Try reading up on what Chernobyl did to Ukraine's neighbor Belarus (where most of the radiation came down, partly thanks to the Russians seeding rain clouds so it didn't make it as far as them).
About 1/3 of Belarus is contaminated. In an already poor country people can't pick wild mushrooms, berries etc in contaminated areas because of it.
The biggest suffers from this are young children - there are much increased rates of blood diseases like Leukemia in Belarus as a result of it.
I doubt there ever will be such a proof. In fact if you read Wolfram's "A New Kind of Science" you'll find just the opposite! Of course that's not to say that non-linear systems can't be modelled at all, but maybe we'll never be able to do better than characterize solutions (e.g. it lies on this strange attractor..).
However as far as the architecture of biological systems go, one would expect the incremental enhancement design technique of evolution to have found very stable designs where incremental changes have mostly incremental effects (i.e. linear dynamics)... and analytical modelling is therefore likely to be quite successful.
Well, I think you *could* correctly deduce the computer's experience of color, but of course there's a big difference in considering subjective experience analytically and actually having that subjective experience yourself... which of course is why it's hard to accept that our own subjective experience is really only the sum of the analyzable factors that we know it is!... If someone correctly and completely described your visual phenomenology then you'd still probably regard that as an empty description even if it captured everying that you experience.
IMO machine consciousness is a very similar issue - if it had the same feedback paths (and lack of them - conscious/subconscious access to various areas) in it's brain, same cognitive architecture, etc, then it'd have the same conscious experience that we do... From visual consciousness to somatic (body sense) consciousness... etc etc.
Well, how do I know that you have similar visual phenomenology to me? I only assume that because your perceptual and cognitive architecture is similar! (and questioning/testing of you would appear confirm that)
If you really thing that there's some inherent "redness" that an AI would be missing out on, then ask yourself what would be left of "red" if you took away it's relative position in the spectrum (i.e. what other colors it's similar/disimilar to), took away your emotional response to it, took away your mental associations (fluorescent, blood, sweater, balloon or whatever), etc, etc. I say nothing!
The only inherent attribute of color is that differently colored objects have different surface appearance (this is essentially the modern definition of color)... For another mental exercise try imagining an AI/robot which could detect color but somehow didn't have the experience of it that we do... But he'd still percieve differently colored objects as having different surface appearance, and it wouldn't be just shades of gray or somesuch as that wouldn't differentiate different colors of same lightness... The phenomenology comes directly from the comparative attributes...
I'm not claiming you can't mimic the behavioral response to color spaces, I'm claiming that the phenomenalogical experience may not be able to be duplicated in a given model.
If an AI were architected with a visual system (incl. visual cortex processing) the same as a human, then the phenomenology would be identical! He'd see green as halfway between blue and yellow just as you do, and it would look "grass green", "lime gree" etc to him just as it does to you!
If the AI were complete enough to have built the mental models of the world that we do then he'd also make the same blue=icey/cold, red=hot type associations/experience that we have, and colors would be wired into his emotions the same way as us so that red is arousing, green calming, etc.
You really want to read Stuart Kauffman's "At home in the universe" to understand the inevitability of the rise of complexity/life in the right circumstances.
Actually color perception is only loosely related to wavelength. What you were taught in highschool was an oversimplification that borders on a lie (as is much of high school science!). Color is really a spatial attribute, not a point one (Google for Edward Land's "Retinex" theory of color preception), and perception of color is not absolute - it depends on the spatially adjacent colors; this isn't an optical illusion - it's the nature of color perception. It doesn't even stop there because color is a compatative attribute - things look "leaf green" because they stimulate your visual cortex in the same say as a leaf, but that is still true if you wear red goggles, and experiments have shown that normal color vision returns after a couple of weeks of wearing colored goggles!
You should also note that humans can only see a fraction of the possible colors (combinations of wavelengths of light) even in the visual part of the spectrum), and there is therefore nothing absolute about what we perceive - it's just what we can differentiate. If instead of having 3 differently tuned color cones in our eye (the cones have bell-curve-like light wavelength response that peak around R/G/B) we had more, then we would be able to differentiate more wavelength combinations. With our eyes the way they are you can differentially stimulate our color cones with only three wavelengths of light, but if we had 4 (peak tuned to R/G/B/Yellow say, or ANY different wavelengths) then you would need 4. Some people in fact do have 4 types of color cones and can therefore differentiate colors that you cannot. Your "red" surface is someone else's patterned one!
That absolute "red" that you are worrying about therefore isn't an irreducible gestalt experience/quale - it's a differential surface attribute detection that a machine will be able to duplicate just fine.
Incidently note also that what you see a color as isn't going to be precisely what I see it as - we may agree on things like "green's a bit like blue and a bit like yellow" that are based on the underlying transducers and brain architecture, but what the color actually looks/feels like is going to be as personal as any other experiental phenomena.
Yep - that's what he's supposedly afraid of, but he could certainly plea bargain to get off anything they might want to charge him with, so it's really nothing more than just wanting to protect his murderer brother.
There's got to be some point at which you draw the line! Would I rat out a family member for something like stealing, or peddling drugs? No. For cold blooded murder (never mind 21 counts)? Yes.
What's really interesting about this isn't that the FBI is using banner ads, but rather why they have to...
The guy they're seeking, #10 on the most wanted list, and suspected of 21 murders, is the brother of the president of the University of Massachusetts, who just plead the 5th to keep his dear brother safe.
It's all about the zero-G. Fucking during a space-walk would be OK in a spacesuit made for two, but I'm more inclined to do it in the station... aside from the arial box munching possibilities you could also use bungees to strap the two of you to the ceiling, and use it as a substitute for gravity. I've got it all worked out!
Never mind the ride up, having a 1/6G 20lb girlfriend in a moonbase hotel would be interesting too!
Really! They should go totally commercial - build a moon station and fill it with $20M/week tourists. One the infrastructure was in place maybe they could up the volume, lower the price, and start ferrying the rest of us poor schmucks up there too.
That's like someone refuting relativity based on the predictions of newtonian physics... If he's onto somethign new here, then it's obvous that existing science isn't going to predict it.
The Tarkovsky version is available in Blockbuster too (at least I've seen it in two different stores - in foreign film section). A very good (albeit a tad wierd as the subject dictates) Tarkovsky movie is "Andrei Rublev" - the life story of the 15th C Russian icon painter.
Acorn itself was sold to Olivetti around 1984-5. ARM (now Advanced RISC Machines, but originally Acorn RISC Machine) was a spinoff, and Element-14 was a separate start-up involving some of the same people such as Sophie (formerly Roger) Wilson.
Re:Initial conditions don't really matter
on
Shapes of Time
·
· Score: 2
Wolfram's exhaustive analysis shows that the grand scale behaviour (most notably whether it generates complexity or not) of systems defined by feedback upon themselves is usually independent of initial conditions.. but not all are, nor do all generate complexity... What I meant by rules competing is that evolution isn't going to just be just a result of the phenotype's success in spreading it's DNA, but also of that DNA to mutate in beneficial ways in the future... For example one theory of junk DNA is that it provides spacing between genes so that typical DNA mutations don't disrupt genes... that's an evolution of the rules rather than the phenotype.
I have to wonder that too! The Doomsday machine was AFAICR nothing more exotic than a BBC Master with SCSI controller and a 12" laserdisc player. The player is hard to come by, but I have seen them on eBay.
Slashdot Mirror
That's a study primarily about fish.
Try reading up on what Chernobyl did to Ukraine's neighbor Belarus (where most of the radiation came down, partly thanks to the Russians seeding rain clouds so it didn't make it as far as them).
About 1/3 of Belarus is contaminated. In an already poor country people can't pick wild mushrooms, berries etc in contaminated areas because of it.
The biggest suffers from this are young children - there are much increased rates of blood diseases like Leukemia in Belarus as a result of it.
I doubt there ever will be such a proof. In fact if you read Wolfram's "A New Kind of Science" you'll find just the opposite! Of course that's not to say that non-linear systems can't be modelled at all, but maybe we'll never be able to do better than characterize solutions (e.g. it lies on this strange attractor..).
... and analytical modelling is therefore likely to be quite successful.
However as far as the architecture of biological systems go, one would expect the incremental enhancement design technique of evolution to have found very stable designs where incremental changes have mostly incremental effects (i.e. linear dynamics)
Well, I think you *could* correctly deduce the computer's experience of color, but of course there's a big difference in considering subjective experience analytically and actually having that subjective experience yourself... which of course is why it's hard to accept that our own subjective experience is really only the sum of the analyzable factors that we know it is! ... If someone correctly and completely described your visual phenomenology then you'd still probably regard that as an empty description even if it captured everying that you experience.
IMO machine consciousness is a very similar issue - if it had the same feedback paths (and lack of them - conscious/subconscious access to various areas) in it's brain, same cognitive architecture, etc, then it'd have the same conscious experience that we do... From visual consciousness to
somatic (body sense) consciousness... etc etc.
Well, how do I know that you have similar visual phenomenology to me? I only assume that because your perceptual and cognitive architecture is similar! (and questioning/testing of you would appear confirm that)
If you really thing that there's some inherent "redness" that an AI would be missing out on, then ask yourself what would be left of "red" if you took away it's relative position in the spectrum (i.e. what other colors it's similar/disimilar to), took away your emotional response to it, took away your mental associations (fluorescent, blood, sweater, balloon or whatever), etc, etc. I say nothing!
The only inherent attribute of color is that differently colored objects have different surface appearance (this is essentially the modern definition of color)... For another mental exercise try imagining an AI/robot which could detect color but somehow didn't have the experience of it that we do... But he'd still percieve differently colored objects as having different surface appearance, and it wouldn't be just shades of gray or somesuch as that wouldn't differentiate different colors of same lightness... The phenomenology comes directly from the comparative attributes...
I'm not claiming you can't mimic the behavioral response to color spaces, I'm claiming that the phenomenalogical experience may not be able to be duplicated in a given model.
If an AI were architected with a visual system (incl. visual cortex processing) the same as a human, then the phenomenology would be identical! He'd see green as halfway between blue and yellow just as you do, and it would look "grass green", "lime gree" etc to him just as it does to you!
If the AI were complete enough to have built the mental models of the world that we do then he'd also make the same blue=icey/cold, red=hot type associations/experience that we have, and colors would be wired into his emotions the same way as us so that red is arousing, green calming, etc.
You really want to read Stuart Kauffman's "At home in the universe" to understand the inevitability of the rise of complexity/life in the right circumstances.
Actually color perception is only loosely related to wavelength. What you were taught in highschool was an oversimplification that borders on a lie (as is much of high school science!). Color is really a spatial attribute, not a point one (Google for Edward Land's "Retinex" theory of color preception), and perception of color is not absolute - it depends on the spatially adjacent colors; this isn't an optical illusion - it's the nature of color perception. It doesn't even stop there because color is a compatative attribute - things look "leaf green" because they stimulate your visual cortex in the same say as a leaf, but that is still true if you wear red goggles, and experiments have shown that normal color vision returns after a couple of weeks of wearing colored goggles!
You should also note that humans can only see a fraction of the possible colors (combinations of wavelengths of light) even in the visual part of the spectrum), and there is therefore nothing absolute about what we perceive - it's just what we can differentiate. If instead of having 3 differently tuned color cones in our eye (the cones have bell-curve-like light wavelength response that peak around R/G/B) we had more, then we would be able to differentiate more wavelength combinations. With our eyes the way they are you can differentially stimulate our color cones with only three wavelengths of light, but if we had 4 (peak tuned to R/G/B/Yellow say, or ANY different wavelengths) then you would need 4. Some people in fact do have 4 types of color cones and can therefore differentiate colors that you cannot. Your "red" surface is someone else's patterned one!
That absolute "red" that you are worrying about therefore isn't an irreducible gestalt experience/quale - it's a differential surface attribute detection that a machine will be able to duplicate just fine.
Incidently note also that what you see a color as isn't going to be precisely what I see it as - we may agree on things like "green's a bit like blue and a bit like yellow" that are based on the underlying transducers and brain architecture, but what the color actually looks/feels like is going to be as personal as any other experiental phenomena.
Your anus opens YOU.
Don't be silly!
That's an old picture.
People and horses were much bigger back then.
Yep - that's what he's supposedly afraid of, but he could certainly plea bargain to get off anything they might want to charge him with, so it's really nothing more than just wanting to protect his murderer brother.
There's got to be some point at which you draw the line! Would I rat out a family member for something like stealing, or peddling drugs? No. For cold blooded murder (never mind 21 counts)? Yes.
Amazon files patent on reading books while taking a dump.
He's hoping to whack the guy at the deli counter?
What's really interesting about this isn't that the FBI is using banner ads, but rather why they have to...
The guy they're seeking, #10 on the most wanted list, and suspected of 21 murders, is the brother of the president of the University of Massachusetts, who just plead the 5th to keep his dear brother safe.
Bulgar takes the fifth
Great to see the head of an institute of learning take such a principled stand. Not.
It's all about the zero-G. Fucking during a space-walk would be OK in a spacesuit made for two, but I'm more inclined to do it in the station... aside from the arial box munching possibilities you could also use bungees to strap the two of you to the ceiling, and use it as a substitute for gravity. I've got it all worked out!
Never mind the ride up, having a 1/6G 20lb girlfriend in a moonbase hotel would be interesting too!
Just need the Russians to seize the opportunity.
Really! They should go totally commercial - build a moon station and fill it with $20M/week tourists. One the infrastructure was in place maybe they could up the volume, lower the price, and start ferrying the rest of us poor schmucks up there too.
One of my lifetime goals is to fuck in space.
Just too funny though - very well done.
Poor Hodges.
.. and run the Apple II original under an emulator.
AppleWin now supports mockingboard emulation.
That's like someone refuting relativity based on the predictions of newtonian physics... If he's onto somethign new here, then it's obvous that existing science isn't going to predict it.
The best ASCII representation to send would be "pi".
The Tarkovsky version is available in Blockbuster too (at least I've seen it in two different stores - in foreign film section). A very good (albeit a tad wierd as the subject dictates) Tarkovsky movie is "Andrei Rublev" - the life story of the 15th C Russian icon painter.
I can't believe you report this and don't even include the value of Pi he calculated in the article!
I guess I'll have to wait for one of the page widening trolls to post it.
Acorn itself was sold to Olivetti around 1984-5. ARM (now Advanced RISC Machines, but originally Acorn RISC Machine) was a spinoff, and Element-14 was a separate start-up involving some of the same people such as Sophie (formerly Roger) Wilson.
Wolfram's exhaustive analysis shows that the grand scale behaviour (most notably whether it generates complexity or not) of systems defined by feedback upon themselves is usually independent of initial conditions .. but not all are, nor do all generate complexity... What I meant by rules competing is that evolution isn't going to just be just a result of the phenotype's success in spreading it's DNA, but also of that DNA to mutate in beneficial ways in the future... For example one theory of junk DNA is that it provides spacing between genes so that typical DNA mutations don't disrupt genes ... that's an evolution of the rules rather than the phenotype.
I have to wonder that too! The Doomsday machine was AFAICR nothing more exotic than a BBC Master with SCSI controller and a 12" laserdisc player. The player is hard to come by, but I have seen them on eBay.