He's also overlooking the factor of time. Evolution is only guaranteed to converge on the global optimum over an infinite time, and, in unsually perverse parameter spaces, it can certainly take an amount of time wholly unavailable to any reasonable application, so that you begin to think the algorithm will "never" converge to the global optimum.
I think this is one reason people find biological evolution fairly easy to swallow when it applies to microbes becoming more drug-resistant, but harder to credit when it applies to dinosaurs evolving wings and turning into birds. It's just a lot harder to wrap your head around the millions of generations such a complex optimization takes, and visualize the exceedingly unlikely semi-random walk through parameter space that happened. We start to rear back and say no way, that's just so damn unlikely it could never happen.
But our daily experience -- our common sense -- does not train us well to distinguish between could never happen and couldn't happen in less than ten million years, and, of course, over a billion years or so the latter is actually functionally equivalent to guaranteed to happen.
Oh let's think positive. Maybe one of the problems in AI is that the part of our thinking of which we're most aware -- the conscious, logical, linear, step-by-step part after which we model our computer programs -- is the part that does the least of our daily-life thinking. Maybe the bulk of our daily-life thinking is half-unconscious and massively parallel, to suit our hardware (10^12 neurons with a 1 kHz clock cycle).
Indeed, I find it fairly plausible that the bulk of our normal "human" ability to usefully and sanely respond to each other might have less to do with peering many steps ahead down complex decision trees and more to do with being able to recognize a huge number of variations on certain basic patterns and situations, and doing so on the basis of incomplete and noisy data. In other words, what makes humans good at, well, being human is more like recognizing a familiar face among thousands you've never seen or solving Where's Waldo? puzzles than playing three-dimensional chess with the M-5.
Alas, the bulk of the computing community has developed fairly little expertise in trying to imagine how to build, program and debug that kind of hardware/software combination.
So maybe that's all the problem is. We might just need to get used to building 500 nm wide nano-CPUs running pico-kernels, and figure out how to wire up a few billion to accomplish pattern recognition tasks as fast as we do.
Just taking a guess here, but I'd say they'll only consider lossless compression schemes (no point in throwing away data it took $400 million to collect), and that photos of Mars are not boring enough (e.g. with vast seas of one-color pixels) to be very compressable via lossless algorithms.
Let us not forget that as the black hat prey evolve the white hat predators do, too, and the latter usually have more funding.
So, if we imagine those stealthy microcrimsubs creeping up to a lonely San Diego beach, we might as well also imagine them bumping their little snouts 100 yards offshore into the dog pod grid. Bzzt snap glub glub glub. Problem solved!
Movies to the contrary, I sort kinda suspect the bulk of criminal smuggling is low tech. You just pay a dumfuk mule a G to take this "package of condoms" across the border, saying it's hers, and then give it to Hernando on the other side, who'll give her another 2 G's. If she gets busted, you're out less than the cost of a minicrimsub, you don't tip off the G-men to your methodology, and there are fewer material clues pointing back to your secret base under the crater of the extinct volcano.
Now that seems like a lot of data. Let's see, I read somewhere it's going to image 1% of the surface of Mars. Mars has a surface area of [Googles radius, punches calculator] 1.4e+14 m^2, so 1% of that is 1.4e+12 m^2. If the smallest thing you can resolve is a coffee table, and that's about 1 m^2, then that suggests each pixel is 1 m^2, so we have 1.4e+12 pixels coming back. Full color, natch, so no less than 32 bpp, totaling 5.8 terabytes.
That's a lot of data. If it has to get back here in a year or so, that's more than 1.4 megabit/second through your deep-space radio modem, even if you transmit around the clock all year.
I conclude either it's going to take substantially more than a year, or they've actually got a deep-space radio T1.
You know, it's also possible that, if implemented right (I know, fat chance, but still...) that RFID identification could increase privacy and security.
Consider: one of the big problems of modern life is you have to prove your ID, credit and legitimacy to all kinds of people all over the place, with the consequences that sooner or later your private info leaks out and bad guys can get ahold of it, zap, identity theft, credit card fraud, and so forth.
But what if something like an RFID tag could be provided by one tightly controlled source, and it gave unimpeachable evidence of right to be there on the airplane, credit enough to buy the laptop from Best Buy, whatever.
Then imagine: you walk into the Best Buy with your bankcard with the RFID tag. You pick up the laptop and walk out. Now, the BB security can let you out, because the bank's card has told them you've credit enough to buy the laptop and given them some secret code that guarantees them payment, and when they get that payment they have to give your digital assistant a secret code that guarantees you can get warranty repairs. But your card hasn't told them a damn thing else about you. You haven't had to tell BB your address or e-mail or bank name or even your own name. No junk mail from Best Buy, no tracking your purchases, no poking their nose into your credit history...
Same thing with the airplane. You get an airline ticket after having proved who you are and where you can be found, and in some way -- OK, things get a little fuzzy here, but bear with me -- that you're safe and can be trusted, and then you walk into the airport and onto the airplane. The chip says "legit, allowed on plane to Boston at 6.47" but nothing else. No one checks your driver's license or passport eight zillion times, no one bends you over to search for bombs up your...but I digress...
Anyway, one of the reasons we have to send out all this extra, unnecessary, privacy-endangering information about ourselves is because we don't have one rock-solid unforgeable way of identifying those narrow aspects of ourselves (our citizenship, credit, student status, license to drive, et cetera) that are legitimate necessities of certain transactions. Maybe RFIDs with some digital signature technology could provide one? Maybe the future could be more private and secure?
Or have I had too much caffeine already? [Peers anxiously into empty mug...]
But you overlook the possibility that the true Buddha wishes to travel incognito, so that (among other things) he need not be bothered by the shallow conversation of false disciples overly impressed by such ephemeral and wordly measures of wisdom, and can keep his select circle of the Elect quietly anonymous...listening...watching...making the occasional, small, barely noticeable Adjustment...
OK, this one befuddles me, so I throw it out hoping a geekier geek will know the answer:
Why hasn't there been a significant move towards digital signatures a la PGP signatures on e-mail? With legit sigs registered with and guaranteed by their firms (work mail) or ISPs (personal), so everyone can find a spammer when he spams and....uh...do something firm but completely legal and moderate to them...nothing at all like staking them on an anthill or trying out setting #10 on The Machine, really.
Anyway, doesn't it seem easier for people to get used to the idea that senders should validate themselves up front, rather than leaving it up to receivers to sift out valid senders -- and with no better route than trying to analyze natural language, for God's sake, a major AI/expert-system challenge itself.
Jesus, tell me about it. I get 30kb attachments merely saying "Got your email, thanks!" with "thanks" done up in some odd curly red font and a six-line sig, not to mention the twenty-seven 8x10 colored glossy JPG attachments with circles and arrows and a paragraph on the back of each one...
I am hard pressed to believe that computers are especially useful in learning, except in two areas, online searching and programming.
Much of what students seem to do with a computer sometimes seems like so much eye candy. You word process your paper instead of type it, put it in proportional Verdana font instead of mono, put bullets and footnotes and italics in with gay abandon. Or you make nice drawings with a vector drawing program instead of by hand. Or, as one poster said elsewhere, you type copious notes and then do searches on them later, instead of (say) thinking hard about what you're hearing as you hear it, and taking minimal but well-organized notes, which worked for generations before this.
But the availability and common use of online information is completely new to this generation, I think. In decades previous we'd truck on down to the library and read books. That meant we rarely got the latest info, and often our info was more restricted -- we didn't find much stuff representing fringe viewpoints, or the viewpoints of people from quite different cultures -- and also, of course, we tended not to find much trash and outright garbage, because it had all been through a lot of review and editing. We didn't get much in the way of unfiltered raw mass opinion.
Now we do. And it seems to me a paradigm shift has occured, and we always will. So exposing kids to the new way we get information -- fast, unfiltered, copious, spasmodically -- and teaching them how to judge it, winnow it, use it responsibly and successfully -- these are very important things we should be doing, and they do need a computer.
I also wish students used those computers to learn programming, because programming teaches the key intellectual skill of debugging your ideas. Too often we think that because an idea is beautiful and successful and perfect in our imagination it will be so in the real world. Programming your ideas teaches very well that, alas, it doesn't work that way, and teaches you a healthy respect for the difference between theory and reality.
Well, here is a list of the world's top 14 oil producers, which are, in order: Saudi Arabia (10.4 bbl/day), Russia, the US, Iran, Mexico, China, Norway, Canada, Venezuela, the UAE, Kuwait, Nigeria, the UK, and Iraq (2.03 bbl/day). Covering the First World (Norway) to the Third (Nigeria), Western democracy (Canada) to Islamic theocracy (Iran) to the Worker's Paradise (China), and five of the seven continents, that seems pretty widely distributed to me!
I realize the Saudis are a pain, and Venezuela has that nutcase Chavez at the reins (for now), but we get along OK with Russia and Mexico and Norway et cetera. It seems a little hard to credit that our relationship with all of the major oil producers is going to go belly up at once.
What you've also got to remember is that economic dependency is a two-way street. It would certainly hurt the US if, say, the Saudis stopped selling oil to us -- but what do you think it would do to the Saudis? What else have they got to sell but oil? How are they going to keep their economy going?
We're talking about oil prices doubling soon, and tripling and quadrupling after that, on to a price point of infinity because there WILL BE NO MORE OIL...
I don't think so. You are assuming that consumption rates will not change with price, but that is silly. Raise the price, and people start to economize and find alternatives. Raise the price a lot, and people flock to the alternatives and consumption plummets. And as I said, there are many alternative sources of energy. They're not used as much as oil now only because oil is cheap. If oil becomes expensive, they will be.
In a decade at most the American military will be taking oil from around the world at gunpoint.
This is a total red herring, but I can't resist saying (just to pull your chain) that this would be fine with me. Life is struggle for survival, my friend. Get used to it or join the dinosaurs.
NO. You won't have a job, and neither will I.
Speak for yourself. I believe I've got the skills that would let me find a niche in any society at all. If the US collapses into the equivalent of 14th century Spain, look for me to be one of those bastards with an ostrich feather in my enormous hat, whose dozens of beefy retainers will proceed in front of me at a respectful distance to thrust commoners into the gutter and out of my way. Bwa ha ha ha!
Well, I might not put it so harshly, but I believe I largely agree with you.
I think part of the longer-term problem is that I really dislike the glamorization of science, and the false impression it gives that science is like a magical lore, something that when you learn it gives you a direct line to the mind of God, so you can solve all kinds of knotty problems bang, like that.
Where the truth is that science is more like systematic pessimism, a way to winnow wheat from chaff when it comes to ideas -- and there is a lot of chaff. I remember going down in the elevator with my chairman once, after some idea or other of mine had gone south, and out of frustration just suddenly grumbling out loud: godsdammit, 90% of science doesn't work! He thought it over, grinned, and replied: yep, that seems about right. About 90% of what I try belly-flops....
See, I feel like it's very important people understand this. That science is very much a case of try and try again, learn from your (copious) mistakes, and that patience, hard work, sobriety and a great cautiousness in your language -- being very careful to not use words that might be misinterpreted -- are actually key hallmarks of the good scientist.
But I recognize science is now big business, and there are salesmanship and PR aspects to it, in order to ensure its continuing political support and all. Oh well.
Right you are. Poor guy became the poster child for all kinds of absurd stuff.
One of my favorite Calvin and Hobbes cartoons says it all: Hobbes asks Calvin if he maybe shouldn't be worried about his math grade, Calvin says: Heck no! Remember, Einstein got a bad math grade once and look how he turned out! And my grades are even worse...!
Well, I respect the editors at Nature -- I even used to date one, ha ha -- but they're only human. So I dunno if this logical proposition is one with which I'd strictly agree:
Nature doesn't publish things in their letters section unless the result is really heavy hitting and solid.
Anyway, I have no idea why my post got modded up, either. I don't see how it could be "informative" since it's just a reflection of my personal taste .
OK, so it blows your mind. Cool. It doesn't blow mine, though, and I don't especially think I'm much less familiar with quantum weirdness than you. De gustabus non disputandum est, eh? Let us agree to disagree. Remember, I'm not saying I don't think the work should have been published, or published in Nature -- only that I personally find my mind solidly unblown. Feel free to count me among the Philistines.
No, no, I don't think you're being rude at all. It's a fair question. And I don't really have a good answer. Yes, you're right that many a merely curious slightly misplaced thread has turned out, when you pick pick pick at it, to turn into major scientific discovery. You can quote Michelson in 1895 to me if you want ("all future discoveries in physics will be in the 5th decimal place" or something like that).
But, on the other hand, it also seems to be true that the purpose of scientific publication has morphed over the decades, and not so much for the better. I recall as a post-doc in 1992 being stunned that one half of the linear feet of the Journal of Chemical Physics on the library shelves had been published since I went to college in 1980. And that journal was founded in 1932 or so! So why was the same amount of material published in the last 12 years as in the first 50? Was there a giant mushrooming in important discoveries?
I don't think so. Not in chemical physics. But when you read modern papers versus what was published 50-75 years ago, what stands out about the older work is its economy. Publication was less a way to achieve or prove status (it seems) and more just a way to communicate essential ideas to others. So, for example, when folks had an idea that just didn't work out, or an experiment that wasn't even an interesting failure, they seemed simply to not publish. That seems less true these days. I can understand why, mind you: nowadays often someone's graduate thesis and possibly post-doc career can hang on whether the work can be spun as some kind of success, and published.
It could just be I'm being curmudgeonly, of course. The old days were better yadda yadda. But I do sort of feel there is more publishing for the sake of proving to all and sundry that you have clever insights and thoughts, than to communicate actual discoveries that others will find useful. More vanity publishing, that is. I don't know this is true, and I respect those who think otherwise. It's just a feeling.
I wasn't arguing that. I'm just saying the reason this is hitting main-stream news and seeming exciting is because of the confusion. If there were no confusion, if everyone knew exactly what quantum information was and how it differed from what we normally think of as information, would this paper get the press it has? Would it be an article on/.? Hardly. Would it even be in Nature? I'm not sure.
I'm not blaming the authors. It isn't their fault people take it out of its strict scientific context and go wild.
But...mmmm, how shall I put this? When I read or review a scientific paper, I like to have some question or other answered definitively. We didn't know the value of X but now I've measured it and we do. No one could explain why Y followed on Z, and now here's a theory that does. That kind of thing.
I'm just not very fond of work that merely "raises interesting questions" or points out curious features of generally well-understood phenomena. I'm not saying such papers shouldn't be published or anything like that, but I see them as fairly low value.
Now, if everyone in quantum computation had heretofore believed strongly that quantum information could not be negative, and they had come along and proved it could, that might be well worth publishing. But I don't think that was the case.
I don't think it's a completely bogus argument, for the following reason:
Yes, it takes generally speaking much more energy to create nice clean-burning H2, so in principle you're foolishly trading pollution at the car tailpipe for even more pollution at the power-plant smokestack.
But this is focussing too much on the basic physics and not enough on the politics and the economics. It overlooks the fact that there are tens of millions of those car tailpipes, owned by people who are hard to regulate and who don't have much capital to upgrade their equipment, whereas there are only tens of those plant smokestacks, and they are owned by big corporations that can't easily hide from regulation or the public eye, and which have large amounts of capital to upgrade their equipment.
It's for this reason that it's commonplace that controlling pollution when it comes from a few large sources is politically and economically easier than controlling pollution when it comes from zillions of tiny dispersed sources. The experience of the AQMD in Southern California is a good example: they've said they are unable to much further impact air quality because all the big polluters are already regulated to the hilt, and they just can't do much at an economical cost about millions of folks using leaky propane grills and so on.
So while it may seem dumbass to start building H2 cars the ultimate power source of which is a few big coal plants, it isn't necessarily. You can apply very stringent regulation to those plants. You can insist they always have the latest and best pollution control technology. You can install extremely expensive equipment with a payback time of 50 years, because the plant is going to live that long. You can site the plants where the pollution is unusually easy to control (good weather, plenty of nearby water for cooling, whatever). You can operate them as efficiently as possible (full bore 24/7, for example). You can afford very expensive training for its few operators. And so on. None of which is possible for dispersed pollution sources like cars.
And lastly, of course, you can remove the very expensive bureaucracy that presently exists to regulate pollution from millions of individual cars.
That is, we get to get rid of the whole giant social cost of pollution regulation on individual cars, from GM's manufacturing costs to the time and money I have to pay to get a smog check every year, and instead we pay for a few highly centralized, tightly monitored, expertly designed and operated power generation plants. It might be a good idea.
Heck no. No one has ever proposed a feature of QM that conflicts with relativity. Even quantum teleportation doesn't transfer information faster than c, since the experimenters need to get together (or exchange radio messages) to discover what was transmitted, and of course that restricts things to less than c.
Generally, as I understand it, if you allow transluminal information flow you run into problems with causality. Since we tend to want to believe in a causal universe, this is a big no-no. Anyone indulging in FTL transmission of information is always arrested shortly before they commit the timecrime by the Illuminati and executed forthwith, if not sooner.
Well, except they're negative probability amplitudes, a horse of a fairly different (e.g. complex) color. But this is a minor quibble, and I agree with you generally.
Geez, don't be so cynical. After I got my PhD from Berkeley, with a dissertation in quantum mechanics, I taught the stuff to graduate students for five years or so. I've published QM papers in PRA and all that, too. So, yeah, I know what they mean. I'm perfectly qualified to review their Nature paper, if it comes to that, and I doubt I'm the only one like this reading/.
I have to say I'm not especially impressed by the work, however. The frisson of defining information as negative emerges ultimately from a semi-deliberate muddling of the distinction between the definition of information in the quantum computing context and information as we use the word in daily life. This is not hard useful scientific discovery so much as the scientific equivalent of making an outrageous pun.
But then I feel similarly about most of what's published in the Bell's Inequality, EPR paradox, quantum tele-whatever field. Getting cynical myself, maybe I am....bah, humbug...grumble...
You know, I was right with you right up until this:
But the best place to spend money, in my opinion, is accelerated research that supports reduced reliance on oil.
How does the US "rely" on oil in any greater way than it relies on the existence and open global market for any other commodity? For that matter, why not make the same isolationist mercantilist argument about cheap foreign labor (which we also "rely" upon to get our manufacturered goods), or foreign capital (which we "rely" upon to fund our trade deficit).
Without doubt, disruption of any internationally-traded commodity or service, including oil, would seriously jar the US economy. How could it not? We are tied to the world by our immense volume of international trade. But why single out oil? Oil is probably one of the least important commodities on which we have a "reliance," for several reasons: First, oil resources are distributed globally, from Russia to Venezuela, and traded widely on the international market. It would take disruptions all over the world to seriously interfere with the ability of the US to buy the stuff, and even then we have a substantial domestic supply. Second, oil is mostly simply burned to generate heat and power, and there are many other possible sources of heat and power, including wood, natural gas, alcohol fermented from crops, solar, wind, nuclear and plain old conservation.
Now, I'm not saying a jump in oil prices to $150/barrel wouldn't hurt, and give us a bit of a recession. Unemployment might jump up to some horrible number like....oh, 10% or something, as in 1982. A million people might lose their homes. Inflation might rise to 12%, even, as it did during the last oil shocks in the 70s. This is harsh, to be sure. But it is hardly Gotterdammerung, hardly the Great Depression and the Black Plague rolled into one. And anyway, think about after the initial shock: Suddenly all kinds of already understood and ready to go alternative energy technologies become relatively cheap, and wind farms get built and new houses come with solar panels, and the price of these things falls steeply with commodization, research into electric cars booms, et cetera and so forth. In a decade at most we're probably right back where we were.
But compare to some other things the government spends money on, like (say) the CDC and NIH research into antivirals. Suppose avian flu finally makes the jump to humans like people think it might and roars through the United States. You know how easy it is to catch the flu from people, and avian flu has a mortality rate of 50%. Now that is a threat the likes of which no conceivable oil shock can match!
Or, to get back to the topic, if a national technical means in orbit can spot the NoKo missile launch facility with an ICBM fueling up and an orbital asset can take it out before it launches for Los Angeles with a 10 megaton warhead aboard....again, the disaster averted seems to me far larger than the threat of $4 or $5 per gallon at the pumps. But maybe that's because I live in Los Angeles, and I will just take the damn bike to work if the car became outrageously expensive.
Slashdot Mirror
He's also overlooking the factor of time. Evolution is only guaranteed to converge on the global optimum over an infinite time, and, in unsually perverse parameter spaces, it can certainly take an amount of time wholly unavailable to any reasonable application, so that you begin to think the algorithm will "never" converge to the global optimum.
I think this is one reason people find biological evolution fairly easy to swallow when it applies to microbes becoming more drug-resistant, but harder to credit when it applies to dinosaurs evolving wings and turning into birds. It's just a lot harder to wrap your head around the millions of generations such a complex optimization takes, and visualize the exceedingly unlikely semi-random walk through parameter space that happened. We start to rear back and say no way, that's just so damn unlikely it could never happen.
But our daily experience -- our common sense -- does not train us well to distinguish between could never happen and couldn't happen in less than ten million years, and, of course, over a billion years or so the latter is actually functionally equivalent to guaranteed to happen.
Oh let's think positive. Maybe one of the problems in AI is that the part of our thinking of which we're most aware -- the conscious, logical, linear, step-by-step part after which we model our computer programs -- is the part that does the least of our daily-life thinking. Maybe the bulk of our daily-life thinking is half-unconscious and massively parallel, to suit our hardware (10^12 neurons with a 1 kHz clock cycle).
Indeed, I find it fairly plausible that the bulk of our normal "human" ability to usefully and sanely respond to each other might have less to do with peering many steps ahead down complex decision trees and more to do with being able to recognize a huge number of variations on certain basic patterns and situations, and doing so on the basis of incomplete and noisy data. In other words, what makes humans good at, well, being human is more like recognizing a familiar face among thousands you've never seen or solving Where's Waldo? puzzles than playing three-dimensional chess with the M-5.
Alas, the bulk of the computing community has developed fairly little expertise in trying to imagine how to build, program and debug that kind of hardware/software combination.
So maybe that's all the problem is. We might just need to get used to building 500 nm wide nano-CPUs running pico-kernels, and figure out how to wire up a few billion to accomplish pattern recognition tasks as fast as we do.
Just taking a guess here, but I'd say they'll only consider lossless compression schemes (no point in throwing away data it took $400 million to collect), and that photos of Mars are not boring enough (e.g. with vast seas of one-color pixels) to be very compressable via lossless algorithms.
Let us not forget that as the black hat prey evolve the white hat predators do, too, and the latter usually have more funding.
So, if we imagine those stealthy microcrimsubs creeping up to a lonely San Diego beach, we might as well also imagine them bumping their little snouts 100 yards offshore into the dog pod grid. Bzzt snap glub glub glub. Problem solved!
Movies to the contrary, I sort kinda suspect the bulk of criminal smuggling is low tech. You just pay a dumfuk mule a G to take this "package of condoms" across the border, saying it's hers, and then give it to Hernando on the other side, who'll give her another 2 G's. If she gets busted, you're out less than the cost of a minicrimsub, you don't tip off the G-men to your methodology, and there are fewer material clues pointing back to your secret base under the crater of the extinct volcano.
Now that seems like a lot of data. Let's see, I read somewhere it's going to image 1% of the surface of Mars. Mars has a surface area of [Googles radius, punches calculator] 1.4e+14 m^2, so 1% of that is 1.4e+12 m^2. If the smallest thing you can resolve is a coffee table, and that's about 1 m^2, then that suggests each pixel is 1 m^2, so we have 1.4e+12 pixels coming back. Full color, natch, so no less than 32 bpp, totaling 5.8 terabytes.
That's a lot of data. If it has to get back here in a year or so, that's more than 1.4 megabit/second through your deep-space radio modem, even if you transmit around the clock all year.
I conclude either it's going to take substantially more than a year, or they've actually got a deep-space radio T1.
It's probably where you put the typewriter and the carbon paper.
But that's not what I came here to tell you about.
I came to talk about the draft.
You know, it's also possible that, if implemented right (I know, fat chance, but still...) that RFID identification could increase privacy and security.
Consider: one of the big problems of modern life is you have to prove your ID, credit and legitimacy to all kinds of people all over the place, with the consequences that sooner or later your private info leaks out and bad guys can get ahold of it, zap, identity theft, credit card fraud, and so forth.
But what if something like an RFID tag could be provided by one tightly controlled source, and it gave unimpeachable evidence of right to be there on the airplane, credit enough to buy the laptop from Best Buy, whatever.
Then imagine: you walk into the Best Buy with your bankcard with the RFID tag. You pick up the laptop and walk out. Now, the BB security can let you out, because the bank's card has told them you've credit enough to buy the laptop and given them some secret code that guarantees them payment, and when they get that payment they have to give your digital assistant a secret code that guarantees you can get warranty repairs. But your card hasn't told them a damn thing else about you. You haven't had to tell BB your address or e-mail or bank name or even your own name. No junk mail from Best Buy, no tracking your purchases, no poking their nose into your credit history...
Same thing with the airplane. You get an airline ticket after having proved who you are and where you can be found, and in some way -- OK, things get a little fuzzy here, but bear with me -- that you're safe and can be trusted, and then you walk into the airport and onto the airplane. The chip says "legit, allowed on plane to Boston at 6.47" but nothing else. No one checks your driver's license or passport eight zillion times, no one bends you over to search for bombs up your...but I digress...
Anyway, one of the reasons we have to send out all this extra, unnecessary, privacy-endangering information about ourselves is because we don't have one rock-solid unforgeable way of identifying those narrow aspects of ourselves (our citizenship, credit, student status, license to drive, et cetera) that are legitimate necessities of certain transactions. Maybe RFIDs with some digital signature technology could provide one? Maybe the future could be more private and secure?
Or have I had too much caffeine already? [Peers anxiously into empty mug...]
But you overlook the possibility that the true Buddha wishes to travel incognito, so that (among other things) he need not be bothered by the shallow conversation of false disciples overly impressed by such ephemeral and wordly measures of wisdom, and can keep his select circle of the Elect quietly anonymous...listening...watching...making the occasional, small, barely noticeable Adjustment...
OK, this one befuddles me, so I throw it out hoping a geekier geek will know the answer:
Why hasn't there been a significant move towards digital signatures a la PGP signatures on e-mail? With legit sigs registered with and guaranteed by their firms (work mail) or ISPs (personal), so everyone can find a spammer when he spams and....uh...do something firm but completely legal and moderate to them...nothing at all like staking them on an anthill or trying out setting #10 on The Machine, really.
Anyway, doesn't it seem easier for people to get used to the idea that senders should validate themselves up front, rather than leaving it up to receivers to sift out valid senders -- and with no better route than trying to analyze natural language, for God's sake, a major AI/expert-system challenge itself.
What am I missing??
Jesus, tell me about it. I get 30kb attachments merely saying "Got your email, thanks!" with "thanks" done up in some odd curly red font and a six-line sig, not to mention the twenty-seven 8x10 colored glossy JPG attachments with circles and arrows and a paragraph on the back of each one...
I am hard pressed to believe that computers are especially useful in learning, except in two areas, online searching and programming.
Much of what students seem to do with a computer sometimes seems like so much eye candy. You word process your paper instead of type it, put it in proportional Verdana font instead of mono, put bullets and footnotes and italics in with gay abandon. Or you make nice drawings with a vector drawing program instead of by hand. Or, as one poster said elsewhere, you type copious notes and then do searches on them later, instead of (say) thinking hard about what you're hearing as you hear it, and taking minimal but well-organized notes, which worked for generations before this.
But the availability and common use of online information is completely new to this generation, I think. In decades previous we'd truck on down to the library and read books. That meant we rarely got the latest info, and often our info was more restricted -- we didn't find much stuff representing fringe viewpoints, or the viewpoints of people from quite different cultures -- and also, of course, we tended not to find much trash and outright garbage, because it had all been through a lot of review and editing. We didn't get much in the way of unfiltered raw mass opinion.
Now we do. And it seems to me a paradigm shift has occured, and we always will. So exposing kids to the new way we get information -- fast, unfiltered, copious, spasmodically -- and teaching them how to judge it, winnow it, use it responsibly and successfully -- these are very important things we should be doing, and they do need a computer.
I also wish students used those computers to learn programming, because programming teaches the key intellectual skill of debugging your ideas. Too often we think that because an idea is beautiful and successful and perfect in our imagination it will be so in the real world. Programming your ideas teaches very well that, alas, it doesn't work that way, and teaches you a healthy respect for the difference between theory and reality.
Well, here is a list of the world's top 14 oil producers, which are, in order: Saudi Arabia (10.4 bbl/day), Russia, the US, Iran, Mexico, China, Norway, Canada, Venezuela, the UAE, Kuwait, Nigeria, the UK, and Iraq (2.03 bbl/day). Covering the First World (Norway) to the Third (Nigeria), Western democracy (Canada) to Islamic theocracy (Iran) to the Worker's Paradise (China), and five of the seven continents, that seems pretty widely distributed to me!
I realize the Saudis are a pain, and Venezuela has that nutcase Chavez at the reins (for now), but we get along OK with Russia and Mexico and Norway et cetera. It seems a little hard to credit that our relationship with all of the major oil producers is going to go belly up at once.
What you've also got to remember is that economic dependency is a two-way street. It would certainly hurt the US if, say, the Saudis stopped selling oil to us -- but what do you think it would do to the Saudis? What else have they got to sell but oil? How are they going to keep their economy going?
We're talking about oil prices doubling soon, and tripling and quadrupling after that, on to a price point of infinity because there WILL BE NO MORE OIL...
I don't think so. You are assuming that consumption rates will not change with price, but that is silly. Raise the price, and people start to economize and find alternatives. Raise the price a lot, and people flock to the alternatives and consumption plummets. And as I said, there are many alternative sources of energy. They're not used as much as oil now only because oil is cheap. If oil becomes expensive, they will be.
In a decade at most the American military will be taking oil from around the world at gunpoint.
This is a total red herring, but I can't resist saying (just to pull your chain) that this would be fine with me. Life is struggle for survival, my friend. Get used to it or join the dinosaurs.
NO. You won't have a job, and neither will I.
Speak for yourself. I believe I've got the skills that would let me find a niche in any society at all. If the US collapses into the equivalent of 14th century Spain, look for me to be one of those bastards with an ostrich feather in my enormous hat, whose dozens of beefy retainers will proceed in front of me at a respectful distance to thrust commoners into the gutter and out of my way. Bwa ha ha ha!
Well, I might not put it so harshly, but I believe I largely agree with you.
I think part of the longer-term problem is that I really dislike the glamorization of science, and the false impression it gives that science is like a magical lore, something that when you learn it gives you a direct line to the mind of God, so you can solve all kinds of knotty problems bang, like that.
Where the truth is that science is more like systematic pessimism, a way to winnow wheat from chaff when it comes to ideas -- and there is a lot of chaff. I remember going down in the elevator with my chairman once, after some idea or other of mine had gone south, and out of frustration just suddenly grumbling out loud: godsdammit, 90% of science doesn't work! He thought it over, grinned, and replied: yep, that seems about right. About 90% of what I try belly-flops....
See, I feel like it's very important people understand this. That science is very much a case of try and try again, learn from your (copious) mistakes, and that patience, hard work, sobriety and a great cautiousness in your language -- being very careful to not use words that might be misinterpreted -- are actually key hallmarks of the good scientist.
But I recognize science is now big business, and there are salesmanship and PR aspects to it, in order to ensure its continuing political support and all. Oh well.
Yeah, but I learned my shell command line geekery on Multics, so I laugh at the arrogance of all you Johnny-come-latelies....har de har har....
Right you are. Poor guy became the poster child for all kinds of absurd stuff.
One of my favorite Calvin and Hobbes cartoons says it all: Hobbes asks Calvin if he maybe shouldn't be worried about his math grade, Calvin says: Heck no! Remember, Einstein got a bad math grade once and look how he turned out! And my grades are even worse...!
Well, I respect the editors at Nature -- I even used to date one, ha ha -- but they're only human. So I dunno if this logical proposition is one with which I'd strictly agree:
Nature doesn't publish things in their letters section unless the result is really heavy hitting and solid.
Anyway, I have no idea why my post got modded up, either. I don't see how it could be "informative" since it's just a reflection of my personal taste .
OK, so it blows your mind. Cool. It doesn't blow mine, though, and I don't especially think I'm much less familiar with quantum weirdness than you. De gustabus non disputandum est, eh? Let us agree to disagree. Remember, I'm not saying I don't think the work should have been published, or published in Nature -- only that I personally find my mind solidly unblown. Feel free to count me among the Philistines.
No, no, I don't think you're being rude at all. It's a fair question. And I don't really have a good answer. Yes, you're right that many a merely curious slightly misplaced thread has turned out, when you pick pick pick at it, to turn into major scientific discovery. You can quote Michelson in 1895 to me if you want ("all future discoveries in physics will be in the 5th decimal place" or something like that).
But, on the other hand, it also seems to be true that the purpose of scientific publication has morphed over the decades, and not so much for the better. I recall as a post-doc in 1992 being stunned that one half of the linear feet of the Journal of Chemical Physics on the library shelves had been published since I went to college in 1980. And that journal was founded in 1932 or so! So why was the same amount of material published in the last 12 years as in the first 50? Was there a giant mushrooming in important discoveries?
I don't think so. Not in chemical physics. But when you read modern papers versus what was published 50-75 years ago, what stands out about the older work is its economy. Publication was less a way to achieve or prove status (it seems) and more just a way to communicate essential ideas to others. So, for example, when folks had an idea that just didn't work out, or an experiment that wasn't even an interesting failure, they seemed simply to not publish. That seems less true these days. I can understand why, mind you: nowadays often someone's graduate thesis and possibly post-doc career can hang on whether the work can be spun as some kind of success, and published.
It could just be I'm being curmudgeonly, of course. The old days were better yadda yadda. But I do sort of feel there is more publishing for the sake of proving to all and sundry that you have clever insights and thoughts, than to communicate actual discoveries that others will find useful. More vanity publishing, that is. I don't know this is true, and I respect those who think otherwise. It's just a feeling.
OK. I'm one of the QM people, then. Thanks for the link.
I wasn't arguing that. I'm just saying the reason this is hitting main-stream news and seeming exciting is because of the confusion. If there were no confusion, if everyone knew exactly what quantum information was and how it differed from what we normally think of as information, would this paper get the press it has? Would it be an article on /.? Hardly. Would it even be in Nature? I'm not sure.
I'm not blaming the authors. It isn't their fault people take it out of its strict scientific context and go wild.
But...mmmm, how shall I put this? When I read or review a scientific paper, I like to have some question or other answered definitively. We didn't know the value of X but now I've measured it and we do. No one could explain why Y followed on Z, and now here's a theory that does. That kind of thing.
I'm just not very fond of work that merely "raises interesting questions" or points out curious features of generally well-understood phenomena. I'm not saying such papers shouldn't be published or anything like that, but I see them as fairly low value.
Now, if everyone in quantum computation had heretofore believed strongly that quantum information could not be negative, and they had come along and proved it could, that might be well worth publishing. But I don't think that was the case.
I don't think it's a completely bogus argument, for the following reason:
Yes, it takes generally speaking much more energy to create nice clean-burning H2, so in principle you're foolishly trading pollution at the car tailpipe for even more pollution at the power-plant smokestack.
But this is focussing too much on the basic physics and not enough on the politics and the economics. It overlooks the fact that there are tens of millions of those car tailpipes, owned by people who are hard to regulate and who don't have much capital to upgrade their equipment, whereas there are only tens of those plant smokestacks, and they are owned by big corporations that can't easily hide from regulation or the public eye, and which have large amounts of capital to upgrade their equipment.
It's for this reason that it's commonplace that controlling pollution when it comes from a few large sources is politically and economically easier than controlling pollution when it comes from zillions of tiny dispersed sources. The experience of the AQMD in Southern California is a good example: they've said they are unable to much further impact air quality because all the big polluters are already regulated to the hilt, and they just can't do much at an economical cost about millions of folks using leaky propane grills and so on.
So while it may seem dumbass to start building H2 cars the ultimate power source of which is a few big coal plants, it isn't necessarily. You can apply very stringent regulation to those plants. You can insist they always have the latest and best pollution control technology. You can install extremely expensive equipment with a payback time of 50 years, because the plant is going to live that long. You can site the plants where the pollution is unusually easy to control (good weather, plenty of nearby water for cooling, whatever). You can operate them as efficiently as possible (full bore 24/7, for example). You can afford very expensive training for its few operators. And so on. None of which is possible for dispersed pollution sources like cars.
And lastly, of course, you can remove the very expensive bureaucracy that presently exists to regulate pollution from millions of individual cars.
That is, we get to get rid of the whole giant social cost of pollution regulation on individual cars, from GM's manufacturing costs to the time and money I have to pay to get a smog check every year, and instead we pay for a few highly centralized, tightly monitored, expertly designed and operated power generation plants. It might be a good idea.
Heck no. No one has ever proposed a feature of QM that conflicts with relativity. Even quantum teleportation doesn't transfer information faster than c, since the experimenters need to get together (or exchange radio messages) to discover what was transmitted, and of course that restricts things to less than c.
Generally, as I understand it, if you allow transluminal information flow you run into problems with causality. Since we tend to want to believe in a causal universe, this is a big no-no. Anyone indulging in FTL transmission of information is always arrested shortly before they commit the timecrime by the Illuminati and executed forthwith, if not sooner.
...things that act like negative probabilities.
Well, except they're negative probability amplitudes, a horse of a fairly different (e.g. complex) color. But this is a minor quibble, and I agree with you generally.
Geez, don't be so cynical. After I got my PhD from Berkeley, with a dissertation in quantum mechanics, I taught the stuff to graduate students for five years or so. I've published QM papers in PRA and all that, too. So, yeah, I know what they mean. I'm perfectly qualified to review their Nature paper, if it comes to that, and I doubt I'm the only one like this reading /.
I have to say I'm not especially impressed by the work, however. The frisson of defining information as negative emerges ultimately from a semi-deliberate muddling of the distinction between the definition of information in the quantum computing context and information as we use the word in daily life. This is not hard useful scientific discovery so much as the scientific equivalent of making an outrageous pun.
But then I feel similarly about most of what's published in the Bell's Inequality, EPR paradox, quantum tele-whatever field. Getting cynical myself, maybe I am....bah, humbug...grumble...
You know, I was right with you right up until this:
But the best place to spend money, in my opinion, is accelerated research that supports reduced reliance on oil.
How does the US "rely" on oil in any greater way than it relies on the existence and open global market for any other commodity? For that matter, why not make the same isolationist mercantilist argument about cheap foreign labor (which we also "rely" upon to get our manufacturered goods), or foreign capital (which we "rely" upon to fund our trade deficit).
Without doubt, disruption of any internationally-traded commodity or service, including oil, would seriously jar the US economy. How could it not? We are tied to the world by our immense volume of international trade. But why single out oil? Oil is probably one of the least important commodities on which we have a "reliance," for several reasons: First, oil resources are distributed globally, from Russia to Venezuela, and traded widely on the international market. It would take disruptions all over the world to seriously interfere with the ability of the US to buy the stuff, and even then we have a substantial domestic supply. Second, oil is mostly simply burned to generate heat and power, and there are many other possible sources of heat and power, including wood, natural gas, alcohol fermented from crops, solar, wind, nuclear and plain old conservation.
Now, I'm not saying a jump in oil prices to $150/barrel wouldn't hurt, and give us a bit of a recession. Unemployment might jump up to some horrible number like....oh, 10% or something, as in 1982. A million people might lose their homes. Inflation might rise to 12%, even, as it did during the last oil shocks in the 70s. This is harsh, to be sure. But it is hardly Gotterdammerung, hardly the Great Depression and the Black Plague rolled into one. And anyway, think about after the initial shock: Suddenly all kinds of already understood and ready to go alternative energy technologies become relatively cheap, and wind farms get built and new houses come with solar panels, and the price of these things falls steeply with commodization, research into electric cars booms, et cetera and so forth. In a decade at most we're probably right back where we were.
But compare to some other things the government spends money on, like (say) the CDC and NIH research into antivirals. Suppose avian flu finally makes the jump to humans like people think it might and roars through the United States. You know how easy it is to catch the flu from people, and avian flu has a mortality rate of 50%. Now that is a threat the likes of which no conceivable oil shock can match!
Or, to get back to the topic, if a national technical means in orbit can spot the NoKo missile launch facility with an ICBM fueling up and an orbital asset can take it out before it launches for Los Angeles with a 10 megaton warhead aboard....again, the disaster averted seems to me far larger than the threat of $4 or $5 per gallon at the pumps. But maybe that's because I live in Los Angeles, and I will just take the damn bike to work if the car became outrageously expensive.