(Commence massive hand waving) The more I think about it, the more intuitive the connection becomes between discrete states of Turing machines and being restricted to the realm of the countably infinite.
If I'm thinking about it correctly, the idea of a distinct class of countably infinite computable numbers is fascinating. With them, you can partition out the computer output of your choice from what the computer is trying to produce or represent. Pseudo-random number generators and simulations of physical phenomena are two things I'm concerned with in particular, and they both try to represent something in the uncountably infinite domain with something that isn't.
After checking your posting history, I'm going to ignore the the inevitable YHBT and respond, since you seem to have garnered a high rating anyway without exhibiting anything more than a buzzword knowledge of a related (but not specifically relevant) subject.
Go into your University Library, sit down at one of the nice P-II's , and point your browser to Science Online, surf to the current issue and article ("Broadband Modulation of Light by Using an Electro-Optic Polymer") and quit whining about paper copies. In fact, you can probably do this from nearly anywhere in.edu.
Also, to correct the opinion of anyone else out there, this article is not about a fiber. This polymer device is an Electro-Optical modulator, a gadget used to transfer electrical signals onto optical waveforms inside a fiber. To repeat, this is not a type of fiber. Modulators are absolutely critical components in optical communications. They have little to do specifically with secure quantum communications.
It is, however, an extremely fast modulator. You can currently buy 40-50 GHz response modulators packaged and ready to go. Historically, polymer modulators have been a bit higher power and tend to degrade (decay) more over time than the competing technology.
One interesting milestone to note is that communications systems go up in factors of four. Current implementation in the ground is 10 Gb/s. Current state of the art (that a slew of telecom startups are crashing and burning selling because no one is buying) is 40 Gb/s. This is the first modulator I've seen that might do 160 Gb/s.
I work for a group that uses and researches mode-locked lasers for a different purpose, and there seem to be a few inaccuracies and misconceptions floating around.
Mode locked lasers have been around since at least 1970, and are widely used for laser and communications research and telecom applications. It's a technology that's getting better as people like Delfyett iron out the substantial kinks, but it's hardly a new technique.
Talking about the size of these things, let's trace a source or two. UCF News release: "smaller than the head of a pin." NY Times: "that potentially could nestle on the head of a pin." Really, the individual optical components (Mach-Zehnder modulator, fiber, electronic oscillator, waveplates, filters, polarization control, isolators, etc.) reqired for a mode locked laser are each at least a few cubic centimeters. With good packaging, you might get a mode locked laser into a lunchbox, but no way as small as a pinhead. That's not to say thirty years down the road they won't have figured out how to put all of these components onto a chip and make it tiny, but it's not happening any time in the next ten years without a serious revolution in optics miniturization.
Mode locked lasers utility as a clock is limited to very fast things, as they're not good clocks for long periods of time. The numerical measure that is quoted for clock accuracy (-140 dBc/Hz, for those of you that follow the stuff) isn't good for time periods longer than 0.1 microseconds. Go below 1-10 MHz (more than 1 - 0.1 microseconds worth of time) or so, and the inaccuracy goes up exponentially for any mode locked laser. Now, that's a thousand clock ticks for a GHz processor, so it's fine for chip clocks. Just don't try timing anything that takes a millisecond.
Compare this to atomic clocks, which get worse over these short time scales, but much, much better over any human time scale. There are physics-based reasons for this, mostly involving the temperature, vibration, and other nusiance noises applied to a few very simple atoms suspended in a shielded vacum trap vs the stability of a bunch of optical components in a box that constantly have optical and electrical power flowing through them.
This is why NIST is looking at linking the two types of clock, not replacing one with the other.
This story isn't really as sensational as it's made out to be. Read the paper a bit, and you'll find, right near the conclusion, this:
Negative transmission times is a wave phenomenon, which together with superluminal propagation
has been observed for light propagation through wave guides and through dispersive atomic media. It is not suprising that the [equations we used] show similar effects, and, indeed, Ray Chiao et. al. [ref] have suggested a [...] mechanism. [...] From our [...] analysis it is not clear whether it is due to this mechanism or if it is more closely related to the time delays which may also be observed in normal wave packet tunneling through barriers
(emphasis mine)
This is a phenomenon that has been observed in other places already, either through quantum tunneling or through some similar quantum effect suggested by the group (Chiao et. al.) mentioned in the article. The (non-sensational) news is that it occurs in BEC's as well.
You know, a current trend in military electronics is to move towards using COTS (Commercial Off-the-Shelf) parts.
Imagine: Now foriegn militaries and terrorists will cover themselves in copywrighted images as the new camouflage. Screw the newly designed sophisticated Marine cammies. If that low-light CCD sees a Coca-Cola logo, it'll just shut down.
Perhaps instead of playing loud american rock music to get holed-up third world dictators to surrender, third-world dictators will be able to play loud american rock music to shut down communications to holed-up american troops. Boy band CD's become military weapons...
There is a big problem with comparing Salon readership to Famous Research Company Tech Report (FRCTR) readership. They're not the same.
The people who buy FRCTR aren't paying out of their own pockets. They are professionals who consider FRCTR advantageous to their work, and so get their organization or grant sugar daddy to pay for the quite substatial cost (I'm guessing at least a couple thousand apiece) of FRCTR as a necessary expense. Because 1) they paid so much for it and 2) are competing with others in their field, they value the information and preserve it, as you point out. The relationship is essentially professional, and works.
Salon readers are a different animal. Thirty bucks ain't that much. And what possible competitive advantage could there be to the information contained in a witty TV review or interesting political analysis? The relationship is weakened because the percieved value is less. The product is largely entertainment, and the relationship reflects that.
And because of the low cost, there are going to have to be many, many more Salon subscribers than FRCTR subscribers. Only a few might treat the personal relationship badly enough to abuse it. But all it takes is for a critical few subscribers to defect each day/month/year, and you can't control the rip-off.
There's another factor when you start moving from a professional product to an entertainment product: the 133t HaX0r factor. J. 133t HaX0r can hack an entertainment service and get social rewards from his or her peers, because they're interested, at least a little. Design engineers don't live in that environment, and the technical product is not very interesting to everyone outside that community. So from FRCTR to Salon, you have two forces changing the behavior of the net community. The first decreases the penalties for spreading the service around, and the second increases the rewards. Relying on the same principles to keep you up is going to fail.
I suppose it's my cynical side showing itself, but I just don't think a subscription relying on relationship and professionalism is going to work for a wide-audience internet news and entertainment magazine. I suppose you could try a public television/radio model, where you count solely on the user's "best interest to make sure that the company fulfilling the other end of your subscription-contract is still around" to provide, but I don't think Salon is that far ahead of the rest of the online pack.
You've got some good (but cynical) points about the overall structure of the agency, but you've left out one major piece. The program managers themselves have a responsibility to find new and interesting projects in their expertise that fulfill this "high-risk, high-payoff" goal. The desire to take credit is quite possibly the motivation of the political appointees at the top of the agency, and the reason why the program was approved and given funds.
The proposal for the program itself probably came from some technically competent program manager who has intrest in and knowledge of open source, and a desire to see what defense applications can come out of it.
I'm sure that there are some people in DARPA who are at least as interested in developing cool new technologies as covering their asses.
In the document itself, they even say that the primary goal of the program is to achieve "Revolutionary advances in the state-of-the-art [...] improving the security functionality, services, and assurance of existing open source operating systems." The question is whether the tens of millions of dollars that DARPA is going to spend will do as much good as the millions they spent trying to realize "distributed networking" did for what is now the internet. It probably won't, but it can't be a bad thing for the community, because it's not like they can buy open source and control the means of production of Free Software.
One other thing that might be motivating this study is the increased worrying in the Pentagon about information warfare. They look around and realize that they don't have a fraction of the best hackers. If it comes down to a real war where the existence of the US is threatened, what are they going to do? They can't draft them and expect them to work, and they probably don't have the resources (human or legal -- as a government agency, the DoD is somewhat limited in what they can pay people) to go on an all-out recruiting binge. So how do you use some of the talent that is out there? Maybe you can get some help from what the best are doing for themselves.
This has been discussed before, when various companies have come out with optical switches.
The trouble is not ampification (Erbium-Doped Fiber Amplifiers, EDFA's, are a stock part at many laser suppliers) or raw switching (see recent slashdot articles), the problem is packet switching. In order to read the header on a packet and decide where that packet should go, you need to convert back to electrical. There's where you take your performance hit. What's really needed for hugely fast networks is all-optical switching logic. And while I can see how you might make some incremental improvements when you bring the electronics closer to the lasers, I don't see a revolution in tech here.
This story is about being able to integrate light and electronics. There are lots of different types of laser and detecting diodes, but to integrate them directly onto a single package, you currently need to fuse different chips onto a single substrate. This process is not the easiest or cheapest way to go. If you can integrate the photonics and the electronics onto a single chip, it would shortcut a bunch of the multi component issues.
The article also tries to hint toward all-optical computing. As far as I can see, this work makes no real steps toward all optical computers. Claims five or ten years ago about optical computing have proved disappointments. No one can get it to work well. As far as I can see, this only is a possibility for better integration between electronics and optics, not a fundamental component of an optical computer. I could be wrong, of course, so if anyone out there knows, I'd be eager to hear how to build one with these.
Well, I don't travel out there that often, but I do make a round now and then. From what I remember, a round trip air shuttle flight between those three cities typically (non-promotional rate) costs between $150 and $300, depending on what times and what airports you use. Major hubs, like JFK, are cheaper to go to/from than regional airports, like Baltimore.
If I remember, the rail between Boston and New York is supposed to take about three hours, and a plane flight takes about an hour, not counting airport check in time and security, etc. If the train stations at both ends are in convenient places, it might be about a wash, time-wise.
The problem is that when you get out to larger distances, the proportion of check in time to travel time gets smaller, and air travel is just faster. So to make it time-economical, you need high-traffic routes that aren't much farther apart than Boston and New York or Washington and Philadelphia, and you just don't get that much of anywhere in the US besides the east coast. They'd have to get significantly cheaper to compete with airlines anywhere else.
You've got to be a bit careful when using these measures. The 'distinct parent' number seems more like it would be useful for thinking about how much population broadening goes on, and how much general inbreeding there is. But I can't see how it changes the likelihood that you have a particular set of (n x great)-grandparents in common. More reasurring generally for the inbreddedness, but no less creepy for thinking about specific examples.
Do we have species differentiation going on within the homo sapiens clan? We must, surely, or at least the very beginnings of it. It's ludicrous to think that we couldn't be evolving, after all, we're still shagging and mutating, and what else do you need for evolution to occur?
You need at least two more things that are assumed by most, but are actually pretty significant: evolutionary pressure (you don't breed) and a population to propogate the mutations.
There's also a timescale issue, which says that groups need to be distinct and pressures felt over long enough of a timescale.
For most serious mutations, this means at least a couple dozen generations worth. (Although this last comes with a caveat or two. See below)
I would argue that only very limited forms of evolution are going on in the human race right now. Mostly this seems to be due to technology.
For evolutionary pressure, we have removed the pressure due to all but the most serious afflictions. There's a bevy of disorders, such as hemophilia, that would have killed enough of the effected population (And yes, I know about dominant and non-dominance: the statistics over generations will still select away from the traits anyway, even if it doesn't show up in every individual) to provide an evolutionary pressure. We have completely relieved many of these pressures.
For the seperate population issue, we have technology again. There is so much interaction between seperate populations via modern transportation, that over the timescales of generations, I sincerely doubt that speciation could occur with humans on our present earth.
Now, the timescale issues become different for certain systems which themselves evolve at an accelerated rate. It also becomes different for selection pressures that occur almost entirely in one generation, and not statistically over many individuals. I would argue that the prime example of this is disease. The immune system has mechanisms built in to evolve over the course of an individual's lifetime, and differences of a single protein, expressed due to a single or a triplet of nucleotides, can make a huge difference.
Look at the African continent. The enormity of the HIV problem over there is staggering. But the selection process for children infected is fast enough to cut them off before they can have children, and the consequences are dire for every single individual. There must be a fantastic amount of evolution going on within that population.
Now, I'm not trying to make light of the problem, or suggest that it be treated as an experiment. It's a real problem, causing real human suffering, and we should try to alleviate it as soon as we can. But those are the instincts and imperatives of modern society, to alleviate suffering, which in most cases means taking off evolutionary pressure.
So yeah, the human race is evolving. But for the world as it currently exists, it's a different and more limited sort of evolution than the simple cases we are taught as classical evolutionary theory.
Also, you can't really leave the non-human element out of this. There is evidence around that homo sapiens and homo neanderthalis mated on numerous occasions.
Well, maybe. I've seen a couple debates go by on that one, and I don't consider myself well informed enough to make that call. Still, it doesn't change the calculus of ancestor number if there were a little extra diversity back there.
Imagine the soap opera potential in that. (chuckle)
Ever read Clan of the Cave Bear?
Not that great a book in my opinion, but it does get some mileage on exactly that premise.
I'm not sure how good such an algoritm would be, but there are trade-offs.
You've got to be a bit careful doing a time-based FFT on your audio, because that means you've got to unencode the entire file before you can play any meaningful segment. You can't stream it through: You've got to pay all your computing expenses up front instead of spreading it out through the entire wave file.
This might be okay for a scheme that doesn't take too much computing power, but if you want to incorporate all the splefty psycoacoustic models and the other stuff that's been flying around this discussion, a piece of software using such a scheme is going to take a hit.
Are you willing to wait for five seconds after selecting a song to hear it play? How about thirty seconds or a minute?
Check out the New Scientist article and NASA's own popularized article on the subject. Evidently, even the inventor hasn't really got how to steer worked out.
Still, he estimates that the specific impulse -- the ratio of fuel to thrust, a measure of efficiency -- is some 1000 times better than chemical rockets. Admittedly, this figure doesn't count the overhead necessary to generate the electrical power.
And don't forget, in addition to getting the thing off the surface, you need to get outside the Earth's own magnetosphere for this sail to work. The magnetosphere is only 10 earth radii on the sun side, but it extends millions of km on the other side, away from the sun. You'd probably kick the spacecraft out perpindicular to the sun line, something that would take about as much fuel as putting a satellite in geosynchronous orbit. Practical, but it takes a lot of overhead to do just that much.
Muon catalyzed fusion uses the muon, which is essentially a heavy electron, to produce fusion at near room temperatures. As others have pointed out, this effect has been experimentally observed since the 50's, mostly as a nusiance effect in particle accelerators. Of course, cold temperatures are relative. 1/40 of an eV is about room temperature, so 1 eV is pretty darn hot.
It works because the muon is some 207 times heavier than the electron, and the math works out that if you put it around a proton, making muon-hydrogen, the average distance of the muon from the proton is about 200 times closer than that of the electron in a similar state around a proton. This means that the muonic hydrogen atom itself is about 200 times smaller than the conventional hydrogen atom. It takes a lot less energy for other hydrogens to get close to the small atom (the couloumb barrier is much thinner), so you can cause fusion at much lower temperatures, less energy is needed to get the particles close enough to fuse.
The problem with muon-catalyzed fusion is that it takes muons, which are incredibly energy-expensive to produce. The efficiency of the particle accelerators needed to make them is miniscule. Less than tenths of a percent. No free lunch there, although there are a few people looking into its viability, especially as a catalyst toward more efficient fission reactors.
According to Jane's (I know most/.'ers don't have access to it, but I'll cite my source: Jane's Intelligence Review, May 01, 1995; June 01, 1995; June 01, 1998 "Homing in on Russia's approach to ASW"; November 01, 1999 "How Shkval ensured Soviet SSBN survivability") The VA-111 Shkval has been in development since 1964. It was accepted into service in 1977, and carried on Sierra, Mike and Akula class attack submarines.
It's rocket powered, and originally it carried a tactical nuclear weapon. It was intended to defend the SSBN's, Ballistic missile submarines, as a defense against the quieter US attack submarines. So it's not too out of line to speculate that an SSBN might be outfitted with them. It has no guidance whatsoever. As soon as an attacking torpedo was fired, a Shkval was supposed to be fired back along the incoming heading. It would reach a point between the incoming torpedo and the attacking sub, then detonate the tacnuke. The shockwave would destroy the attacking sub and torpedo.
Since treaties have forbade the use of tactical nukes on submarines, this role has been retired. As previously discussed, a new version of the weapon, the Shkval-E is now being made as a conventional weapon. The rocket motor has a 90 second burn time, it is 533.4mm in diameter, 8.2m long, and weighs in at 2700kg, with a 250kg warhead. Range is supposed to be about 10,000 yards, pretty short for a torpedo. It's supposed to be guided, although it would be damn hard to get it to turn very fast.
The Regulon -- Katz' model incomplete
on
The Regulon
·
· Score: 1
There is a significant problem with this assertion that information will devour us all. The model is wrong. Katz contends that information on modern and future networks can promulgate itself, that computers can copy and spew information ad infinitum, without regard to a selection for quality.
The problem is, that the worries in the article are about humans being overwhelmed with this spew of information. But human minds are the medium for information, and are a limiting resource to exponentiation. The regulon, in Katz-speak.
The dead information is just that, dead, and will be ignored by everyone but the computer it resides on. The information codfish that no one looks at are dead.
The new economy business model that has driven the proliferation of corporate web information cannot continue indefinitely. Companies that do not derive a profit from the information they put out will eventually be selected against. Money is the regulon of corporations, and will apply to corporate information.
"We have no way to keep CNN, weatherman, flamers, spammers, Web site designers, e-do gooders and nit-picking coders, pundits, zealots, smart-asses and grumps in check" As many previous posters point out, we do have a way to keep these in check, and our defences are evolving. Witness email filters, recent suits against spammers, the de-spamming of my address above, choosing the URLs you visit and no others, Refusing to buy from advertisements pushed upon you. I, as a reasonably intelligent modern person, can keep all this in check and more. Perhaps John Katz, as a public person, has more trouble than most in filtering his information, as so many desire to push theirs onto him. But there are options for public people. Witness Neal Stephenson's method of filtering.
The real problem is in the human capacity to understand and process. In the current state of the world, we have not yet reached this carrying capacity, and so the growth continues. The real problem is human. Are we intelligent enough to filter our information for the useful bits and keep the spew out? Are we intelligent enough to create tools to do this effectively for us? I would argue that we are.
I think that the cold fusion reference might come from a not-well-known phenomenon called muon-catalyzed fusion. Muon catalyzed fusion uses a subatomic particle called the muon, which is essentially a heavy electron, to produce fusion at near room temperatures. This effect has been experimentally observed since the 50's, mostly as a nusiance effect in particle accelerators.
It works because the muon is some 207 times heavier than the electron, and the math works out that if you put it around a proton, that means that the average distance of the muon from the proton is about 200 times closer than that of the electron in a similar state around a proton. This means that the muonic hydrogen atom itself is about 200 times smaller than the conventional hydrogen atom. It takes a lot less energy for other hydrogens to get close to the small atom (for you science types, the couloumb barrier is much thinner), so you can cause fusion at much lower temperatures.
The problem with muon-catalyzed fusion is that it takes muons, which are incredibly energy-expensive to produce. The efficiency of the particle accelerators needed to make them is miniscule. No free lunch there, although there are a few people looking into its viability.
So anyway, this sounds a fair bit like Mills' hydrinos, although with no claims whatsoever towards a Grand Unified Theory or antigravity or whatever. If Mills took a cue from this well explored (but somewhat obscure) theory to provide some background to his fusion claims, he sure tacked on a heap extra.
Oh, and from what I've seen, Pons and Fleischmann, when they did the experiments leading to the cold fusion incident, actually might have found something interesting. However, all their measurements relied on a bunch of calorimetery that is somewhat difficult to interpret, and no evidence that actual nuclear fusion was going on was accurately reproduced. Mills might have stumbled on to the same electrochemical effect (if it exists), but I sincerely doubt that he's got any genuinely stunning new physics.
I guess the proof is in the pudding: If he comes up with something that actually works and produces energy, I'll believe him.
Slashdot Mirror
Actually, if you look into it, they're using PPM.
Curious.
It must be said, IANAM(athematician).
(Commence massive hand waving) The more I think about it, the more intuitive the connection becomes between discrete states of Turing machines and being restricted to the realm of the countably infinite.
If I'm thinking about it correctly, the idea of a distinct class of countably infinite computable numbers is fascinating. With them, you can partition out the computer output of your choice from what the computer is trying to produce or represent. Pseudo-random number generators and simulations of physical phenomena are two things I'm concerned with in particular, and they both try to represent something in the uncountably infinite domain with something that isn't.
Thanks for the thought-provoking post.
BMagneton
After checking your posting history, I'm going to ignore the the inevitable YHBT and respond, since you seem to have garnered a high rating anyway without exhibiting anything more than a buzzword knowledge of a related (but not specifically relevant) subject.
.edu.
Go into your University Library, sit down at one of the nice P-II's , and point your browser to Science Online, surf to the current issue and article ("Broadband Modulation of Light by Using an Electro-Optic Polymer") and quit whining about paper copies. In fact, you can probably do this from nearly anywhere in
Also, to correct the opinion of anyone else out there, this article is not about a fiber. This polymer device is an Electro-Optical modulator, a gadget used to transfer electrical signals onto optical waveforms inside a fiber. To repeat, this is not a type of fiber. Modulators are absolutely critical components in optical communications. They have little to do specifically with secure quantum communications.
It is, however, an extremely fast modulator. You can currently buy 40-50 GHz response modulators packaged and ready to go. Historically, polymer modulators have been a bit higher power and tend to degrade (decay) more over time than the competing technology.
One interesting milestone to note is that communications systems go up in factors of four. Current implementation in the ground is 10 Gb/s. Current state of the art (that a slew of telecom startups are crashing and burning selling because no one is buying) is 40 Gb/s. This is the first modulator I've seen that might do 160 Gb/s.
BMagneton
I work for a group that uses and researches mode-locked lasers for a different purpose, and there seem to be a few inaccuracies and misconceptions floating around.
Mode locked lasers have been around since at least 1970, and are widely used for laser and communications research and telecom applications. It's a technology that's getting better as people like Delfyett iron out the substantial kinks, but it's hardly a new technique.
Talking about the size of these things, let's trace a source or two. UCF News release: "smaller than the head of a pin." NY Times: "that potentially could nestle on the head of a pin." Really, the individual optical components (Mach-Zehnder modulator, fiber, electronic oscillator, waveplates, filters, polarization control, isolators, etc.) reqired for a mode locked laser are each at least a few cubic centimeters. With good packaging, you might get a mode locked laser into a lunchbox, but no way as small as a pinhead. That's not to say thirty years down the road they won't have figured out how to put all of these components onto a chip and make it tiny, but it's not happening any time in the next ten years without a serious revolution in optics miniturization.
Mode locked lasers utility as a clock is limited to very fast things, as they're not good clocks for long periods of time. The numerical measure that is quoted for clock accuracy (-140 dBc/Hz, for those of you that follow the stuff) isn't good for time periods longer than 0.1 microseconds. Go below 1-10 MHz (more than 1 - 0.1 microseconds worth of time) or so, and the inaccuracy goes up exponentially for any mode locked laser. Now, that's a thousand clock ticks for a GHz processor, so it's fine for chip clocks. Just don't try timing anything that takes a millisecond.
Compare this to atomic clocks, which get worse over these short time scales, but much, much better over any human time scale. There are physics-based reasons for this, mostly involving the temperature, vibration, and other nusiance noises applied to a few very simple atoms suspended in a shielded vacum trap vs the stability of a bunch of optical components in a box that constantly have optical and electrical power flowing through them.
This is why NIST is looking at linking the two types of clock, not replacing one with the other.
BMagneton
This is a phenomenon that has been observed in other places already, either through quantum tunneling or through some similar quantum effect suggested by the group (Chiao et. al.) mentioned in the article. The (non-sensational) news is that it occurs in BEC's as well.
BMagneton
You know, a current trend in military electronics is to move towards using COTS (Commercial Off-the-Shelf) parts.
Imagine: Now foriegn militaries and terrorists will cover themselves in copywrighted images as the new camouflage. Screw the newly designed sophisticated Marine cammies. If that low-light CCD sees a Coca-Cola logo, it'll just shut down.
Perhaps instead of playing loud american rock music to get holed-up third world dictators to surrender, third-world dictators will be able to play loud american rock music to shut down communications to holed-up american troops. Boy band CD's become military weapons...
There is a big problem with comparing Salon readership to Famous Research Company Tech Report (FRCTR) readership. They're not the same.
The people who buy FRCTR aren't paying out of their own pockets. They are professionals who consider FRCTR advantageous to their work, and so get their organization or grant sugar daddy to pay for the quite substatial cost (I'm guessing at least a couple thousand apiece) of FRCTR as a necessary expense. Because 1) they paid so much for it and 2) are competing with others in their field, they value the information and preserve it, as you point out. The relationship is essentially professional, and works.
Salon readers are a different animal. Thirty bucks ain't that much. And what possible competitive advantage could there be to the information contained in a witty TV review or interesting political analysis? The relationship is weakened because the percieved value is less. The product is largely entertainment, and the relationship reflects that.
And because of the low cost, there are going to have to be many, many more Salon subscribers than FRCTR subscribers. Only a few might treat the personal relationship badly enough to abuse it. But all it takes is for a critical few subscribers to defect each day/month/year, and you can't control the rip-off.
There's another factor when you start moving from a professional product to an entertainment product: the 133t HaX0r factor. J. 133t HaX0r can hack an entertainment service and get social rewards from his or her peers, because they're interested, at least a little. Design engineers don't live in that environment, and the technical product is not very interesting to everyone outside that community. So from FRCTR to Salon, you have two forces changing the behavior of the net community. The first decreases the penalties for spreading the service around, and the second increases the rewards. Relying on the same principles to keep you up is going to fail.
I suppose it's my cynical side showing itself, but I just don't think a subscription relying on relationship and professionalism is going to work for a wide-audience internet news and entertainment magazine. I suppose you could try a public television/radio model, where you count solely on the user's "best interest to make sure that the company fulfilling the other end of your subscription-contract is still around" to provide, but I don't think Salon is that far ahead of the rest of the online pack.
BMagneton
You've got some good (but cynical) points about the overall structure of the agency, but you've left out one major piece. The program managers themselves have a responsibility to find new and interesting projects in their expertise that fulfill this "high-risk, high-payoff" goal. The desire to take credit is quite possibly the motivation of the political appointees at the top of the agency, and the reason why the program was approved and given funds. The proposal for the program itself probably came from some technically competent program manager who has intrest in and knowledge of open source, and a desire to see what defense applications can come out of it.
I'm sure that there are some people in DARPA who are at least as interested in developing cool new technologies as covering their asses.
In the document itself, they even say that the primary goal of the program is to achieve "Revolutionary advances in the state-of-the-art [...] improving the security functionality, services, and assurance of existing open source operating systems." The question is whether the tens of millions of dollars that DARPA is going to spend will do as much good as the millions they spent trying to realize "distributed networking" did for what is now the internet. It probably won't, but it can't be a bad thing for the community, because it's not like they can buy open source and control the means of production of Free Software.
One other thing that might be motivating this study is the increased worrying in the Pentagon about information warfare. They look around and realize that they don't have a fraction of the best hackers. If it comes down to a real war where the existence of the US is threatened, what are they going to do? They can't draft them and expect them to work, and they probably don't have the resources (human or legal -- as a government agency, the DoD is somewhat limited in what they can pay people) to go on an all-out recruiting binge. So how do you use some of the talent that is out there? Maybe you can get some help from what the best are doing for themselves.
BMangneton----------------
Care for a Spin?
This has been discussed before, when various companies have come out with optical switches. The trouble is not ampification (Erbium-Doped Fiber Amplifiers, EDFA's, are a stock part at many laser suppliers) or raw switching (see recent slashdot articles), the problem is packet switching. In order to read the header on a packet and decide where that packet should go, you need to convert back to electrical. There's where you take your performance hit. What's really needed for hugely fast networks is all-optical switching logic. And while I can see how you might make some incremental improvements when you bring the electronics closer to the lasers, I don't see a revolution in tech here.
This story is about being able to integrate light and electronics. There are lots of different types of laser and detecting diodes, but to integrate them directly onto a single package, you currently need to fuse different chips onto a single substrate. This process is not the easiest or cheapest way to go. If you can integrate the photonics and the electronics onto a single chip, it would shortcut a bunch of the multi component issues.
The article also tries to hint toward all-optical computing. As far as I can see, this work makes no real steps toward all optical computers. Claims five or ten years ago about optical computing have proved disappointments. No one can get it to work well. As far as I can see, this only is a possibility for better integration between electronics and optics, not a fundamental component of an optical computer. I could be wrong, of course, so if anyone out there knows, I'd be eager to hear how to build one with these.
BMagneton
Any opinions on what the best way to go would be?
Well, I don't travel out there that often, but I do make a round now and then. From what I remember, a round trip air shuttle flight between those three cities typically (non-promotional rate) costs between $150 and $300, depending on what times and what airports you use. Major hubs, like JFK, are cheaper to go to/from than regional airports, like Baltimore.
If I remember, the rail between Boston and New York is supposed to take about three hours, and a plane flight takes about an hour, not counting airport check in time and security, etc. If the train stations at both ends are in convenient places, it might be about a wash, time-wise.
The problem is that when you get out to larger distances, the proportion of check in time to travel time gets smaller, and air travel is just faster. So to make it time-economical, you need high-traffic routes that aren't much farther apart than Boston and New York or Washington and Philadelphia, and you just don't get that much of anywhere in the US besides the east coast. They'd have to get significantly cheaper to compete with airlines anywhere else.
BMagneton
1 child with 2/3 of a distinct parent.
You've got to be a bit careful when using these measures. The 'distinct parent' number seems more like it would be useful for thinking about how much population broadening goes on, and how much general inbreeding there is. But I can't see how it changes the likelihood that you have a particular set of (n x great)-grandparents in common. More reasurring generally for the inbreddedness, but no less creepy for thinking about specific examples.
Do we have species differentiation going on within the homo sapiens clan? We must, surely, or at least the very beginnings of it. It's ludicrous to think that we couldn't be evolving, after all, we're still shagging and mutating, and what else do you need for evolution to occur?
You need at least two more things that are assumed by most, but are actually pretty significant: evolutionary pressure (you don't breed) and a population to propogate the mutations. There's also a timescale issue, which says that groups need to be distinct and pressures felt over long enough of a timescale. For most serious mutations, this means at least a couple dozen generations worth. (Although this last comes with a caveat or two. See below)
I would argue that only very limited forms of evolution are going on in the human race right now. Mostly this seems to be due to technology.
For evolutionary pressure, we have removed the pressure due to all but the most serious afflictions. There's a bevy of disorders, such as hemophilia, that would have killed enough of the effected population (And yes, I know about dominant and non-dominance: the statistics over generations will still select away from the traits anyway, even if it doesn't show up in every individual) to provide an evolutionary pressure. We have completely relieved many of these pressures.
For the seperate population issue, we have technology again. There is so much interaction between seperate populations via modern transportation, that over the timescales of generations, I sincerely doubt that speciation could occur with humans on our present earth.
Now, the timescale issues become different for certain systems which themselves evolve at an accelerated rate. It also becomes different for selection pressures that occur almost entirely in one generation, and not statistically over many individuals. I would argue that the prime example of this is disease. The immune system has mechanisms built in to evolve over the course of an individual's lifetime, and differences of a single protein, expressed due to a single or a triplet of nucleotides, can make a huge difference.
Look at the African continent. The enormity of the HIV problem over there is staggering. But the selection process for children infected is fast enough to cut them off before they can have children, and the consequences are dire for every single individual. There must be a fantastic amount of evolution going on within that population.
Now, I'm not trying to make light of the problem, or suggest that it be treated as an experiment. It's a real problem, causing real human suffering, and we should try to alleviate it as soon as we can. But those are the instincts and imperatives of modern society, to alleviate suffering, which in most cases means taking off evolutionary pressure.
So yeah, the human race is evolving. But for the world as it currently exists, it's a different and more limited sort of evolution than the simple cases we are taught as classical evolutionary theory.
Also, you can't really leave the non-human element out of this. There is evidence around that homo sapiens and homo neanderthalis mated on numerous occasions.
Well, maybe. I've seen a couple debates go by on that one, and I don't consider myself well informed enough to make that call. Still, it doesn't change the calculus of ancestor number if there were a little extra diversity back there.
Imagine the soap opera potential in that. (chuckle)
Ever read Clan of the Cave Bear?
Not that great a book in my opinion, but it does get some mileage on exactly that premise.BMagneton
I'm not sure how good such an algoritm would be, but there are trade-offs.
You've got to be a bit careful doing a time-based FFT on your audio, because that means you've got to unencode the entire file before you can play any meaningful segment. You can't stream it through: You've got to pay all your computing expenses up front instead of spreading it out through the entire wave file.
This might be okay for a scheme that doesn't take too much computing power, but if you want to incorporate all the splefty psycoacoustic models and the other stuff that's been flying around this discussion, a piece of software using such a scheme is going to take a hit.
Are you willing to wait for five seconds after selecting a song to hear it play? How about thirty seconds or a minute?
-BMagneton
Still, he estimates that the specific impulse -- the ratio of fuel to thrust, a measure of efficiency -- is some 1000 times better than chemical rockets. Admittedly, this figure doesn't count the overhead necessary to generate the electrical power.
And don't forget, in addition to getting the thing off the surface, you need to get outside the Earth's own magnetosphere for this sail to work. The magnetosphere is only 10 earth radii on the sun side, but it extends millions of km on the other side, away from the sun. You'd probably kick the spacecraft out perpindicular to the sun line, something that would take about as much fuel as putting a satellite in geosynchronous orbit. Practical, but it takes a lot of overhead to do just that much.
Muon catalyzed fusion uses the muon, which is essentially a heavy electron, to produce fusion at near room temperatures. As others have pointed out, this effect has been experimentally observed since the 50's, mostly as a nusiance effect in particle accelerators. Of course, cold temperatures are relative. 1/40 of an eV is about room temperature, so 1 eV is pretty darn hot.
It works because the muon is some 207 times heavier than the electron, and the math works out that if you put it around a proton, making muon-hydrogen, the average distance of the muon from the proton is about 200 times closer than that of the electron in a similar state around a proton. This means that the muonic hydrogen atom itself is about 200 times smaller than the conventional hydrogen atom. It takes a lot less energy for other hydrogens to get close to the small atom (the couloumb barrier is much thinner), so you can cause fusion at much lower temperatures, less energy is needed to get the particles close enough to fuse.
The problem with muon-catalyzed fusion is that it takes muons, which are incredibly energy-expensive to produce. The efficiency of the particle accelerators needed to make them is miniscule. Less than tenths of a percent. No free lunch there, although there are a few people looking into its viability, especially as a catalyst toward more efficient fission reactors.
According to Jane's (I know most /.'ers don't have access to it, but I'll cite my source: Jane's Intelligence Review, May 01, 1995; June 01, 1995; June 01, 1998 "Homing in on Russia's approach to ASW"; November 01, 1999 "How Shkval ensured Soviet SSBN survivability") The VA-111 Shkval has been in development since 1964. It was accepted into service in 1977, and carried on Sierra, Mike and Akula class attack submarines.
It's rocket powered, and originally it carried a tactical nuclear weapon. It was intended to defend the SSBN's, Ballistic missile submarines, as a defense against the quieter US attack submarines. So it's not too out of line to speculate that an SSBN might be outfitted with them. It has no guidance whatsoever. As soon as an attacking torpedo was fired, a Shkval was supposed to be fired back along the incoming heading. It would reach a point between the incoming torpedo and the attacking sub, then detonate the tacnuke. The shockwave would destroy the attacking sub and torpedo.
Since treaties have forbade the use of tactical nukes on submarines, this role has been retired. As previously discussed, a new version of the weapon, the Shkval-E is now being made as a conventional weapon. The rocket motor has a 90 second burn time, it is 533.4mm in diameter, 8.2m long, and weighs in at 2700kg, with a 250kg warhead. Range is supposed to be about 10,000 yards, pretty short for a torpedo. It's supposed to be guided, although it would be damn hard to get it to turn very fast.
There is a significant problem with this assertion that information will devour us all. The model is wrong. Katz contends that information on modern and future networks can promulgate itself, that computers can copy and spew information ad infinitum, without regard to a selection for quality.
The problem is, that the worries in the article are about humans being overwhelmed with this spew of information. But human minds are the medium for information, and are a limiting resource to exponentiation. The regulon, in Katz-speak. The dead information is just that, dead, and will be ignored by everyone but the computer it resides on. The information codfish that no one looks at are dead.
The new economy business model that has driven the proliferation of corporate web information cannot continue indefinitely. Companies that do not derive a profit from the information they put out will eventually be selected against. Money is the regulon of corporations, and will apply to corporate information.
"We have no way to keep CNN, weatherman, flamers, spammers, Web site designers, e-do gooders and nit-picking coders, pundits, zealots, smart-asses and grumps in check" As many previous posters point out, we do have a way to keep these in check, and our defences are evolving. Witness email filters, recent suits against spammers, the de-spamming of my address above, choosing the URLs you visit and no others, Refusing to buy from advertisements pushed upon you. I, as a reasonably intelligent modern person, can keep all this in check and more. Perhaps John Katz, as a public person, has more trouble than most in filtering his information, as so many desire to push theirs onto him. But there are options for public people. Witness Neal Stephenson's method of filtering.
The real problem is in the human capacity to understand and process. In the current state of the world, we have not yet reached this carrying capacity, and so the growth continues. The real problem is human. Are we intelligent enough to filter our information for the useful bits and keep the spew out? Are we intelligent enough to create tools to do this effectively for us? I would argue that we are.
BMagneton
-------------"Care for a little Spin?"
I think that the cold fusion reference might come from a not-well-known phenomenon called muon-catalyzed fusion. Muon catalyzed fusion uses a subatomic particle called the muon, which is essentially a heavy electron, to produce fusion at near room temperatures. This effect has been experimentally observed since the 50's, mostly as a nusiance effect in particle accelerators.
It works because the muon is some 207 times heavier than the electron, and the math works out that if you put it around a proton, that means that the average distance of the muon from the proton is about 200 times closer than that of the electron in a similar state around a proton. This means that the muonic hydrogen atom itself is about 200 times smaller than the conventional hydrogen atom. It takes a lot less energy for other hydrogens to get close to the small atom (for you science types, the couloumb barrier is much thinner), so you can cause fusion at much lower temperatures.
The problem with muon-catalyzed fusion is that it takes muons, which are incredibly energy-expensive to produce. The efficiency of the particle accelerators needed to make them is miniscule. No free lunch there, although there are a few people looking into its viability.
So anyway, this sounds a fair bit like Mills' hydrinos, although with no claims whatsoever towards a Grand Unified Theory or antigravity or whatever. If Mills took a cue from this well explored (but somewhat obscure) theory to provide some background to his fusion claims, he sure tacked on a heap extra.
Oh, and from what I've seen, Pons and Fleischmann, when they did the experiments leading to the cold fusion incident, actually might have found something interesting. However, all their measurements relied on a bunch of calorimetery that is somewhat difficult to interpret, and no evidence that actual nuclear fusion was going on was accurately reproduced. Mills might have stumbled on to the same electrochemical effect (if it exists), but I sincerely doubt that he's got any genuinely stunning new physics.
I guess the proof is in the pudding: If he comes up with something that actually works and produces energy, I'll believe him.