shows some sales figures from 2006. Of the youth-dominated genres we get something along the lines of: Alternative+Rock: 280Mil Metal:62Mil R&B+Rap:176Mil
There's no listing for "pop" and individual albums can be in multiple genres, so these are only very rough numbers. By this prediction, of course, we should see somewhere in the neighborhood of 12% of youth at large listening to metal. That's larger than 6%, but smaller than the 33% who listed it in their top-three genres. The only really suggestive discrepancy is between R&B having 18% in the survey and somewhere in the neighborhood of 34% sales. For which, clearly, there are many possible explanations having little to do with the intelligence of those who listen to R&B.
Of course, the only *real* way to test is to compare it against the answers to the same questions in the population at large. Perhaps the original study does so, but the article doesn't give enough to go on.
And of course, as others have pointed out, if Western governments weren't so agriculturally anticompetitive, African economies would be in much better shape.
"Bringing Yesterday's Masters to Today's Minds Springer expands the realm of scientific research through the Online Journals Archive package. Scientists and researchers can access over a century of scientific evolution and complete historical information. Springer has created a comprehensive body of scientific, medical and technical research documentation accessible to scientists, researchers and other professionals. We will offer approximately 1,200 journals starting in 2005 by adding all content formerly published by Kluwer, as well as, all content from Volume I, Issue 1, where available, from both the former Springer-Verlag and Kluwer."
Perhaps the expectation is that we'll go research it and then report back if there's actually something interesting going on somewhere. Somehow. Maybe even related in some way.
While I admit the virtualization of what was physically instantiated connection information is costly in terms of brute storage as well as memory bandwidth, there's not that many bits in the bus, once stripped of addressing overhead.
That said, managing memory bus speed could turn out to be a considerable technical constraint: the whole virtualizer would need a fairly agile daemon that tracked evolving connection topologies and kept neurons resident with their neighbors to minimize inter-cpu module bandwidth needs. This would be as transparent to the person those neurons instantiate as dendrite formation is to us.
It has been updated to account for IE7's ideosyncracies, but does it also cover the other browsers well? Every time I've started really using CSS, I 've gotten frustrated by the way each browser decided to support a different subset of CSS so I had to sit there and try everything before figuring out what non-IE-on-Windows browsers would make of the page. Probably the major browsers support much more of the CSS spec, but I was burned so bad last time I haven't wanted to touch it with a ten foot pole in several years.
To some extent, googlebombs *were* abuse, which leads me to think Google needed to upgrade their heuristics. This appears to be much of what they've done, though I think their response was too focused on killing that specific form of abuse and not focused enough on improving analytic depth.
I've never seen anything pernicious and accidental* come into a corporate network except through the marketing folks. They always seem to be the ones who like the use gadgets they don't understand, leave extraneous services on because they seem kinda neat and so on. They're exactly the sort of people who connect to ad-hocs all day long. After all, if their computer is compromised, it's IT's problem.
The summary is right - anyone who is a big enough geek to read/. isn't the sort of person the perpetrators are looking for.
*Pissed off IT guys have occasionally been foolish enough to actually sabotage their employers. This is pure shitting where you eat, no matter how big an asshole your CEO is (or whatever).
Problems like these will kill any technology in the cradle unless it has absolutely no competitors. If the 2006 models turn out to be anywhere near as bad, the game's already over. I mean, the Newton's second version vastly improved on the original and had no credible competitors, yet the horrible handwriting recognition on the first cast a pall over the whole enterprise that never lifted.
The X Prise had more to do with stoking an incipient avenue of development than anything as narrow as looking for an immediate solution. It shows that whatever it is can be done, or done better. There's publicity for the contestants, yes, but also for the contest. In cases where a company puts up the money, I'm sure that the prime functions are to create buzz for its industry (as well as the company's place in it) and as a method to identify hireable talent.
It would seem the next generation of bandwidth rollout predicates the true boom of pirating, considering that official services aren't likely to drop you and can usually keep a good data rate. P2P and the like are too unreliable for downloading a huge movie to be worth the hassle to most people.
Nothing (plausible) can detect everything that might be explosive, but it would seem this targets the most easily-obtained explosives and should raise the bar significantly in terms of the technical competence it takes to defeat the security measures.
That said, it would seem to have a ways to go before it's practical.
Having partial data from peoples' brains seems likely to be about as useful as 48 bits of a 128bit key - it might help in some way for those crunching the ciphertext, but on its own it's not going to make anything intelligible.
With machines, you can just replace damaged parts. Biological parts usually wear out even faster than machines, but as you point out there's mechanisms in place to repair the bioparts along the way. Think of modern manufacturing as being the repair process writ large and simplified. In that view, a number of 16th century windmills have far outlasted any of the biological machines that built them. Perhaps those are insufficiently complex to qualify in your mind, or maybe they are insufficiently non-stop in their operation. I would say they are nonetheless excellent candidates for comparison to, say a knee, which is a very complex device designed to accomplish a fairly simple task. Also, a knee in constant use without rest will break down in an awful hurry.
It would seem fairly easy to, when one's cybernetic arm is mechanically damaged, to take the spare out of your closet or whatever, flash the firmware with the virtual neural state of the damaged arm, and blammo, it feels exactly the same. Then you ship the damaged arm back to the manufacturer.
There are limits, of course. As long as the brain remains biological (and it seems likely to remain so for a long time) then one must have functioning metabolism feeding the brain, so full body replacements don't appear to be immediately in the offing. It might even be easier to "migrate" the mind from wetware to hardware than to work out ways to accommodate the incredibly convolute requirements of the various metabolic systems. "Easy" in the sense of being plausibly tractable by the end of the century.
It's all well and good to "biopatch" if you will, but machine-biological interfaces are really the holy grail. Machine engineering is far easier than biological engineering, more replaceable, more durable, and eventually more versatile. If your arm is amputated, we can either restore some basic functionality with a neural extension cord, or we can put a big fat processor connected to precise abiological sensors on it to provide all the proprioceptive and tactile data the original arm would have supplied. The only problem is presentation of that data to our biological brains. For that we'd still need some sort of electrode grid or something. Not an easy problem, but at least if it's solved once it's more or less solved for all time. Trying to regrow biological parts involves a gajillion types of tissue and membranes and so on in bewildering variety. Nature did not design us for easy reverse engineering.
In any case, biopatching is great and tractible for reconnecting pieces that already fundamentally work, but for wholesale replacement at a high grade of function we still need that bridge.
If we model each neuron as an object composed of state equations, quasi-spatial information and in-out buffers and synapse sets as registry entries that link a given out buffer to the in buffer of another neuron, then we can expect that we'll have a total memory overhead of about 50KB per neuron (when accounting for the average number of connections per neuron) requiring about 50KFLOPS per neuron. For 2x10^8 neurons, that implies that we can comfortably allow for human-scale intelligence using one PByte of memory and about one PFLOP. Of course, there are some subsystems in the brain we may not need to instantiate to such a high degree of verisimilitude and can be replaced with more traditionally coded software (I'm thinking here of more hard-coded lizard brain stuff).
All of this is out of reach today, but in 20 years, I doubt it will seem insurmountable in the slightest. The truly hard question is how to set up an initial arrangement of those neurons (with all the various kinds, initial connections and so on) that allows for a person to grow inside. You know, like any baby.
I didn't mean to imply that only Strong AI is militarily useful. In fact, I would say that Strong AI is *not* useful, if one thinks about the ethics of forcing anything sentient to go to war in one's place.
Also, I have no trouble recognizing that cleverly-designed "Weak" AI is nonetheless quite strong enough in more conventional senses to be a monumental aid to human problem solving, in the same manner and to the same degree as an ICBM is a great aid to human offensive capabilities.
I don't mean to imply humanoid robotics qua robotics is necessary to AI development. Rather, only in a creature that acts as an agent inhabiting the world at large can one expect anything like human-level understanding thereof to develop. It's all very well to develop clever as-if software widgets to simulate understanding in carefully controlled circumstances, but they won't scale to true global context richness because 1) they interact with the world over narrow modalities and 2) they don't have the rich internal structure necessary on which predicate agents with deep and flexible competencies.
It's like we build ever more elaborate visual perception analogues, but they backend into databases that only ask for enumerations of objects discriminated. I don't care how competent the visual system is, it's never going to achieve sentience because it's not part of a whole agent that travels around (in some sense), processes the answers it's getting from the visual system in a multimodal way related to the agent's goals, edits those goals based on new information and so on. It can't just see, it has to look, and it can't just look because someone typed in a domain name, it has to look for a reason and the reason has to be a reason in the sense of being the result of a decision or discrimination, not just an action with a physical cause.
It would seem that the easiest way to allow for all that is to build something that really moves around in the real world. In short, building a robot with all the appropriate competencies might be really hard, but it's still the most tractable way to achieve Strong AI.
The only programming that leads to context-rich understanding that could be called "knowledge" in the human sense is self-programming. Like babies. We're all born with a some basic software and a lot of hardware, but it's interaction over time with our environment that we self-program. One might call it learning, but it's more fundamental than just accumulating facts: it's self-creation.
Dennett calls us self-created selves. Any AI more than superficially like a human would be the same.
The DoD funds a huge percentage of AI research, but at the end of the day they're interested in things that can be easily weaponized or used for simple intelligence sifting heuristics. The most fundamentally interesting research in AI is in the humanoid robotics projects such as those at the MIT shop, and it is from these more humanly-modeled projects that anything like HAL could ever issue. Search-digest heuristics like PAL aren't much like humans and will never lead to anything approching a human's contextually rich understanding of the world at large any more than really advanced racecar design will lead to interstellar craft.
The difference, as Searle would say, between Strong (humanlike) AI and Weak (software widget like) AI is a difference of type, not scale.
I think we agree that there are generally two aspects to an item's environmental impact: production and use. Both panels and big screen TVs have negative production values, but panels thereafter have a positive or at least approximately neutral value* while TVs have a clearly negative one.
If we assume that panels and TVs have the same environmental production cost, then clearly spending money on the panels would be better. But even better than that might be just throwing your money in a hole - then you don't incur the production cost at all! Is that the environmentally-responsible course of action to which we should try to guide consumers?
Clearly this argument is a charicature, but I wanted the narrative to illustrate the extreme version of the consumer spending displacement argument. Obviously you want something more along the lines of people investing their money in environmental solutions rather than spending it on ephemeral pleasures. The problem, however, is that many environmentalists don't feel it useful to incorporate economic analysis into their decision making regarding in what sorts of environmental solutions they wish to invest.
Sometimes this makes me suspicious of the committment of certain self-described "environmental activists" to real solutions as opposed to, say, punishing greed or advancing some pet vision. I don't know how many times I've pointed out to some would-be savior of the earth living out in suburbia (or worse, some forest village) that for all their recycling and power conservation and all that if they were willing to live in urban high-density housing that their environmental footprint would be automatically cut up to 90%, but then the person says they can't bear to live in the soulless concrete wasteland or whatever. That's insincertity to me. I spend ~$400 extra a month to situate myself amenably to mass transit, living in a high density area generating very little waste and blah blah blah because I do care about the environment and I think it's a shame the way we subsidize** destructive styles of living (suburban, rural, road-commuter).
Costs exist for a reason and they signal something. Sometimes they signal government bias (tax structures, regulation and infrastructure projects massively favore low-density living) and sometimes they just signal for resources expended by the activity. In the case of solar panels, the government actually lowers their cost through subsidies and tax breaks, but these arguably only offset the various government advantages afforded to traditional power generation. The remaining high cost - the one we actually see - somewhat represents the capital costs of establishing manufacturing capability for solar panels plus the marginal cost of production, but is also significantly representative of research and development costs. All those researchers have to be educated, housed and fed. Further, devotion of research resources to the solar panel power field to some degree those resources for non-panel solar power, wind power, geothermal, nuclear, and so on, as well as for high-efficiency electronics, environmental-control smart materials and so on.
If I felt that in the future solar panels would prove to be a better solution to the big problems, I would probably have something else to say, but this to me appears to be a rich man conducting a distracting stunt, the message of which is: "If you're wealthy enough, you con't have to change your lifestyle or assumptions to maintain the pretense of environmental consciousness."
Yeah, I'm a little bitter, but I think a lot of these environmentalists' distain for taking economic analysis seriously represents a real threat to broader success.
Money represents resources, which ultimately have an environmental cost, whether through feeding more people to provide labor, mining more for metals, clearing land for roads... There are, of course, ways of decreasing the impact of any input on the environment - solar power generation is, ceteris paribus, one of them. However, the ceteris is not paribus, if you will. Though the array and its successors might compare well in use to a fossil fuel plant, their manufacture is no more environmentally neutral than nuclear and further their manufacturing expense means that there's a bunch of resources that could be allocated to producing other things sinking into solar panel manufacturing.
Really solar panels are great for power generation in places where transmission from centralized or semi-centralized is very resource intensive. Solar plants are (closest to) competitive for new generation in isolated communities or to serve communities that have slightly outgrown their existing power generation capabilities. Outside of that, thinking of solar power as a holy grail is just going to lead to a lot of waste and not a lot of environment saving.
Slashdot Mirror
http://rogerbourland.com/blog/2007/01/07/2006-genr e-album-sales-report/
shows some sales figures from 2006. Of the youth-dominated genres we get something along the lines of:
Alternative+Rock: 280Mil
Metal:62Mil
R&B+Rap:176Mil
There's no listing for "pop" and individual albums can be in multiple genres, so these are only very rough numbers. By this prediction, of course, we should see somewhere in the neighborhood of 12% of youth at large listening to metal. That's larger than 6%, but smaller than the 33% who listed it in their top-three genres. The only really suggestive discrepancy is between R&B having 18% in the survey and somewhere in the neighborhood of 34% sales. For which, clearly, there are many possible explanations having little to do with the intelligence of those who listen to R&B.
Of course, the only *real* way to test is to compare it against the answers to the same questions in the population at large. Perhaps the original study does so, but the article doesn't give enough to go on.
"Alternately, if your kids likes rock, they are 6.6 times more likely to be a genius"
I'm sure you have no difficulty figuring out how this statement is wrong.
Who makes Africa buy seed from the West?
And of course, as others have pointed out, if Western governments weren't so agriculturally anticompetitive, African economies would be in much better shape.
Had no idea about Verlag being a German word, obviously.
SpringerOnline Journals Archives
n ger/00244/index.asp
from http://www.environmental-expert.com/magazine/spri
"Bringing Yesterday's Masters to Today's Minds
Springer expands the realm of scientific research through the Online Journals Archive package. Scientists and researchers can access over a century of scientific evolution and complete historical information. Springer has created a comprehensive body of scientific, medical and technical research documentation accessible to scientists, researchers and other professionals. We will offer approximately 1,200 journals starting in 2005 by adding all content formerly published by Kluwer, as well as, all content from Volume I, Issue 1, where available, from both the former Springer-Verlag and Kluwer."
Perhaps the expectation is that we'll go research it and then report back if there's actually something interesting going on somewhere. Somehow. Maybe even related in some way.
But expectations are sometimes disappointed.
While I admit the virtualization of what was physically instantiated connection information is costly in terms of brute storage as well as memory bandwidth, there's not that many bits in the bus, once stripped of addressing overhead.
That said, managing memory bus speed could turn out to be a considerable technical constraint: the whole virtualizer would need a fairly agile daemon that tracked evolving connection topologies and kept neurons resident with their neighbors to minimize inter-cpu module bandwidth needs. This would be as transparent to the person those neurons instantiate as dendrite formation is to us.
It has been updated to account for IE7's ideosyncracies, but does it also cover the other browsers well? Every time I've started really using CSS, I 've gotten frustrated by the way each browser decided to support a different subset of CSS so I had to sit there and try everything before figuring out what non-IE-on-Windows browsers would make of the page. Probably the major browsers support much more of the CSS spec, but I was burned so bad last time I haven't wanted to touch it with a ten foot pole in several years.
To some extent, googlebombs *were* abuse, which leads me to think Google needed to upgrade their heuristics. This appears to be much of what they've done, though I think their response was too focused on killing that specific form of abuse and not focused enough on improving analytic depth.
I've never seen anything pernicious and accidental* come into a corporate network except through the marketing folks. They always seem to be the ones who like the use gadgets they don't understand, leave extraneous services on because they seem kinda neat and so on. They're exactly the sort of people who connect to ad-hocs all day long. After all, if their computer is compromised, it's IT's problem.
/. isn't the sort of person the perpetrators are looking for.
The summary is right - anyone who is a big enough geek to read
*Pissed off IT guys have occasionally been foolish enough to actually sabotage their employers. This is pure shitting where you eat, no matter how big an asshole your CEO is (or whatever).
Problems like these will kill any technology in the cradle unless it has absolutely no competitors. If the 2006 models turn out to be anywhere near as bad, the game's already over. I mean, the Newton's second version vastly improved on the original and had no credible competitors, yet the horrible handwriting recognition on the first cast a pall over the whole enterprise that never lifted.
The X Prise had more to do with stoking an incipient avenue of development than anything as narrow as looking for an immediate solution. It shows that whatever it is can be done, or done better. There's publicity for the contestants, yes, but also for the contest. In cases where a company puts up the money, I'm sure that the prime functions are to create buzz for its industry (as well as the company's place in it) and as a method to identify hireable talent.
It would seem the next generation of bandwidth rollout predicates the true boom of pirating, considering that official services aren't likely to drop you and can usually keep a good data rate. P2P and the like are too unreliable for downloading a huge movie to be worth the hassle to most people.
Nothing (plausible) can detect everything that might be explosive, but it would seem this targets the most easily-obtained explosives and should raise the bar significantly in terms of the technical competence it takes to defeat the security measures.
That said, it would seem to have a ways to go before it's practical.
Having partial data from peoples' brains seems likely to be about as useful as 48 bits of a 128bit key - it might help in some way for those crunching the ciphertext, but on its own it's not going to make anything intelligible.
With machines, you can just replace damaged parts. Biological parts usually wear out even faster than machines, but as you point out there's mechanisms in place to repair the bioparts along the way. Think of modern manufacturing as being the repair process writ large and simplified. In that view, a number of 16th century windmills have far outlasted any of the biological machines that built them. Perhaps those are insufficiently complex to qualify in your mind, or maybe they are insufficiently non-stop in their operation. I would say they are nonetheless excellent candidates for comparison to, say a knee, which is a very complex device designed to accomplish a fairly simple task. Also, a knee in constant use without rest will break down in an awful hurry.
It would seem fairly easy to, when one's cybernetic arm is mechanically damaged, to take the spare out of your closet or whatever, flash the firmware with the virtual neural state of the damaged arm, and blammo, it feels exactly the same. Then you ship the damaged arm back to the manufacturer.
There are limits, of course. As long as the brain remains biological (and it seems likely to remain so for a long time) then one must have functioning metabolism feeding the brain, so full body replacements don't appear to be immediately in the offing. It might even be easier to "migrate" the mind from wetware to hardware than to work out ways to accommodate the incredibly convolute requirements of the various metabolic systems. "Easy" in the sense of being plausibly tractable by the end of the century.
It's all well and good to "biopatch" if you will, but machine-biological interfaces are really the holy grail. Machine engineering is far easier than biological engineering, more replaceable, more durable, and eventually more versatile. If your arm is amputated, we can either restore some basic functionality with a neural extension cord, or we can put a big fat processor connected to precise abiological sensors on it to provide all the proprioceptive and tactile data the original arm would have supplied. The only problem is presentation of that data to our biological brains. For that we'd still need some sort of electrode grid or something. Not an easy problem, but at least if it's solved once it's more or less solved for all time. Trying to regrow biological parts involves a gajillion types of tissue and membranes and so on in bewildering variety. Nature did not design us for easy reverse engineering.
In any case, biopatching is great and tractible for reconnecting pieces that already fundamentally work, but for wholesale replacement at a high grade of function we still need that bridge.
If we model each neuron as an object composed of state equations, quasi-spatial information and in-out buffers and synapse sets as registry entries that link a given out buffer to the in buffer of another neuron, then we can expect that we'll have a total memory overhead of about 50KB per neuron (when accounting for the average number of connections per neuron) requiring about 50KFLOPS per neuron. For 2x10^8 neurons, that implies that we can comfortably allow for human-scale intelligence using one PByte of memory and about one PFLOP. Of course, there are some subsystems in the brain we may not need to instantiate to such a high degree of verisimilitude and can be replaced with more traditionally coded software (I'm thinking here of more hard-coded lizard brain stuff).
All of this is out of reach today, but in 20 years, I doubt it will seem insurmountable in the slightest. The truly hard question is how to set up an initial arrangement of those neurons (with all the various kinds, initial connections and so on) that allows for a person to grow inside. You know, like any baby.
I didn't mean to imply that only Strong AI is militarily useful. In fact, I would say that Strong AI is *not* useful, if one thinks about the ethics of forcing anything sentient to go to war in one's place.
Also, I have no trouble recognizing that cleverly-designed "Weak" AI is nonetheless quite strong enough in more conventional senses to be a monumental aid to human problem solving, in the same manner and to the same degree as an ICBM is a great aid to human offensive capabilities.
I don't mean to imply humanoid robotics qua robotics is necessary to AI development. Rather, only in a creature that acts as an agent inhabiting the world at large can one expect anything like human-level understanding thereof to develop. It's all very well to develop clever as-if software widgets to simulate understanding in carefully controlled circumstances, but they won't scale to true global context richness because 1) they interact with the world over narrow modalities and 2) they don't have the rich internal structure necessary on which predicate agents with deep and flexible competencies.
It's like we build ever more elaborate visual perception analogues, but they backend into databases that only ask for enumerations of objects discriminated. I don't care how competent the visual system is, it's never going to achieve sentience because it's not part of a whole agent that travels around (in some sense), processes the answers it's getting from the visual system in a multimodal way related to the agent's goals, edits those goals based on new information and so on. It can't just see, it has to look, and it can't just look because someone typed in a domain name, it has to look for a reason and the reason has to be a reason in the sense of being the result of a decision or discrimination, not just an action with a physical cause.
It would seem that the easiest way to allow for all that is to build something that really moves around in the real world. In short, building a robot with all the appropriate competencies might be really hard, but it's still the most tractable way to achieve Strong AI.
The only programming that leads to context-rich understanding that could be called "knowledge" in the human sense is self-programming. Like babies. We're all born with a some basic software and a lot of hardware, but it's interaction over time with our environment that we self-program. One might call it learning, but it's more fundamental than just accumulating facts: it's self-creation.
Dennett calls us self-created selves. Any AI more than superficially like a human would be the same.
The DoD funds a huge percentage of AI research, but at the end of the day they're interested in things that can be easily weaponized or used for simple intelligence sifting heuristics. The most fundamentally interesting research in AI is in the humanoid robotics projects such as those at the MIT shop, and it is from these more humanly-modeled projects that anything like HAL could ever issue. Search-digest heuristics like PAL aren't much like humans and will never lead to anything approching a human's contextually rich understanding of the world at large any more than really advanced racecar design will lead to interstellar craft.
The difference, as Searle would say, between Strong (humanlike) AI and Weak (software widget like) AI is a difference of type, not scale.
Replied to the wrong parent
I think we agree that there are generally two aspects to an item's environmental impact: production and use. Both panels and big screen TVs have negative production values, but panels thereafter have a positive or at least approximately neutral value* while TVs have a clearly negative one.
If we assume that panels and TVs have the same environmental production cost, then clearly spending money on the panels would be better. But even better than that might be just throwing your money in a hole - then you don't incur the production cost at all! Is that the environmentally-responsible course of action to which we should try to guide consumers?
Clearly this argument is a charicature, but I wanted the narrative to illustrate the extreme version of the consumer spending displacement argument. Obviously you want something more along the lines of people investing their money in environmental solutions rather than spending it on ephemeral pleasures. The problem, however, is that many environmentalists don't feel it useful to incorporate economic analysis into their decision making regarding in what sorts of environmental solutions they wish to invest.
Sometimes this makes me suspicious of the committment of certain self-described "environmental activists" to real solutions as opposed to, say, punishing greed or advancing some pet vision. I don't know how many times I've pointed out to some would-be savior of the earth living out in suburbia (or worse, some forest village) that for all their recycling and power conservation and all that if they were willing to live in urban high-density housing that their environmental footprint would be automatically cut up to 90%, but then the person says they can't bear to live in the soulless concrete wasteland or whatever. That's insincertity to me. I spend ~$400 extra a month to situate myself amenably to mass transit, living in a high density area generating very little waste and blah blah blah because I do care about the environment and I think it's a shame the way we subsidize** destructive styles of living (suburban, rural, road-commuter).
Costs exist for a reason and they signal something. Sometimes they signal government bias (tax structures, regulation and infrastructure projects massively favore low-density living) and sometimes they just signal for resources expended by the activity. In the case of solar panels, the government actually lowers their cost through subsidies and tax breaks, but these arguably only offset the various government advantages afforded to traditional power generation. The remaining high cost - the one we actually see - somewhat represents the capital costs of establishing manufacturing capability for solar panels plus the marginal cost of production, but is also significantly representative of research and development costs. All those researchers have to be educated, housed and fed. Further, devotion of research resources to the solar panel power field to some degree those resources for non-panel solar power, wind power, geothermal, nuclear, and so on, as well as for high-efficiency electronics, environmental-control smart materials and so on.
If I felt that in the future solar panels would prove to be a better solution to the big problems, I would probably have something else to say, but this to me appears to be a rich man conducting a distracting stunt, the message of which is: "If you're wealthy enough, you con't have to change your lifestyle or assumptions to maintain the pretense of environmental consciousness."
Yeah, I'm a little bitter, but I think a lot of these environmentalists' distain for taking economic analysis seriously represents a real threat to broader success.
Money represents resources, which ultimately have an environmental cost, whether through feeding more people to provide labor, mining more for metals, clearing land for roads... There are, of course, ways of decreasing the impact of any input on the environment - solar power generation is, ceteris paribus, one of them. However, the ceteris is not paribus, if you will. Though the array and its successors might compare well in use to a fossil fuel plant, their manufacture is no more environmentally neutral than nuclear and further their manufacturing expense means that there's a bunch of resources that could be allocated to producing other things sinking into solar panel manufacturing.
Really solar panels are great for power generation in places where transmission from centralized or semi-centralized is very resource intensive. Solar plants are (closest to) competitive for new generation in isolated communities or to serve communities that have slightly outgrown their existing power generation capabilities. Outside of that, thinking of solar power as a holy grail is just going to lead to a lot of waste and not a lot of environment saving.