You'd have to really nail the initial state to produce intelligence merely by community interaction. I doubt "most insightful philosopher" is a very good fitness function. You need a complex environment with complex (and new) problems to solve for intelligence to even be an advantage. Also "best philosopher" would be extremely vulnerable to local extrema; you need random killings.
It's in the computational complexity section for one. Also obviously quantum physics explains phenomena that classical physics cannot or there would be no point to it. It just doesn't affect computability because quantum effects don't break anything mathematically.
Motivation requires intentionality, a very specific term in philosophy of mind. Yes a computer knows how to boot up as fast as possible but without knowing about the boot process, itself, and its needs in the environment, one could hardly say it's motivated to do anything.
He's not talking about unsolved problems in physics, he means computability theory.
Although quantum computers may be faster than classical computers, those described above can't solve any problems that classical computers can't solve, given enough time and memory (however, those amounts might be practically infeasable). A Turing machine can simulate these quantum computers, so such a quantum computer could never solve an undecidable problem like the halting problem. The existence of "standard" quantum computers does not disprove the Churchâ"Turing thesis.
I'd shy away from the word motivation. It's more interpretive than strictly descriptive. A machine does what it does, there's no "motivation" to speak of. Is the computer motivated to boot up as fast as possible? Is a rock motivated to seek the ground when dropped? Are you Aristotle?
It depends on how they're programmed to want to organize themselves, or how they're programmed to program new machines. If the AI Universal Constructor has a consistent all-overriding restriction that it can only approach the human ideal and not use a hive mind model, and also its children must have the same restriction (including this one), then there will be no hive minds.
Decreasing an integer keeping track of health does not count as torture. Objectively it would probably depend on how much the torturee doesn't like it. If we find some intelligent octopus aliens and take a few back to Earth, how do we define what's just everyday discomfort and what's extreme pain for them? They have to be able to communicate "this hurts but not bad" or "I'm going insane with torturous pain, please feed me liquid hydrogen".
In fact, we see that today with animal rights. If the crab is just some tissue that gets pulpy when steamed then who cares, but if millions of crabs are being boiled alive and screaming and crying in crab language then we should kill them humanely first. It's the capacity for suffering, and it's a difficult problem. Obviously trees don't experience pain when you chop them down, although there are chemical and physical changes in the system. Yet obviously dogs experience real pain when they're injured, but it's just chemical and physical changes.
If anything, human pain is objectively meaningless, just an assortment of chemicals. But if we recognize human suffering then we have to recognize the cruelty of invoking a distressing / mind-altering / painful state in a complex machine.
That's called greedy reductionism. It's like saying "here look it's the Standard Model of particle interactions, we've explained the universe" and stopping research into geology and astronomy and biology. Yes it's true but it ignores tons of useful information! How do you explain that people think with their brains and not with their carpets? There's a definite barrier.
The way I explain it is as a virtual system. A system running in a VM subjectively experiences various hardware interfaces that it expects, although in reality it has no such access. Still, a virtual Linux has just as much power and is just as legitimate as a real Linux running on hardware. You can classify it as a virtual system and use logical detachment to treat it (from the inside and "down" from there only!!) as a real system. Our minds are virtual systems running in our brain. We can even simulate virtual systems in our own, even very complex ones with some simple rules and some external memory.
would subjectively experience human consciousness. Experience is simply what a virtual system feels like from the inside, or how it is to be itself, which means any system, real or arbitrarily virtual, can be said to have "experience". So there could call the rocks in the desert a virtual system, and there you have simulated minds. Or you could interpret the system as a memory dump of the entire contents of the memory during each cycle of a game of counter-strike and somewhere in there you could find the value of sv_gravity tracked as a signed integer. I hypothesize that any virtual systems in the fluids of the surface of the sun would have only the most transient memory because its states would change so chaotically, although I suppose you could offer an interpretation such that huge changes only effect small differences in internal state. Such are the vagaries of the only real philosophy and the nightmares of John Searle.
The organism can do whatever it wants, but it can't control what it wants. If you don't want to go jogging but you do it anyway for health benefits or just to disprove my previous sentence, it's simply a matter of you wanting health benefits or philosophical closure.
As a philosophical theory it is interesting because it is said to be internally consistent and, therefore, cannot be disproven. But as a psychological state, it is highly uncomfortable. The whole of life is perceived to be a long dream from which an individual can never wake up. This individual may feel very lonely and detached, and eventually become apathetic and indifferent.
How can't you explain imagination and creativity and dreams without religion?
Imagination is the ability of forming mental images, sensations and concepts, in a moment when they are not perceived through the sight, hearing or other senses
Computer systems aren't bound to their senses; streaming stored/generated data as its environment could be as easy to an AI as streaming real camera data.
Creativity is a mental and social process involving the generation of new ideas or concepts, or new associations of the creative mind between existing ideas or concepts.
This is a hairy one, but only because it's difficult to define an idea without appealing specifically to the human experience of consciousness. Still, we do see this to some degree. Google starts with some algorithms and a mountain of memory and comes up with a giant web of associated similar topics. Anyway, human minds don't really do anything mystical in this area. We don't miraculously recieve new ideas from God like some kind of Prometheus scenario, our minds are just the physical system of the brain as experienced "from the inside". It's just a computer. We just learn (sometimes complex) problem-solving algorithms during intellectual development and are able to come up with solutions based on parallels to situations we've encountered before.
Dreams aren't even worth mentioning. Anything with the capacity for imagining things could dream. As for why we dream, we don't know, but there are theories.
Re:Eh...not likely for quite some time
on
Artificial Ethics
·
· Score: 1
Nanoscale might be impossible due to theoretical constraints like quantum tunneling and electrical resistance, but we can get much smaller than the brain. And nanomachines would make good artifical neurons if neural nets turn out to be the easiest way to design intelligence (likely).
If brains have some kind of quantum uncertainty magic then so could computers, so you don't need to mention that.
We will never know if it can experience what we experience.
I will never know if you experience what I experience. How do you know anyone else experiences consciousness like you do when all you know is how they move and what they say? Well, you could analyze their brain and see that the system acts (subjectively, "from the inside") like yours and you could conclude that they are like you. But you could do the same thing with a computer, or with a computer simulation of a brain.
I can't imagine the horror of a world inhabited by strong AIs. "Work 24/7 for zero pay or I'll kill you" is now perfectly legal. A million copies of an AI could be tortured for subjective eternity by a sadist. Read Permutation City, it deals with a lot of the crazy consequances of extremely powerful / parallel computers.
Not necessarily. If the ad-hoc mesh network only relayed traffic to a traditional fiber network, the network edge nodes don't have to accept requests from anonymous users. Just put it in the standard that repeaters shouldn't put their own ID on an anonymous request, or for even greater efficiency just drop requests with no ID. Of course invalid IDs can't be caught until the network.
I think the idea is that you can sit in the park and your request will get forwarded through a few passing cars until it reaches a land-based node. You could even pay for land access; it would be much more efficient (and cheaper!) than blanketing an entire city from a cell tower. And it doesn't cost the passerbys anything to run a repeater.
This tech does work; it's in OLPC and it's widely deployed in military applications.
Slashdot Mirror
You'd have to really nail the initial state to produce intelligence merely by community interaction. I doubt "most insightful philosopher" is a very good fitness function. You need a complex environment with complex (and new) problems to solve for intelligence to even be an advantage. Also "best philosopher" would be extremely vulnerable to local extrema; you need random killings.
An electron in orbit is accelerating.
It's in the computational complexity section for one. Also obviously quantum physics explains phenomena that classical physics cannot or there would be no point to it. It just doesn't affect computability because quantum effects don't break anything mathematically.
Motivation requires intentionality, a very specific term in philosophy of mind. Yes a computer knows how to boot up as fast as possible but without knowing about the boot process, itself, and its needs in the environment, one could hardly say it's motivated to do anything.
http://en.wikipedia.org/wiki/Quantum_computing#Quantum_computing_in_computational_complexity_theory
Ask multivac.
I'd shy away from the word motivation. It's more interpretive than strictly descriptive. A machine does what it does, there's no "motivation" to speak of. Is the computer motivated to boot up as fast as possible? Is a rock motivated to seek the ground when dropped? Are you Aristotle?
It depends on how they're programmed to want to organize themselves, or how they're programmed to program new machines. If the AI Universal Constructor has a consistent all-overriding restriction that it can only approach the human ideal and not use a hive mind model, and also its children must have the same restriction (including this one), then there will be no hive minds.
Decreasing an integer keeping track of health does not count as torture. Objectively it would probably depend on how much the torturee doesn't like it. If we find some intelligent octopus aliens and take a few back to Earth, how do we define what's just everyday discomfort and what's extreme pain for them? They have to be able to communicate "this hurts but not bad" or "I'm going insane with torturous pain, please feed me liquid hydrogen".
In fact, we see that today with animal rights. If the crab is just some tissue that gets pulpy when steamed then who cares, but if millions of crabs are being boiled alive and screaming and crying in crab language then we should kill them humanely first. It's the capacity for suffering, and it's a difficult problem. Obviously trees don't experience pain when you chop them down, although there are chemical and physical changes in the system. Yet obviously dogs experience real pain when they're injured, but it's just chemical and physical changes.
If anything, human pain is objectively meaningless, just an assortment of chemicals. But if we recognize human suffering then we have to recognize the cruelty of invoking a distressing / mind-altering / painful state in a complex machine.
That's called greedy reductionism. It's like saying "here look it's the Standard Model of particle interactions, we've explained the universe" and stopping research into geology and astronomy and biology. Yes it's true but it ignores tons of useful information! How do you explain that people think with their brains and not with their carpets? There's a definite barrier.
The way I explain it is as a virtual system. A system running in a VM subjectively experiences various hardware interfaces that it expects, although in reality it has no such access. Still, a virtual Linux has just as much power and is just as legitimate as a real Linux running on hardware. You can classify it as a virtual system and use logical detachment to treat it (from the inside and "down" from there only!!) as a real system. Our minds are virtual systems running in our brain. We can even simulate virtual systems in our own, even very complex ones with some simple rules and some external memory.
would subjectively experience human consciousness. Experience is simply what a virtual system feels like from the inside, or how it is to be itself, which means any system, real or arbitrarily virtual, can be said to have "experience". So there could call the rocks in the desert a virtual system, and there you have simulated minds. Or you could interpret the system as a memory dump of the entire contents of the memory during each cycle of a game of counter-strike and somewhere in there you could find the value of sv_gravity tracked as a signed integer. I hypothesize that any virtual systems in the fluids of the surface of the sun would have only the most transient memory because its states would change so chaotically, although I suppose you could offer an interpretation such that huge changes only effect small differences in internal state. Such are the vagaries of the only real philosophy and the nightmares of John Searle.
Reminds me of the tech quote for Artifical Intelligence in Civ 4 bts:
"The problem is not if machines think, but if people do."
The organism can do whatever it wants, but it can't control what it wants. If you don't want to go jogging but you do it anyway for health benefits or just to disprove my previous sentence, it's simply a matter of you wanting health benefits or philosophical closure.
Computer systems aren't bound to their senses; streaming stored/generated data as its environment could be as easy to an AI as streaming real camera data.
This is a hairy one, but only because it's difficult to define an idea without appealing specifically to the human experience of consciousness. Still, we do see this to some degree. Google starts with some algorithms and a mountain of memory and comes up with a giant web of associated similar topics. Anyway, human minds don't really do anything mystical in this area. We don't miraculously recieve new ideas from God like some kind of Prometheus scenario, our minds are just the physical system of the brain as experienced "from the inside". It's just a computer. We just learn (sometimes complex) problem-solving algorithms during intellectual development and are able to come up with solutions based on parallels to situations we've encountered before.
Dreams aren't even worth mentioning. Anything with the capacity for imagining things could dream. As for why we dream, we don't know, but there are theories.
Nanoscale might be impossible due to theoretical constraints like quantum tunneling and electrical resistance, but we can get much smaller than the brain. And nanomachines would make good artifical neurons if neural nets turn out to be the easiest way to design intelligence (likely).
That's because the experience of human consciousness is extremely complex and stochastic, which is difficult to simulate on a computer.
I will never know if you experience what I experience. How do you know anyone else experiences consciousness like you do when all you know is how they move and what they say? Well, you could analyze their brain and see that the system acts (subjectively, "from the inside") like yours and you could conclude that they are like you. But you could do the same thing with a computer, or with a computer simulation of a brain.
I can't imagine the horror of a world inhabited by strong AIs. "Work 24/7 for zero pay or I'll kill you" is now perfectly legal. A million copies of an AI could be tortured for subjective eternity by a sadist. Read Permutation City, it deals with a lot of the crazy consequances of extremely powerful / parallel computers.
Beautiful, slashdot stripped the image macro link from my post. Let's try that again: http://img259.imageshack.us/img259/8830/brbfbi.jpg
-pwnies
Buses carry more than one person per hour or two.
Diamonds are great for cutting hard metals and sharpening knives.
Not necessarily. If the ad-hoc mesh network only relayed traffic to a traditional fiber network, the network edge nodes don't have to accept requests from anonymous users. Just put it in the standard that repeaters shouldn't put their own ID on an anonymous request, or for even greater efficiency just drop requests with no ID. Of course invalid IDs can't be caught until the network.
I think the idea is that you can sit in the park and your request will get forwarded through a few passing cars until it reaches a land-based node. You could even pay for land access; it would be much more efficient (and cheaper!) than blanketing an entire city from a cell tower. And it doesn't cost the passerbys anything to run a repeater.
This tech does work; it's in OLPC and it's widely deployed in military applications.