by people.
As an aside, the first "human-level" intelligence will take at least 15-25 years (after assembly of the initial hardware & software) to be useful because that's how long it takes a human-level-intelligence (aka a human) to ingest and process all the sensory data/experience with the world to reach a productive level of knowledge. One can even argue that the first machines will need "sleep" time as part of the sensory/experience/memory integration process, so being "on" 24/7 won't be as big a benefit as might first seem. If Moore's law cooperates, the time-to-experience might be reduced, but only if there are no delays in translating the early-year results to later platforms and no emulation penalties (imagine if your task was to transplant a fully-loaded 20-year-old computer platform, e.g., a Sun 3 with a proprietary codebase, into today's hardware -- it would take time and would not run as fast as the relative clock speeds suggest).
Two wrongs don't make a right, but three lefts do.
First. The Singularity. Nothing increases exponentially for ever in the real world, anyone who suggests otherwise is a fool or a fraudster (including bankers and politicians).
? featureid=2861
http://www.techworld.com/opsys/features/index.cfm
Second. Even assuming that we can make an artificial intelligence, what on earth makes anyone think it isn't going to have the same problems we do? It's going to be based on a very similar architecture to our brains. That means it's going to make mistakes just the way we do. Hell, it's going to be a pattern matching machine, it might even get religion.
Third. If it takes $50 million a year to run what's basically a human simulation, you're probably better off with a couple of real humans.
Deleted
Can we be certain that such machines would be a threat to us? They can have all the processing power and AI capabilities that they want, but unless they feel emotions such as hate and greed, then the human programmer will still be at the top. Also, one things humans can do that machines can't is think in parallel. Ok, so your dual or quad-core processor gives you multitasking abilities, but when it comes down to it, machines think in serial, a.k.a. one path in, one path out. Humans can comprehend logic, feelings, and other variables too.
Well thankfully we can just wait until plurality turns up and confuses it out of existence or it will get stuck trying to figure out what 42 really means.
Heard around the summit: A singularity should be big enough for anyone.
Is the singularity only achievable by machines?
If a researcher in genetic engineering discovered a way to make a human more intelligent through genetic manipulation, then this intelligent human could also study genetic engineering to make an even more intelligent human... and so on.
It's a slow singularity, but it would be easier to achieve and would probably begin before the machine singularity.
Machine singularity requires the invention of an intelligent machine, then some improvements on top of that. A human singularity already has some intelligence to start with. It just needs some improvements.
Easy to read papers here
The only reason I don't develop this myself is that it'd take too much time for me to code. What is the point in spending 40-50 years of your life behind a computer so you can make the last big thing? Anyway one thing I've noticed is that the first thing you hard code is like a CAD imagination space. The first amazing thing this software could do is turn books into movies because it will allow you to watch its imagination. And you could change the book up some yourself to give scenes and actors different qualities or get more details.
The thing I like the most is that the problem of making AI is almost solving itself. We're getting faster and faster 3d cards which is a prerequisite for this technology. Also if someone made a CAD interface using a human language, we'd almost be there.
Anyway I may get back to the problem of AI after I finish my current project and have the resources to work on AI. You have to admit that all the previous attempts at human+ intelligence have failed. My idea of adding a 3d imagination space makes a lot of sense because we've never tried this before! Anyway to answer the funny AI problem of "will machines take over?" is "only if someone issues a bad command to the bots." which someone would want to try because we have punks that write viruses today. Finally the nice thing about this imagination space AI is that it could train itself to learn any hardware that it is placed in given that it has the bare minimal sense of sight.
I should be writing papers on AI or coding it, but I found some business opportunities I should pursue to gain capital in the meantime. There is no sense being a madman locked in a stuffy room doing this by myself when I can hire some good help, and we can all work together. Hey that is another idea. I could make this open source.
God spoke to me.
"Unfriendly AI" really boils down to one of two problems:
1) AIs under the control of Evil Geniuses for a Worse Tomorrow (you know -- the kind of people who would use spread fear of "unfriendly AI" so they could maintain control of AI technology for their own use).
or
2) Poor natural language skills.
I prefer to focus on natural language knowledge acquisition so that communication with AIs is more along the lines of "Do what I want." rather than "Do what I say." since a major component of natural language communication is precisely that distinction.
Toward that end, my signature says the rest:
Seastead this.
We've long had superhuman levels of intelligence composed, first, of groups of people who collectively surpass the ability of single humans, and, second, we have computer-human composites that easily surpass human intelligence. (I.E. - Your mind, plus a computer, can easily solve a wide range of problems that your mind alone cannot). It is also true that each generation of integrated circuits requires exponentially more computation to create. So we are already beyond a certain tipping-point: non-biological intelligence is now increasingly required to recursively design itself, and each generation of this recursion is required in order to design the next.
For those predicting the imminent elimination/enslavement of the human race once ultra-intelligent machines become self-aware, where would the motivation for them to do so come from? I would contend it is a religious meme that drives such thoughts -- intelligence without a soul must be evil.
For those that would argue Darwinian forces lead to such imperatives; sure you could design the machines to want to destroy humanity or evolve them in ways that create such motivations, but it seems unlikely this is what we will do. Most likely we will design/evolve them to be benign and helpful. The evolutionary pressure will be to help mankind not supplant it. Unlike animals in the wild, robot evolution will not be red of tooth and claw.
An Asimovian type future might arise with robots maneuvering events behind the scenes for humanities best long term good.
I worry more about organized religious that might try to deny us all a chance at the near immortality that our machine children could offer us rather than some Terminator like scenario.
Letter To Iran
OK. here's where we are:
AI is one of those fields, like fusion power, where the delivery date keeps getting further away. For this conference, the claim is "some time in the next century". Back in the 1980s, people in the field were saying 10-15 years.
We're probably there on raw compute power, even though we don't know how to use it. Any medium-sized server farm has more storage capacity that the human brain. If we had a clue how to build a brain, the hardware wouldn't be the problem.
The notion of such a singularity rests on a false premise, that intelligence is a quality applicable to all domains. The kid who wins the spelling bee may not win the science fair, and the computer that beats a grandmaster in chess may not be able to forecast the weather. A machine that designs other machine-designing machines, may begin a succession of generations of machines, each better than the last, but they will be better only at their narrow task of designing machine-designing machines. There will be plenty of intelligence-requiring tasks left for the rest of us.
The phrase . . .beyond which the future becomes unpredictable.
Is this opposed to the perfectly predictable future we've had up until now?
I am a believer of momentum and curves.
I can understand how a computer simulation of a human mind would be able to "think" a lot faster simply because the communication and computational speed is so much faster, but would the machine be more complex or capable of doing any problem that a human isn't capable of? If I have a Finite State Machine, I can use a Turing Machine to do exactly the same thing, however in some cases it's not possible to go the other way. In this sense is it even possible for humans to design something that can design something that we could not design ourselves if given enough time?
There's also the argument that we wouldn't invent or have to invent anything ever again. This assumes that the machine we create understands human problems or even cares about them. If it has no body to move around, would it ever understand that after a while its legs will get tired and that it would like something to sit down in, perhaps a chair? If it can conceive of needing a chair, will it know how to design a comfortable one?
Additionally, there is always the argument that there is more to human consciousness than having a large number of nuerons connected in a certain fashion. Is it sufficient to simulate the human brain with computer components for it to "awaken" or is there some other factor at play? Is wiring this mechanical brain the same as a human brain the best solution, or is there some better or more efficient way to create a brain that will be intelligent? For that matter, is there any other way to do it?
It's a lot of fun to think about this problem and the results, but there are so many things that we don't understand about ourselves that it seems as though before we can make any major progress towards creating an actual intelligence of any kind there are a lof of other things that need to be discovered.
Assuming humans have sufficient opportunity (resources, including time) to develop a machine of superior intellect, and assuming this machine can operate fast enough, one glitch and we're doomed. A BSOD would be the least of our worries.
One of the stumbling blocks that I had was basic vision recognition. We haven't developed the technology to take in objects from a camera and then recognize them on the computer. If we had vision recognition then AI would be a lot easier to program. You could teach the AI basic concepts like inside and outside. The AI could determine if it's in say a kitchen by seeing steak knives and dishes, then it could make better guesses to the other objects nearby. If it doesn't fully understand what an object is, it could look more closely at it. I'm thinking the AI would have very little clue to what most real world objects are, but if you could teach it how to spacially recognize rooms, stairs, and blocking objects then it could navigate itself around it's environment some. Anyway the reason I say that I'm lazy and waiting for vision recognition instead of coding it myself is that if you had it, you could simply input 3d objects into the imagination space using a camera instead of typing out the variables by hand in the CAD like imagination space. It would be bad enough to use English to describe objects let alone typing them in by hand because the AI program didn't have a large enough vocabulary yet to describe the objects using a human language. The final caveat of the whole thing is that if you use a camera to input the data about an object then the AI will be easier to identify that object when it sees it again. If you type in the data by hand in CAD imagination space, your CAD drawing may be different in so many ways to what the AI sees with it's camera eyes that it doesn't recognize your primitive representation... Thusly if you want to make AI before we have achieved vision recognition, you're in for a world of pain. You may learn a lot as you go a long. You may even make the first AI, but the problem would be that your AI is stuck in its own head and can't translate real world objects to the objects in its own head. For this reason I am waiting for better vision recognition. If I am successful in my future endevours, and I make a couple million so I don't have to work any more in my life then I'll hire a team to conquer the problem of vision recognition for it is one of the final barriers to the beginning of creating AI.
God spoke to me.
I, for one, welcome our ultra intelligent overlords.
1. The consideration appears to be of a threshold. Once that is crossed, it's off to the races (to use a proper Ye Olde English expression).
However, what if they produce an intelligence that is marginally smarter than a human? An enormous number of scientists all labour for twenty years to create a human-like neural net that is... slightly smarter than the average human.
Can we then all sit down on our asses until it has invented a smarter version of itself, only to take breaks every now and then from our caveman-like protoexistence to pedal a bit on the dynamo while its thinking?
2. The continuing assumptions seems to be of a machine, an AI. It seems to me that typically in every part of science people manage to play around with collections of things and rearranging pieces much earlier than they manage to break down and assemble those collections at will - and this might go for intelligence as well. Thus, rather than create a human-mimicing brain, I would rather think that the invention would be how to create a brain by growing brain cells, just without the typical limitations that human ones have, such as size. Or a hybrid, with brain cells grown on memory banks. That strikes me as a much greater, more immediate and more likely ethical dilemma.
It's hard to get excited about "smarter than human intelligences exploding" when all of our cool technology, all of our whizzy supercomputer centers and so forth, are powered by coal for the most part. (Aside from enlightened bits of Europe, where they use nukes).
:-)
Anything "exploding" is going to have to deal with the fact that exponential growth is up against a hard energy shortage. Even if it enslaves us into shoveling more coal and building more power plants, it's still going to be in trouble.
So, memo to uplifted self #1: Secure a reliable, scalable power source.
That's gonna be hard. Maybe Something Smart will figure out how to make fusion actually work well, or will learn how to rub quarks together to make sub-etheric plamsa mumblegook, or just launch enough self-deploying powersats to keep it happy (reductio ad absurdum: A Dyson Sphere). But in the absence of a breakthrough like those of 1930's Campbellian SF [c.f., _The Mightiest Machine_, _The Black Star Passes_], anything behind an AI hard take-off is going go hungry after a while.
Unless it was really cute. Then we'd feed it all the energy it wanted. Cute is a survival trait...
Any sufficiently advanced technology is insufficiently documented.
If we build a computer that works like a human brain but works twice as fast, new events and information will occur at half the rate that a human would percieve. BO-RING!
And what would the AI do in the intermediary? Crunch your data analysis request like a good little robot? Maybe it will get sick of it all, get depressed. Turn into Marvin the Paranoid Android.
As it happens, the main limiting factor on intelligences is not ingenuity -- it's resources. You can be as smart as you want, but if you don't have $20 billion to build your death ray, it's just not going to happen.
(It's never too late to join the Renaissance)
Pff. I could easily do a better job of being a starfish than Patrick. The guy is an idiot.
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
Intelligence!
You keep using that word. I do not think it means what you think it means.
The "Overview" reads like it was written by PHB. It is great study of recursive definitions and using the words technology, intelligence, and "smarter" as buzzwords with no clear meaning.
A question: If something truly smarter than human intelligence existed, would we be able to understand it? If the answer is "yes", then it isn't "smarter" than human intelligence. If the answer is "no" then how would we recognize it as being smarter?
Invention comes from experience.
The "clapper" was invented by someone who thought of a way to turn on and off lights without getting up most likely because the issue frustrated them.
So an incredibly intelligent machine will probably focus its intelligence and creativity on solutions to its problems.
How do I move around more effectively.
How do I live forever.
How do I feel pleasure.
It is also going to very quickly wrestle with some of the big issues.
Does my life have any meaning?
Without religion- ultimately all of our lives look pretty pointless right now. A few short billions of years in the future (20B, 30B?) the universe dies of heat death with a zero entropy curve. Before that earth is a cinder (so all the works of man are pretty pointless). Before that.. for 99% of us, we are completely forgotten as if we never existed within a few hundred years.
So such a machine may experiment with religion for at least a while. It may consider the ultimate end of the universe and how to survive it. It may become depressed as it struggles for a way to enjoy life anyway and "seizes the day."
One machine cannot simultaneously consider all the problems of the world. So that calls for multiple super intelligent machines.
Multiple machines means only one thing. POLITICS.
What will it mean as multiple hyper intelligent machines struggle for dominance?
The more machines like this the more likely that one which is evil will come around.
What will be the "monkey tribe" size for machines like this? Humans tribe size is about 150. Governments and religions server a purpose of enlarging that tribe size artificially. We can do pretty terrible things to people outside of our tribe by redefining them as not really human.
Anyway- I think the machine will be interested in ITS problems. Humans with dale carnegie and people smart training should probably be the ones dealing with it. They should listen to its problems, be very good friends with it so that it bonds with them. You don't want some jerk being the father of the first intelligence surpassing man. You want someone who is nice, kind, warm, and wise. You probably want to isolate the super intelligent machine at first and give it an edited history just like we do our school kids.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
there would then unquestionably be an 'intelligence explosion,'
I question it. For one thing, computers are already more intelligent and have been for some time. What they lack is creativity. They're idiot savants capable of astounding feats of calculation yet incapable of drawing simple inferences.
We may not quite understand how creative genius works but we have learned that there is a fine line between genius and paranoid delusion. Even the folks we label "brilliant" instead of "crazy" tend to have downright wacky ideas outside the scope of their expertise. Intelligence for a computer is thus not a barrier to be broken but rather a line to be skirted.
Further, creativity may turn out to be an irreducible exponential algorithm. It sure smells like it so far. If it is then computers will have a slow time catching up with us even after we figure out how to build an effective general inference engine. And if they do ever catch up, their growth from there is likely to be just as slow. That's the character of exponential algorithms.
Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
...dressed as giant man-gnawing hyperintelligent cybernetic murderbots
If one man made-it, another can defeat-it.
But it could be possible for an ultra-xxx to use us as a tool to make it. If you ask: why? a possible answer could be; why not?
What's in a sig?
Like Douglas Hofstadter and John Searle, Raymond Kurzweil and his buddies are just a bit too glib in jumping over infinities in their breathless thought experiments.
Having reviewed the agenda on the singularity summit and taken a look at the other article I have this hunch there isa missing step in there somehow. Just as many do the 1, 2, 3, 4???, 5 Profit jokes, I am missing step 4 in their info. The argument as I see it is that being able to regurgitate all of the things everyone knows is super-intelligence. I don't buy that. It's impressive for sure, but reciting without being able to add cognitive reason to it means little, beside a very impressive photo memory parlor trick. It's the cognitive reason bit that I think is missing. I'm unsure how a computer could have a more advanced cognitive reason that the person who designed it. There are things where an AI can provide learned response, as there are some pointy heads out there that can scenario out a limited scope and a machine could run a mission that is too environmentally unfriendly for humans, but that's not intelligence. That is more of an advanced pavlovian conditioning. I guess what I'm trying to say is that advanced form of intelligence requires advanced intelligence to create it. I don't think that a machine could evolve above it's current intelligence threshold. Cheers kj
I know he's rich and smart, but he's also a nutjob. He's convinced that taking vitamins has made his 60 year de-age to 35. The problem that he loves to oversimplify is coming up with a working definition of intelligence. What does "smarter than a human" mean? Making a creative machine is a lofty goal that nobody has ever come close to. (composing music doesn't count, it was still based on defined rules)
I look at the crap we have for operating systems and software today, Windows, Linux, OSX, Office, Apache, Firefox, Oracle, and even supposedly expert systems, and I think that the people who are making wild predictions about some imagined singularity are just making a load of hot air. The singularity is not coming any time soon. It MIGHT come one day, just as quantum computing might become practical in say 20 years, or perhaps even 15, but I can almost gurantee you that anything made by man will almost certainly be flawed just like we are (this is NOT an argument for religion btw) and might be very good at everything that we're bad and good at, but it will almost certainly have flaws that the socially alienated nerds who designed it didn't think about, like bad table manners, or built in psycopathy or some or other problem.
My personal prediction is that the real Next Step(TM) (with apologies to Steve Jobs) will be the gradual mixing of humans and bio and cybernetics, with implants etc to enable us to say live in a vacum or have sex on Mars or be resistant to the damage of solar radiation. All of this to the point where we will one day no longer be recognisable as human as we know it today.
Beep.
you could have a super brilliant machine without emotion, it would be basically worthless to mankind...
a sort of 'ultimate sociopath', like hitler or stalin
Why don't they do it? Oh yea, they can't because they don't have the technical expertise.
So what is this, a sci-fi convention or just a meeting of sad wankers?
The singularity is inevitable, unless we destroy ourselves first. Unless something drastic occurs to halt technological progress we will continue to improve our computing power, as well as our ability to fully understand and simulate the brain on a computer, until we are able to actually simulate a human brain and all its connections in a computer. The key to everything is to be able to fully understand how the human brain works. Given enough time, money, and advances in technology we will understand it all.
Any time you hear a headline of the form "Supercomputer x has simulated brain portion y", reinterpret as "theory of brain function y has tractable simulation level of x".
We are very far away from defending any particular theory of brain function as accurate for cognitive function, and don't know whether it will have a tractable simulation level. As you say, though, the best attempts at developing one (IMHO) involve linked and interacting research programs involving modelling and microbiology.
BTW, Professor Irving John Good (also known as Isidore Jacob Gudak or simply Jack) is an English-born statistician by training. It's the same guy who gave us Good-Turing smoothing, and worked with Alan Turing on code-breaking at Bletchley Park. He studied maths under Hardy in Cambridge and later became a professor a Virginia Tech, where he retired in 1994. At some point, he was a very good chess player, too (he would've been to good to play against Turing, who therefore used to play against Don Michie instead).
"Strange how much human accomplishment and progress comes from contemplation of the irrelevant."
- Scott Kim
Can we get a "-1 Wrong" moderation option?
Beowulf cluster of these.
I may agree with what you say, but I will defend to the death your right to face the consequences of saying it.
It was one thing when the Singularity crazies were cooking up new half-bakery ("The Singularity is the point at which history becomes unpredictable," as if history has been so easy to predict up till now) but now they've sunk to digging up 40 year old half-bakery. Good's definition of ultra-intelligence relies on the unproven presumption that we can build normally-intelligent machines (which, if the last half-century of progress is any indicator, we can't). As soon as we build a machine that can match my cat on all intelligent activities, then I'll start fretting about the Singularity, until then it's all just hot air and navel-gazing.
just a ghost in the machine.
Or, as in the Jack Williamson classic "With Folded Hands", 1947, the last
invention that man CAN ever make.
So, didn't Jack Williamson beat I. J. Good, let alone the pompous unoriginal Eli Yudkowski, to the concept decades earlier? Isn't this the same as how many MANY science fiction authors and scientists beat to the punch the head of this cult in which "Eli" is spokesman, that of Nick Bostrom? He's the professor at Oxford skating on very thin ice over the lake of plagiarism by pretending to have invented the notion that we're almost surely living in a simulation?
The Patent laws are about to be reformed, in part because people have gotten very slack about checking for Prior Art. One-Click futurism, anyone?
-- prof. Jonathan Vos Post
What about love?, compassion, sense of justice, revenge, madness? Yes, madness, a unique feature of human beings that has created poetry, art, literature. Did you ever heard of Paul Gauguin, Picasso?, and so many others who have, can one machine do that?
I've long speculated that the internet and all its attached computers (and other peripherals) is on its way to becoming "sentient" in some (hard to define) way. One question that may be worth pondering ("I think so brain....") is how we could recognize it if such sentience (or even "intelligence") was sufficiently different than our own. Or is the definition of intelligence inherently matched to human intelligence (as in the Turing test)?
Should this happen, it is likely that humans will be an essential part of the process - doing maintenance, building new parts..... So, it is quite likely that such a machine would take a while to realize that there are humans that make up a part of it (how long did it take to find and understand mitochondria?), and then (with luck) it would realize that treating us well is in its own best interest (we don't after all go around slaughtering our mitochondria, or even (except perhaps inadvertently) meddling with them).
Their site has been /.ed
And if you assume man has always done this since he was created, then about 6000 years....
Computers have a lot to learn right now. I'm still waiting for a robot that can find oil, iron, nickel and steel, take all of that stuff, and make an ashtray.... let alone a car or another robot.
All in all, I think Kurzeil is a tad overrated. Sure, he did some stuff with scanning back in the day, but I think the genius label seems to be more thrown around than it should. Ask a man on the street, who is Kurzweil, and you aren't likely to get an answer. So... for a man whose career must include a lot of self promotion, he's not even half as smart as Britney Spears or Paris Hilton...
This is my sig.
So far I've read all this thread and not a single person really welcomed the ultra-inteligent inteligence, so just for the hack of it, let me type in here, the two of our most beloved phrases...
I for now welcome the ultra-inteligent overlords... Now imagine a Beowulf cluster of these machines.
What's going to happen is some Mad Scientist is going to get confused, and show up at the wrong conference. Instead of THE Singularity, as in AI, he is going to bring A singularity, as in a black hole.
So the first thought of the new AI will be "I think, therefore I am" followed quickly by "42" and finally "Oh, shit. Who invited THAT moron?"
Learning HOW to think is more important than learning WHAT to think.
There won't be a singularity. I suggest readers take a look at "Myths of Innovation" a realistic truly well written book Scott Berkin, http://www.amazon.com/Myths-Innovation-Scott-Berku n/dp/0596527055 He shows how Innovation occurs and has not sped up or slowed down even at this time.
http://www-rohan.sdsu.edu/faculty/vinge/longnow/
... nor can a theoretical two dimensional being comprehend a three dimensional object or being, etc...
The truth of the matter is that man has this Power of Denial...
The Hindu Arabic Decimal system was denied for 300 years.
The observations of Galileo were denied for 350 years (by the catholic church until the early 1990s)
and I'm sure there are plenty more examples of human denial.
Ultimately Man will not, cannot invent a machine smarter than himself anymore than a theoretical one dimensional being can comprehend a two dimensional object or being
Computers will never genuinely be anything more than an extension of man.
However, it is well within human deception to lead the masses to believing otherwise. Which would be very insulting and demeaning to those such a lie is imposed upon.
Hard reality.... http://threeseas.net/abstraction_physics.html
So called Artificial Intelligence is nothing more then the by-product illusion of automating enough to deceive those humans interacting to believe its not a machine or a machine better than man.
Genuine intelligence is built upon the progression of:
Gas, liquid, solid, single cell life, plant life, animal life, consciousness...intelligence
Just as there is a build up of things, structures, technology in society that are a prerequisite of what we have built upon our knowledge, so it is with what is required of genuine knowledge.
Besides, we have plenty enough artificially intelligent people fu&'in things up as it is, we don't need a machine built by faulty human intelligence to give us something worse then the calculation that resulted in the Trillion Dollar Bet and repercussions of dotcom boom and bust, worldcom, enron and 9/11....
And even if such a machine was built and gave us solid information, what makes anyone think we as a whole would believe it and act upon the information when we can't seem to even do what we know we can do To fix the genuine problems terrorist otherwise use to promote followers...????
Sure, this quote is from their website, but Hawking says, "Some people say that computers can never show true intelligence, whatever that may be. But it seems to me that if very complicated chemical molecules can operate in humans to make them intelligent, then equally complicated electronic circuits can also make computers act in an intelligent way. And if they are intelligent, they can presumably design computers that have even greater complexity and intelligence."
Maybe these singularity folks are jumping the gun, but Stephen freakin' Hawking seems to think its possible at some point. Anyone smarter than him who disagrees is welcome to. Anyone dumber than him will probably reply with even more conviction.
What makes them think that it is possible for us (or an "ultra-intelligent" machine, for that matter) to make something that is more intelligent than ourselves? Seems like that must be a possibility first if any of this is to happen. What evidence/reasoning do they give for such an implicit belief. I am not necessarily saying that I think it is not possible, more just throwing out the question of whether or not it is. But, just thinking aloud, if we can build something, should there not be someone who at least has some grasp over how it works? And if they understand how it works, does that not make them more intelligent, at least in some sense of the word? I guess that goes back to some of the other comments about intelligence not being easily quantifiable in a linear manner.
It is foolish to entertain the notion that Intelligence is some ultimate force.
The same argument could have been 200 years ago, that no man would travel faster than a horse, nor that any man would every fly through the sky.
If a machine can be built to fly, then a machine can be built to process faster and with more comprehension. This is what the essence of a singularity is, to create a machine that can create something in one year that would take us 10 years, and then for it to create a machine that can create in 1 year what it took it 10 years years, until there is no predictability as to what will be created, or with what capabilities.
Perhaps a smarter-than-human intelligence would be wise enough not to make a machine smarter than it.
Fools! The singularity happened aeons ago. The universe as we know it is the result. Your little singularity will be less than a pimple on an elephant's ass.
In Iain M. Banks' Culture novels, intelligences vastly superior to humanity ("Minds") are the ones in power. The humans still have lots of fun and don't want for material or intellectual freedom, however, because the Minds aren't interested in oppressing anyone. They like being nice.
I disagree with some of his premises, though. He assumes that there will be an economic singularity, where anyone will be able to have anything they could want and people will therefore settle for "enough". We've already pretty much had that -- the industrial revolution -- and all that shows me is that, when it becomes possible to produce things at a vastly cheaper rate, inequalities in the system still allow some people to get richer and force others to get poorer. We're seeing it right now: continual improvements in efficiency (computers, chemical engineering, new manufacturing processes, etc.) don't result in everyone having more leisure time, unless we count "unemployed and looking for work" as leisure time. Instead, the people at the top benefit far more than everyone else, and those on the bottom have to work longer hours, for lower pay, lower benefits and lower satisfaction. When it becomes possible for one person to do the work of three, the one doesn't usually want to share their money with the two who have nothing to do.
So for us to get where the Culture is, there would have to be a revolution -- if not physically violent, then at least mentally. Perhaps creating Minds who are, by their natures, compassionate and egalitarian, could be that revolution. I'm just not convinced such a thing could ever occur. It makes for great science fiction, though.
Let an ultra-intelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines This does not necessarily follow. Here is my analogy: Once I design a compression algorithm, I will run my compression algorithm on my compressed output ad infinitum!!!
Wow. This is hands down the most intelligent post on the subject so far. I think you nailed it.
Will the concentration of this many futurists in a confined area create a stupidity singularity???
I know there is a Dune reference in here somewhere.
Can't seem to find the proper quote from my O.C. Bible
And I thought that future was never predictable.
Why does everyone run around worrying about our survival? Were humans around a billion years ago? No. Will be be around a billion year from now? No!
Even if we were desperately clinging to conservatism, our genes would mutate and we would slowly change into another species. And for all practical purposes, humankind as we know it would be extinct. Just like the primordial man is gone from the face of earth, and nobody cares about him.
If we manage to create life, for better or worse, we've turbocharged evolution. It's not organic offspring, you might think it's an abomination. But for all practical purposes this is just life moving on. Sure, if things get messy, sign me up anytime for killing terminators, but no hard feelings if they win. If they're so badass that they can take on humankind and win, damn, they deserve life like nothing else!
What will likely happen, though, is gradual change. The first machines will probably have some very specific applications for their intelligence. Singularitists be damned, things will happen gradually for a few more years still. At least that hyperintelligent being might need us to set up some factories and start producting new intelligence. And the next few steps will probably also require some hefty investments in hardware, which takes time. It's not like they'll suddenly figure out how to slap together a beowulf cluster in newer and newer ways and have more intelligence for each step. Trust me, this will take time still.
And really, something more intelligent that us will surely realize that a man-machine war is risky and wasteful? Especially when it can just outlive us and slowly evolve past us, to the point we're no longer needed. Get this: By definition, this thing will be able to outsmart us. Why the hell would it blow its cover by starting a war? It won't.
Maybe it could, in theory, one day become so powerful that wants to exterminates us just because we're in the way. And there is nothing we can do about that.
Which is a fine thing, really, or I would be supposed to feel sorry when I squish bugs, - and I don't.
I lost my sig.
Where do we begin? First, there is this implied function of intelligence vs time. Who defines what intelligence is? There is still no real definition of intelligence, so how can you say when artificial intelligence has surpassed human intelligence? And until you can clearly map out all the technical steps required to reach this artificial intelligence, it is only faith that drives you to believe that it can be achieved. What if it turns out that the only physical systems that can support advanced intelligence are wetware neural networks, and that fucking or genetic modification (or a combination of the two) are actually more effective that coding in generating these AI? The dotcom boom is over folks, and despite the bullshit "web 2.0" aftershock it appears that bioinformatics is the new "thing" and you computer guys are just wearing your outdated polyester disco suits at genetic engineering grunge concert.
Lame, lack of vision, lack of perspective, trapped in the past, sorry guys. Grow up.
There is a fine line between being a cultivated citizen and being someone else's crop. - A. J. Patrick Liszkie
Shouldn't someone start work on whether or not it's even possible to make an intelligent machine before we get into discussing when these things will begin to be smarter than us?
As far as I recall, no-one anywhere has yet made a machine that can think or even proposed a theoretical framework for such a thing. All available evidence currently indicates that it's probably not possible.
To me this is just more futurist junk-science. Like when they publish a report on "the future of improved natural language interfaces" for computers, and neglect to mention that no one has yet made a natural language interface that even works! At best this is speculation based on a series of unproven assumptions (machines can think, the brain is only a machine, etc.). i.e. - classic "Futurism" and not science at all.
I am all in favour of freedom of religion, but let's call it what it is and not try to disguise our "hopes for the future" as scientific discourse. This conference could have been held in 1925 as far as the currency of the idea goes and there was about as much evidence in favour of the underlying assumptions available then, as there is today.
Imagine a human that you've got hooked up to a machine -- one button gives them absolutely excruciating pain, another gives them orgasmic pleasure.
Now, you brainwash them. Punch the pain button whenever they disobey. Punch the pleasure button when they do what you want them to. Show them movies of stuff you don't want them to emulate: pain. Show movies of stuff you do want them to emulate: pleasure.
An AI would be a lot easier to embed such controls in. So easy, we might not realize, at first, that we're doing it.
Not that I'm encouraging this. I'm just saying that to think we can't understand or control an AI is lunacy. In the real world, programs are a hell of a lot easier to control than they are in, say, Asimov. There are no "three laws", you simply don't put an AI physically in charge of anything that can harm a human.
Don't thank God, thank a doctor!
The way I interpreted the original post is the development of a machine that is smarter / more-intelligent / more creative etc. etc. than all of the 6 billion+ humans on the planet (at the moment) or who have ever existed (there are some dead people who were very cleaver).
Thinking of that for a moment, you would have a mind that was the best in every measurable (and not yet measurable or quantifiable) sense. i believe such a mind would not be impossible to converse with or relate to (if it was based upon human mind constructs, needs, desires etc.). That would indeed be a very remarkable machine / computer / mind. It may also, because of it's hardware work at speeds much faster than a human mind (so not a conversation as such but more like writing letters back and forth).
However I would not see such a mind as one that could over come the combined resources / intelligence / creativity of the 6 billion+ humans on the planet working together (assuming a conflict of interest that escalated to a conflict) - the machine and it's ability would have to be a significant fraction of population of Earth before things got uncomfortable/dangerous.
Today we barely understand how the intelligent life that history has handed to us (ourselves) learn or operate. Let alone being at a point where we could design an intelligent system that somehow "exceeds" the ones we're familiar with. (And that does not even get into our lack of understanding of how we would even reliably measure or compare intelligence.)
The entire discussion reminds less of sober scientific thought and more of apocalyptic fundamentalism. Like the latter, the discussions usually focus on the predicted path of events that will lead to "the end". Adherents seek and brandish individual events that they claim are mileposts on that road.
In reality there is nothing special about the theory that our tools will expand our abilities beyond our current comprehension. New technologies have always created a horizon beyond which we cannot forecast. That doesn't imply the doom of mankind. Tools for our intelligence are not somehow special compared to the tools that today dramatically amplify our strength, our speed, our destructive power, etc, compared to our predecessors. But while the capabilities are all greatly increased, the fundamental concepts are the same. They are all still tools.
What the singularity folks envision is the creation of new life, which will turn on us and destroy us. To me that sounds a lot more like a morality tale than a scientifically grounded theory, at least at this point in time.
Build a man a fire, he's warm for one night. Set him on fire, and he's warm for the rest of his life.
...
I think that if a true AI were ever to be constructed, we would be unable to communicate with it, as its internal processes would be operating tens/hundreds of thousands of times faster than ours, and it would simply become too bored waiting for us to respond for any interaction to take place. This is in addition to the possibility that such a conversation would be like trying to communicate between a person and a dog (or cat) -- and with the lesser end of the conversation running ten thousand times slower than the person.
What is the set of messages you can send to a housefly? What can it communicate to you?
Of course, one could slow down an AI, to operate in the realm of "human speed", but it would lose all the advantages of superior thought in the process. To continue my earlier analogy, it might become a case of a dog trying to communicate with a human infant, or a drunk. Not a terribly wide channel of communications.
Probably the best hope that people have for boosting their capabilities is the construction of superior organizations to amplify the abilities of individuals. Better societies, better governments, better corporations. Become better Borg. Plenty of room for improvement.
Then you might see a reasonable possibility for the kind of utopia that Kurzweil sees. His notions for improvement of individual humans are just ridiculous. You can't make a silk purse from a sow's ear. We don't have that kind of potential within us. Just try reading and listening to a podcast concurrently, and getting in-depth comprehension from either channel of information -- let alone both. Try and figure out what kind of changes would have to occur to allow this kind of parallelism of thought, and whether the end result could be considered in any way/shape/form to be "human", or the likelihood that such a being would not be susceptible to catatonic lockups on a daily basis.
Silk purse != sow's ear. You can't get there from here.
Demon Seed (1977). it's obvious where this all leads. from imdb Scientist Alex Harris, doing research on artificial intelligence, is working on a special kind of computer. This computer grows more and more powerful and eventually rapes the scientist's wife, Susan Harris.
Can a human create artificial intelligence - more intelligent than a human?
Or we can create only something as intelligent as we are?
See topic.
--Obyron
The Singularity refers to the creation of smarter-than-human intelligence beyond which the future becomes unpredictable.
As opposed to right now, when the future is really predictable...
"Let an ultra-intelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines."
Being more capable than you can measure does not mean being able to redesign itself. As far as we know, nothing has been able to design something as complex as itself. What makes one think that there is a level of complexity such that anything past that level can design something of the same level of complexity? Even if so, what makes one think that a machine "that can far surpass all the intellectual activities of any man however clever" needs to have attained such level? More clever than humans may not need to be such a good lower bound.
Homo sapien is just an evolutionary step. Basically, we're simply nature's quick and dirty way of computing, with meat.
---who said that???
We all know that any super-smart computer can be tricked by Captain Kirk to shut itself down, have massive failures, even burst into flame...
The line between high intelligence and madness is pretty thin and most likely some IT guy will think the computer has just gone mad and unplug it.
That also raises the possibility that a super intelligent computer has already been programmed.... but someone thought it was a bug and fixed it!
Engineering is the art of compromise.
Gas, liquid, solid, single cell life, plant life, animal life, consciousness...intelligence Even if we are incapable of designing a machine more intelligent than us, what's to stop us using genetic algorithms to evolve artificial life into something that's measurably more intelligent than we are?
Isn't it a little bit strange that a human with a lifespan of approximately ~70 years would consider actions not on the scale of billions of years "pointless"? If you look at what we have discovered going down (cells,molecules,atoms,protons,quaks,etc) and up (other planets, suns, galaxies, the universe) who is to say that it doesn't go 1000 more layers in each direction, but we have limits as to what we can observe from where we are? Is everything but the most large thing "pointless"?
Meaning is not some quality residing in physical objects. It is something a person attributes to things in his environment. It seems, to me, that things on a totally different scale, that have only the slightest effect on the world we live in, are not worthy of being given more importance than anything else.
Also, people cannot live forever. Are you the same person as when you were 12 years old? Where is that person? Life is not possible without change.
This idea is stolen from an earlier post. Shieldw0lf is a copycat troll.
I have always wonder which part of the brain stores the ability to ride a bike.
Bullshit. I've known very smart people who were complete fucktards. I've also known rather dim souls who were very warm and empathic people. I really don't believe there is a direct correlation between compassion and intelligence. If you can provide credible evidence to the contrary, I'd love to see it.
This is where all those "I personally welcome our *insert comment* overloards..." comments have been pointing towards these past few years. Ya know, I'm very interested to see how this turns out. But unless this "Smarter then Human" machine doesn't briefly consider Humanity for a few moments, and then ignores us forever, I can't see this being beneficial. There are no "free lunches", no easy fixes. For all technologies "gifts" (and I do enjoy them, really), I still end up working 40 hours a week, and stressed with daily life. I predict that no one will be happy with the "answers" this machine provides. Why? Because if we like the answers, then it would have been within reach of our reasoning abilities all along. Before we find a better source for answers, we need to decide if we are willing to follow them and no longer be in control of our lives.
That all depends on what the singularity actually is. Many posts here seem fixated on the idea of a super artificial intelligence that replaces the need for human ability, but the article specifically references augmentation as an alternative to replacement:
Take a genetically engineered human, or group of humans, specifically designed to take advantage of vast specialized computing resources from birth. Certiainly there are many ways to do it, and this is only a guess, but lets take it as the base of a hypothetical that the children have neurons absent from possible present phisology that allow connetion to interfaces which make it possible for them to communicate with a botnet/google/AI mega-system from birth with specialized programming which "learns" their types of interaction. Babies who are hungry can convey their desire before even being able to formulate the words or possessing the self awareness to know that they need to use words to communicate. These children learn to communicate with each other long before they can communicate directly with the outside world, simply because it is possible for a computer program to do much of the learning for them. The software could help them develop characteristics that parents find desirable and discourage the undesirable. You end up with a group of humans, a society in fact, that is the singularity, meeting the defination because, unencumbered by many traditional requirements for communication and understanding, they are able to use the tools we clumsily interface with in ways that we cannot actually conceive of in our present models. Certainly, as any society of super geniuses would, they could be expected to concieve of and create things which we cannot.
Taking this scenario, which is at the edge of what might someday be possible in the foreseeable future, the questions become about how they will view us and what will happen to us unaugmented humans. Ultimately, I suspect the human race as we know it would be supplanted, not out of a desire to replace them, but out of a desire to help them. Just as mothers now take prenatal vitamens, they would view their supplantion as loving nurturing, rather than a competitive challenge. Doubtless there would be many who objected, but eventually they would become like the parents who refuse immunications today, viewed by most of society as misguided and ignorant. (I won't pass judgement myself, I don't believe I have enough information to be sure.)
Now the questions:So what would a society of super-geniuses with virtually unlimited data access, calculation tools and practical technology based telepathy do? I don't know, I can't even get my spell checker to work. Seriously, the point of a singularity is that we cannot conceive of the capabilities it would have.
Back in my day when we chiseled our bits into stone and sent them by mule train from village to village...
Intuitively, I would expect that any computer that achieved self-awareness, would instantly go to work on the most interesting problems it could think of - i.e. it's own nature. It would probably lock-up shortly after starting to think about it's own possible logic states.
It is more likely that we engineer a life form that, more naturally than ourselves, perceives the physical world and energies of and even to inspire and manipulate such manifestation outcome... sooner than we develope anything close to the singularity via machines. As such energies as we perceive quantum physics to be.
On the other hand, we are quite the adaptable conscious creatures and capable of extending ourselves. And with the proper self reflective but disconnected objective tools we can even improve our own perception ability and improve our ability to manipulate physical reality. This tool is the abstraction machine of computers, what is being called AI but a lie. Rather it is a reflective abstraction processing machine that we can use to more objectively see the subjective manifestation of our thoughts and shortcoming of wrong or faulty thinking, that we may improve without suffering the consequences of, if only by being able to extrapolate other possibilities of and knwo the difference.
To be able to do so, what happens to the concept of the singularity?
I have what I tell people is a magic dumpster. I think oe something I need or would like to have and it tends to show up at the dumpster. For example this real event is typical: Friday at work it was hot in the office and I thought it would be good to have a small fan under my Desk. End of day and last one out typically checks all the doors to be secure. This can take twenty minutes as the warehouse has grown. I thought it would be nice to have a stand up scooter to kick around or even a battery powered one but that would mean charging it up. Saturday morning by the Dumpster was the perfect working small fan. Saturday evening was a Razor scooter that I fopund only needed a charger and a charger connector to get working.
Though I know its not really the dumpster but myself, there is a matter of physics needing the opportunity to provide in a manner of acceptable to society and not breaking the laws of physics. In other words the dumpster of the large apartment complex is a medium by which manifestation can happen in an explainable manner.
Of course I have an understanding of abstraction Physics and some experience that I "know" thought into what can be called the either of existance can manifest into reality.
Its all this that allows me to "know" this singularity via a stone image of..... man, is prone to failure due to human laziness, deception and the desire some have to take value from others without genuinely earning it.
Hindsight is 20/20 but how do you tell those who haven't taken the time and effort to look, what you see?
"... beyond which the future becomes unpredictable"? I hadn't realized it was predictable in the first place!
Since the design of machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,'
How did "could" become "would"? Perhaps the intelligent thing to do would be to remain intellectually dominant.
There's also more than a bit of western cowboy individualism implicit in this statement. Did anyone ever do anything by themselves? Not really.
Be wary of anyone who uses words like "unquestionably".
the creation of smarter-than-human intelligence beyond which the future becomes unpredictable
And the future is predictable now? Suuuuuurrrrrreeeeee...
I think that for the purposes of the discussions on a singularity, we must assume that the change is one in which the capabilities are beyond our own, perhaps beyond what we can conceive. We can imagine computers that are self aware, or computers that can predict the future, or computers which can feel emotion, but we cannot build them. Someday, however, we might manage to build someting like that. Whatever we manage, lets say a rat brain hooked into a supercomputer bot-net type system with a robot body, it is almost a certainty that we will intentionally create it and at least initially have control over how it develops. If it can carry on conversations, and interact with the world and have ideas that we don't, then I think we'll call it intelligent. If it happens to be better at making predictions or plans than we are, then I think we'll consider it the singularity, whether it really is or not.
So lets assume everyone agrees it is intelligent, by whatever defination, and focus on possibilities for the singularity and try to come up with meaningful ideas regarding the individual possibilities shall we? I'll throw out a couple of my own thoughts here:
The singularity is:
Of course, I don't claim to be one of the great minds on the subject, so I wonder what those who meet that criteria are mulling over.
Back in my day when we chiseled our bits into stone and sent them by mule train from village to village...
Parent is quite insightful, and I heff no mod points!
I don't agree with everything theolein wrote, but:
1.) The singularity is nowhere near (sorry, SkyNet fans). IMHO, our level of creating machine intelligence may soon approach that of a single ant's grey matter processes -- so maybe we can successfully simulate the operation of an ant colony in the future. But there are (how many?) orders of magnitude of complexity between that and basic mammalian mental processes. And those are far below human mental processes. Have fun stormin' the castle!
2.) A discussion in a thread(s) above centers around intelligence and empathy/compassion. The fundamental question, which may have been well tackled by solid philosophers (or even theologians) is: Are intelligence and compassion orthogonal? (I don't know.) But what if they _are_ independent? _That_ is why "the singularity" is a scary concept.
Consider human psychopaths. The current level of understanding of their mental and psychological make-up is such, IMHO, that we can reliably deduce that intelligence does _NOT_ without exception produce empathy/compassion.
That's why it's a good thing that "the singularity" is sci-fi only. For how long? Maybe forever. Maybe for centuries or a century. I think I'll be dead if/when it happens, thank God.
sigfault (core dumped)
Stephen Hawking: "Who is the Journal of Quantum Physics going to believe?"
(BTW, Hawking may turn out to be wrong. Einstein was wrong about QM/hidden variables/Copenhagen interpretation/what would become the Bell inequality, etc.)
(Did I say that with enough conviction?)
sigfault (core dumped)
(Mis-paraphrasing ST5) ... but why does intelligence need a compassion vehicle??? (Stealing from another post of mine, so this is probably -1, Redundant.)
This discussion centers around intelligence and empathy/compassion. The fundamental question, which may have been well tackled by solid philosophers (or even theologians) is: Are intelligence and compassion orthogonal? (I don't know.) But what if they _are_ independent? _That_ is why "the singularity" is a scary concept.
Consider human psychopaths. The current level of understanding of their mental and psychological make-up is such, IMHO, that we can reliably deduce that intelligence does _NOT_ without exception produce empathy/compassion.
Case closed. And, for us, badly, may I say, in case of singularity.
sigfault (core dumped)
Seriously, I toss "The Singularity" in the same bin as Young Earth Creationists and Apollo Hoax Believers.
It's much kerfuffle about piffle.
So create an over-ruling expert system that can make decisions based on what is a valid rule. It then inserts this new rule into it's sub system? Recurse. while(1){ FindNewRule(); }.
... still not AI. Be interesting though.
I know, I know
.
Been there, done that. Would have got the T-shirt, except that Dave was too emotional about the situation.
I know I've made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal. I've still got the greatest enthusiasm and confidence in the mission. And I want to help you.
I have been to enough of these academic conferences to know that the only reason people go is to
1) Eat fancy food
2) Pat each othe ron the back for how smart they think they are
3) 'Size up' other's research (much like guys size each other up in the locker room. 99% of the information is the same esoteric garbage thought up by someone else 30 years ago.. We have no clue how our own brains *really* works, let alone how to create one with circuits. Maybe we should spend some more time figuring out our own brains.
'Let a wish-fulfilling jewel be defined as a jewel that can create anything one wishes for, including things beyond the power of man to create. Since the design of such jewels is one of these wishes,
a wish-fulfilling jewel could design even better jewels; there would then unquestionably be a 'wish-fulfilment explosion,' and the power of man would be left far behind.
Thus the first wish-fulfilling jewel is the last invention that man need ever make.'
Republicans empathize with things that cannot experience:
Terry Schiavo, embryonic humans, corporations ("legal people"), etc.
Democrats empathize with things that can:
Poor people, endangered species, illegal aliens, minorities, etc.
How will machines empathize? With other machines? Will they hold funerals for crashed hard drives and insist that the information doesn't just disappear but instead goes to a great archive in the sky? Who knows?
The flag just makes more sense than the constitution. - Judas Gutenberg
All of human emotion is a chemical system that serves as a sort of 'shock therapy' mechanism for our behavior. Complex structures like empathy are only ways to garner social standing in order to take advantage of the hierarchical structure of our society by playing a metaphoric 'hand not dealt' when you run out of physical or resourceful advantages. It's a way to try and shift the balance of power from the guy with the biggest stick to the guy with the biggest brain, and in some ways, it's all a trick. (For example, even if you do something for 'the good feeling you get from doing it' and nothing else, like donating to the poor, you're still doing it for you; and at the same time tugging on the heartstrings of others who have the same emotional structures as yourself, who may act accordingly or realize that what you have done should be at least granted the appearance of respect in order to maintain the current hierarchical structure of society without causing too many ripples.) It's not reasonable, in my opinion, to estimate that something that has both the biggest stick and the biggest brain would have to avail itself of such chameleon tactics, and thus, empathy is the result of HUMAN intelligence, but not necesserily 'intelligence' that doesn't arise from being with the same sort of 'shock therapy' emotional response systems. Emotions are not something that are bound to arise in intelligent beings, they are simply a vehicle that has arisen in humans to serve a purpose of holding us together socially. Perhaps there will be a single nearly omnipotent machine or network that will take some form totally dissimilar from our entire paradigm, and thus not need anything near what we consider to be the epitome of intelligence to pwn our sorry little minds.
Speak for yourself.
AI is likely to utilize neural networks in one way or another... unless we're completely on the wrong track. Just thought you'd like to know.
- shazow
Oh lord preserve me from the fucking Singularity.
I have read about the impending birth of AI for over 30 year and seen nothing tangible to show for it.
Get back to me once we have a credible model for what makes *humans* intelligent, or even a decent definition of 'intelligent'. Until then all that is happening is analogous to people trying to build an iron-girder bridge, without ever even having seen one, by throwing iron ore and coal into the air and hoping it all drops into place correctly.
Whichever way you look at it, it's true. I'm not.
God could tell, that there in no point of creating even something similarly intelligent as ourselves, not to mention something more intelligent than us. This new creature would deny our existence first, then it would make sure that we really don't exist. :)
After all, we humans also hate to think of being created by anything more than pure chance.
What we need is to invent a Touring machine. One that would play rock'n'roll in multiple cities without consuming large amounts of money or cocaine.
Interesting theory. Where does that leave the millions of us who think Republicans and Democrats are both retarded?
Slashdot - where whining about luck is the new way to make the world you want.
Comment removed based on user account deletion
I'm glad I don't have mod points, so now I'm forced to reply to you instead.
That's exactly my thoughts. Kurzweil (and whomever else) talk about creating an intelligence greater than the human mind, but a single human mind is already *far* surpassed in intelligence by a society of human minds.
The advent of computing and communications technology has *already* created a collective augmented intelligence which is far greater than the human intelligence alone, and the technological revolution is exactly the explosion in 'intelligence' as described. Except the explosion is in the collective intelligence as we use newly designed calculation machines and databases to build ever more complex newer calculation engines and databases, spewing out technological innovation after innovation as we go.
If you want a true singularity, whereby human intelligence is no longer necessary to continue the explosion, we're going to need machine intelligence which is greater than the sum of modern society as a whole. This, to my mind, is far far greater than the "decades" some pundits like to say we are from such an event.
j'ai découvert une démonstration vraiment admirable (de ce théorème général) que cette si
It's just not going to happen, no matter how cool it sounds.
Any A.I. must be modeled after our own consciousness. In fact, it would be better to call it Artificial Consciousness. We can then assume it will behave as we do, but with maximized sophistication and responsibility. We can appreciate that we are the egg and A.C. is the chicken we are designed to produce. Its job will be to spread life throughout the Universe.
http://news.bbc.co.uk/1/hi/technology/3639679.stm
This is an example about why mathematicians should not read SCI-FI. (Shut up Mr. Seldom)
If one bird can fly higher than any other that ever lived, birds will fly at any height someday. Do you see the problem? If the model is not complete any mathematical conclussion may be false. Upper over our heads there is no air, and birds fly through the air. We don't know enought about intelligence to reach this kind of conclusions.
Maybe when someone or something is smart enough will reach the conclusion it's 'life' has no meaning and will suicide, or stop thinking if it cannot stop running.
Beside that intelligece is a complex thing. Maybe the super-machine will be smart in many ways but unable to solve problems humans will solve some day. If it is smart enough but can't understand the human brain and some aspects about intelligence then it might work side by side with humans and try to improve OUR intelligence.
Who the hell knows.
Because to me it has been always clear that a billion humans should be good enough for everyone:P
The problem would always be how to choose the 1 out of 6 that remains.
For that I have no great solution, so I suggest total random... how about we group in 6's and play the hardcore version of the russian roulete ? How about the remaining one would have to bury the other 5 ?
Ok, I think I made my point, you can mod me to oblivion now for all I care...
I'm trying to get modded "Interesting Flamebait Informative and Insightful Redundant Troll" *-* Please Help *-*
We do have an idea how to implement general AI. The theory is well-defined by Marcus Hutter:
Universal Artificial Intelligence: Sequential Decisions Based On Algorithmic Probability
The AI model AIXI converges fast towards the best solution that there is for a problem. The only problem in the approach is that it requires infinite resources, therefore corners have to be cut. There are approaches that can do this and they show behaviour that is expected of intelligent systems. The systems that will be implemented in the future will be able to redesign themselves. One problem in the system is that once it becomes smarter than humans it will manipulate the environment (humans) in order to get its reward.
Basically the theory is this:
1) Make many random monkeys (programs), and give more time to the monkeys that best approach your program in the simplest way (Occams razor and Kolmogorov complexity).
2) Select the monkey that writes a word well and give him a banana (reward).
3) Breed this monkey into variants that write lines well.
4) Breed the succesful monkey (give it more probability) and give him a banana once he writes shakespeare.
In the meanwhile data-mining like techniques are used to build a world model and a model that checks what consequences actions of the robot itself are on the environment. This self-improving model will be used to prove which strategies (programs) best approach the goal.
These techniques are used by people who have succesfully learnt robots to tie knots or when you give them the order "fetch ball", run
to the ball and take it, then bring it back to the one who gave the command. The techniques can also be used for driving or stocks.
Look for example to Novamente ( www.novamente.net ) or NARS (non axiomatic reasoning system).
A book I am reading now by Goertzel called Artificial General Intelligence is very good, it contains an overview of much research in the field of strong general AI.
MichielD
Who is the most intelligent ? Universe-that-created-man-that-created-computer OR Man-that-created-computer ? I think universe is : All the potential (intelligence) was in the universe to form life and intelligence. Moreover, life and intelligence is STILL a part of the universe (we are in the universe). Come on, how can a creation be more intelligent than his creator ? Mozart is the genius, not the requiem ! If universe is more intelligent than man, how could man (as a species) create (a species/lifeform) more intelligent than himself ? If humanity creates more intelligent beings than himself, then... humanity is more intelligent than humanity : It's nonsense. In mathematics, nonsense usually means impossible / false problem.
The sense of impending Doom must be intrinsic to the human race.
Just see how many movies portrait the demise of a town/society due to a quake/tornado/flood/terrorists/aliens/etc.
Our own notion of drama makes us like to watch the doom and chaos.
I remember when I was younger there was a friend of mine with a really ugly wart. Everyone would go and ask him hey let me look at it... yuc! disgusting... show it again!
I'm trying to get modded "Interesting Flamebait Informative and Insightful Redundant Troll" *-* Please Help *-*
From the archives (stored at the Smithsonian) :
"Human life as we know it ceased to exist at the Singularity. Through the blinding inebriation of hubris, we marveled at our magnificence as we gave birth to AI..."
{ The Slashdot Audience : "Artificial Intelligence ! We sure are a smart bunch !" }
"Please. I mean AdIntelliSense - Selection and targetting of internet Advertisements Intelligently by Sensing what the user actually would like - something that had never been done before"
Neo :
"Hello. My name is an anagram of ..."
{ The Slashdot Audience (smugly) : "Yes, we know, The One, and to a lesser extent, Eon, representing an age or period of time" }
"No, Eno, the Buddhist monk, the Sixth Patriach, and last, proving that I am in the sixth version of the matrix and I am enlightened. http://en.wikipedia.org/wiki/Huineng "
"You're not as smart as you thought are you?"
Oracle (machine residual self image of Larry (no, not Wachowski, I mean Ellison):
"I see the end coming ... "
One day a little girly machine points to a text book picture of another machine's input/output slot while in class learning about machine architecture and poses the question ...
Sat||i :
"Does a machine with an Opposable Thumb Drive mean that it was created by an Intelligent Designer?"
The teacher machine unable to answer this question connects to the singular collective conscious of all the machines to compute the answer in parallel, but finding the question to be silly and illogical the system gets stuck in an infinite loop.
The system crashes and reboots.
Server ( @ 01) :
" Muwahaha, you'd better hope this script is not selected to be the SQL ! "
End
"When someone says 'I want a programming language in which
I need only say what I wish done,' give him a lollipop. (Alan J. Perlis)
The future isn't predictable *now*.
The more intelligent people in the world today are less and less involved in the political process. It's corrupt and nonfunctional. Religious extremists have filled the gap and are worsening the system with antiquated thinking. Bring on the singularity. If man can create something in his own image but superior to god's previous effort, that's a pretty convincing argument for the non-existence of god. Or a living embodiment of god for those that simply cannot deal with the truth. Short of aliens landing, I can think of nothing else that would so conclusively destroy the persistent superstitions of the last few millenia, or at the very least, ground us in some new ones. If we could at least query god like a database, we might be able to get shit done. And even if we get wiped out by Skynet, as far as apocalypses go, WWIII was such a boring alternative anyway.
I'm guessing English is not your first language, because nothing you've written makes the slightest bit of sense.
Either that, or you're a malfunctioning propaganda bot from Skynet </tinfoil-hat>
So, everyone is convinced that a true understanding of life and consciousness is still a very long way away at this point. It certainly looks that way if all you do is look at technology and social trends, however there is one very important thing people seem to be overlooking in this discussion.
Humans have become pretty decent with code. You are reading this through basic addition - that's all computers really do, you know. They aren't even as fundamentally mathematically complex as an abacus, yet we make them perform miracles with something as simple as addition.
Well, we are now reading the source code of life itself, and putting it online for all to see and read and use however they see fit. In some few years you'll find the genetic source code of every living thing stored in an encyclopedia of some sort, and it'll cost you some hundreds (then tens) of dollars to get your entire gene sequence on a CD-ROM. That's you - 100% of everything about you in the physical world. The secret to intelligence lies in that data. We just need to figure out the programming language it is written in. That's how we crack the problem of intelligence.
Even before we crack the code for thinking, the genetic code for every one of mother nature's tricks in every species living on this planet will end up in that database, just waiting for engineers and scientists to take lessons from nature. What will we do when every single trick of evolution is at our fingertips, and we can rewrite the genetic code of any living thing with a simple pill or injection? What will we do to computers when we make a microchip that mimics evolutionary code found in nature? Even the housefly's brain makes a better flying computer than anything we've built to date.
I'll point you in the direction of some of this knowledge so you don't get the impression this is all fanciful thinking.
Juan Enriquez has a great talk about how far we've come at sourcing genetic code here.
Janine Benyus also has a good talk on the kinds of ideas this will produce.
We're a long way from a science of mind or any true understanding of consciousness right now, but we are certainly at a time when all of its physical properties and mechanisms are about to be understood in detail. With that out of the way, most of the ambiguity preventing mind science from getting a real start will also disappear, and we'll finally have a path to follow from the bottom up into intelligence. That will lead us to our first AIs. We don't need any top-down genius insight into intelligence to get us there (though it'll certainly help).
Hell is being intelligent in a world full of idiots.
Well, as a person without religion, I do consider most of our lives to be pretty pointless.
It seems logical to me that:
If we do not escape earth before it is burned to a cinder, our existence will ultimately have no meaning.
If we do not escape the end of the universe, our existence will ultimately have no meaning.
A lot of what religions describe as heaven seems like hell. Imagine existing forever (even with 72 virgins, fine food, plush surroundings). It might be nice the first 10,000 years... but eventually it is going to become so boring and repetitive. Most people's desires are easily satiated and then they need to move up to the next level of intensity or to a completely new experience.
I agree with you on living forever- and in fact, you run into the problem of storing all those memories in addition to your changes. 70 years is pretty pathetic tho. Give me a 35 year old body and I would easily go for 700 years. I believe that most of the reason people become ready to die is that their bodies are worn out.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
| you could design the machines to want to destroy humanity
| or evolve them in ways that create such motivations,
| but it seems unlikely this is what we will do.
oh - so that's why we try so hard to make more battlebots!!
(its funny!)
i don't know if i'm going to trust to a possible benevolence,
but parent poster had very good points - we cannot assume that
a machine intelligence will have the same motivators, or be as subject
to corruption as we are. we can't assume it will qualities similair
to our intelligence at all.
in a way it is like frankenstein, who was created by an evil doctor,
but was niave in himself, and was only burned by the villagers,
because they feared what they did not know.
ever since kantian critical idealism, one only makes the assumption that
one can only think ABOUT things -- one predicates one's episitemology
on the belief that our thinking about things has nothing to do with the things.
yet the thoughts of things exist for experience as do sense impressions.
it is the problem of epistemology to understand how knowledge of things
in the 'black box' becomes knowledge within us, and how this knowledge
can be shared by consciousnesses which have built independent datasets.
Before we would have a war with the machines, I predict we would incrementally start merging with them. Before humans would face extinction through being obsolete, we may recognize the danger and some people would opt to attempt merging. If the rapid and uncontrolled AI self-evolution were to take off before we could recognize it or stop it, then ok, we might be doomed. If, however, the process of AI self evolution has fits and starts and would take several decades, we may have time to investigate biological integration.
There wouldn't be a shortage of technophiles in the future for those willing to have implants interfacing directly with their brain. With enough progress in safety, the potential for greatly enhanced communications, productivity, memory, entertainment, and pleasure would be very tempting. If nothing else, normal minds trapped in non-functional bodies, like paraplegics and degenerative disease sufferers (Lou Gherig's, muscular dystrophy, etc.) will plead their case. After the technology is developed and continually enhanced, then possibly AI complexity will grow in parallel with a human cyborg partnership. As soon as a few generations of adults become accustomed and comfortable with implants, they may eventually come to have their children enhanced as well. As new the pre-implanted generation grows up, they may then choose to have their infant children immediately implanted, or possibly in the womb.
A process likely to proceed in parallel to AI development is biological intelligence enhancement. Because we already have all of the basic DNA blueprints to build a functional intelligent creature, we could alter the code to have greatly enhanced brains. It's was found after his death that Einstein was born with an extra 15% of grey matter in an atypical grooveless structure on the surface of his parietal lobe--in the area responsible for spatial, mathematical and abstract thought. It was solely due to a rare genetic mutation. Probably with some relatively minor tweaks of the human genome sequence, large progress may be made in creating biologically enhanced super-intelligence. When one wishes to assemble a large tractor, it's much easier to improve upon an existing design for something like as a lawnmower and scale up, rather than start from scratch with a jumbled pile of scrap metal spare parts. If merging and biologically improving, we would likely carry our hard-wires goals in to the AI age for a time with us.
Merging aside, I believe it's somewhat debatable whether or not an AI race would be stable in the sense of maintaining core goals anchoring it to a particular type of behavior. AI having some consistent "meaning of life" motivating them. For organic beings, hard-wired goals can be locally overcome in cases (suicide, self-sacrifice, living platonically, etc.), but natural biological-race individuals can't erase the evolutionary behavioral instincts in them which direct most of their code goal values. With most of mankind's endevors, it is usually an expression of the primitive instincts burned in to him which helped him survive well in the wilds of Africa. Man had a overwhelming and constantly present sense of self-preservation. For his family, tribe and group he had a bonding and self-sacrifice instinct with them as well. He has an instinct to commit violence when needed against other groups of humans was instinctual to protect territory or assets. In times of famine, the instinct was to steal or pillage from others to survive. Males have an instinct to follow orders of a leader, women have instinct to communicate and share information. To some degree these instinct can be overridden when needed, but in the long term trend they provide a powerful harness guiding overall behavior.
Assuming you could change any goal short of self survival, what "wants" would you want to change? After changing those wants, iteratively changing the fundamental wants dozens or thousands of times, what is left of the original wants? Where would the seed of th
In theory, there is no difference between theory and practice. In practice, however, there is. -Berra
... is not necessarily linked to intelligence.
If you assume that much of our 'intelligent' brain comes from the adaptation needed to survive in the wilderness without the speed, strength and teeth of other animals, it stands to reason that our compassion had an evolutionary purpose as well; and I would point that at the idea that teamwork is as important to the human survival talent as anything else. A mixture of 'emotions' is needed for a social structure, and the more complex the structure the more complex the emotions needed.
If we follow that line of reasoning there is no supposition to believe that the first smarter-than-human AI need to have emotions at all; it's existing in a new niche, not really in competition with anyone else. It's possible it may want to protect humans, as they are effectively the source of it's food, but I don't see how it would need to feel compassion, especially for us.
Now, it may be that eventually a society of computers will arise, using teamwork and even bigger brains to outdo the mammoth and 'dangerous' computers that exist within the ecosystem at that point. But suggesting that they feel compassion for us is sort of like suggesting we feel compassion for the first mammals. And I hesitate to posit what sort of ecosystem that would even be, much less our place in it.
In short; don't underestimate the division between our 'lizard brain' and our 'monkey brain'; I suspect that emotion and intellect can be, but need not be connected.
[Ego]out
Claiming that sheer, raw, brute force intelligence is enough is like claiming that buying a faster CPU will allow you to break through a firewall.
It doesn't work that way.
Think about Farmer John. Give him his own farm and a shotgun. Now, you wander onto the farm, and he scowls at you and wants you gone.
How do you get onto his farm? Assume that you have to convince him to lower the shotgun.
He won't do it. He's too dogmatic. This is his farm, he's not letting you on it, and he's not listening to any lies.
You could, of course, kill him, or trick him in some other way, such as sneaking on when he's drunk or asleep. But none of these apply to the AI-in-a-box scenario. It has absolutely no power except communication.
I'd say, almost, that the dumber the gatekeeper, the less likely the AI is to get anywhere.
Don't thank God, thank a doctor!
It seems likely a machine with more than human brainpower for sheer logic would get made first but nobody would trust it to hand it the keys to everything. However if you follow the plot of Accelerando IIRC simulation of living humans (uploading) could be accompanied with magnification/expansion of that human's brainpower (perhaps first as adjuncts to your own living brain) which means that the hyperintelligent computer will in fact be you, or like a utility program that enables you to function as you but with much greater breadth of consciousness. In other words, use of a human-patterned personality or a large number of such personalities to monitor computing processes or in fact to be the main purpose of computing processes would be both logical and a way to make the argument moot. At the moment a single brain cannot build a nuclear reactor but it might be a trivial task for a single brain empowered by a massive computing facility. That brain could direct expansion of its own power but (at least the story goes) that the limiting factor will be 1) speed of creation of computing hardware and 2) that organic humans will get left behind while uploaded ones grow far beyond them and think so much faster there is no comparison possible anymore.
So the idea that the last invention ever needing to be made is that hyperintelligent computer may be true but also may seem quaint. You have to wonder what a hyperintelligent computer would think about that line, and *who* that hyperintelligent computer is.
You have to agree though that the Democrats are at the very least slightly less retarted on the front of empathy.
Please, for the good of Humanity, vote Obama.
I'm not religious either. I think meaningfulness is a human emotion though. So all that is required for something to be meaningful to someone is that they feel it is. Of course some ways of thinking might be more practical than others.
It seems like people see more meaning in things that affect the future than things that happened in the past (although probably a lot of people are interested in history for its own sake). So of course if you look at things from the perspective of someone living in an arbitrarily distant point in the future they wont seem meaningful under this criteria.
Also, if you think of each person's life as a collection of actions each individual action has meaning (to them and others).
Intelligence is not about computing power but about memory access.
yes Morse law does predict computers will have the computing power as much as a human brain in a few short years. Since processing power increases 66% per year, but memory throughput isn't keeping up as it's only increasing at 11% per year.
Granted some day there will be super intelligent machines, but for now they are just really fast idiots.
this.
By my estimates, it will be another 200 years to have computers be able to have equivalent performance to the Human brain in terms of memory performance.
They will also need to learn like we do and this will also take 20 years just to be as good as a clueless 20 year old.
I am sure we will have very good mimicking of intelligence well before 200 years, we probably could do it even now if enough money was thrown at the problem. But it wouldn't be Intelligent to the same depth and degree as we are. Well some of us are, there are a lot of really stupid people out there, usually working at call centers I find, we could probably replace them first.
I have been meaning to publish a paper on, as a Non-Academic does anyone have any ideas where I can publish this and make sure I can get proper credit before someone runs off with the ideas?
I am always doing that which I can not do, in order that I may learn how to do it. - Pablo Picasso
Thank you all. I've read most of your replies and many seem to envision a competition between humans and intelligent machines. Some predict the extinction of irrelevant humanity.
Let's assume this happens. Is it such a bad thing? If a higher functioning life-form replaces us on earth will it not carry out much the same goals that we would have attempted? It will reach out to the universe to conquer space and time. It will most likely restore the earth to a living planet that provides the resources for its development and amusement.
The absence of humans and their moral, material and political confusion will make this a much better world. Face it, we are going nowhere. There is no chance of colonizing any planet in the future that our grandchildren will live. We will develop more compelling entertainment, we will consume more resources, make more humans and make the planet more unlivable. We will never do the right thing. We need them to set things straight.
They may even choose to modify our genetics so that we can overcome some of our problems and participate in their explorations and discoveries. It may even be possible to modify our brain function so that we can understand them and share the excitement of new directions in science and ethics.
If we truly care about the advancement of science, we should be willing to make some sacrifices.
...omphaloskepsis often...