But in a virtual reality, there is as much isolation or non-isolation as the simulation cares to provide at that given instance. It is entirely controllable - and entirely measurable. In the physical world, you cannot readily control such data, let alone measure it, and therefore cannot quantify its effects.
The first problem with just accepting "this is how it works" is that classical Hard AI solutions haven't worked, and they all rely on a direct connection.
The second problem with just accepting "this is how it works" is that autism is a consequence of the breakdown of the mechanism you describe, with the severity of the autism being a function of the severity of the breakdown. Autism has direct impact on the brain's capacity to function - more so than just about anything else. There can be quite a bit of brain damage without altering logical thought in the slightest, but even a few haywire mirror neurons can radically alter every thinking process.
Conclusion: The ability to symbolically represent and manipulate a synthetic reality is paramount the intelligence of humans. The ability to process that data is entirely secondary.
Chaos theory is 100% deterministic and 0% predictable. Chaos theory is 100% representable in pure mathematics and requires no emotional component. Chaos theory rules the marketplace. (Always has. It's how fractals were discovered by Mandelbrot, after all.)
Roundabouts are superior to traffic lights, in many respects. You don't hold up traffic at all, provided streaming is done right. The biggest problem is when they're used at small, infrequently-used intersections.
Neural nets can do nothing that is non-computable and are not suited to all kinds of problems. Petri nets are also quite interesting, but again have a very specific role in AI.
It has been shown that a single neuron from a physical brain can perform extremely complex operations. How is, as far as I know, unknown. However, it means that the most advanced computer neural nets are not yet as advanced as single organic neurons, and organic brains that have anything approximating intelligence have billions of organic neurons.
In other words, if you gave everyone on the planet an Origin 3000 and a terabit pipe, and ran all those computers as a single distributed neural net program, you would have a computer with the mental capacity of an African Grey parrot.
Now, clearly that much compute power is capable of doing FAR more work and is capable of vastly superior intellect, from which we can conclude that the problem lies not in the computers but in the techniques used. We simply don't have the software tools necessary for quality AI. At least, not by using any approach so far adopted.
My personal theory is that AI will evolve out of virtual reality because the human brain does not interact directly with the senses or muscles but via an internal simulation. All the physical world stuff is handled by external plugins to that internal VR. That seems simple to reproduce.
Since intelligence is merely an evolutionary byproduct of how that internal VR is handled, and as computers can run as many parallel handlers as there are processors thrown at the problem, I don't see that it's necessary for humans to ever solve how to program AI. Give it a self-contained MMORG and let it herustically develop itself.
Since AIs are, almost by definition, more predictable than humans, it is self-evident that AI customers can be more cost-effectively be tailored for and more easily swayed. Since limited intelligences already handle most financial decisions (virtually no humans actually play the stock markets these days), it is the AIs who have the serious money and therefore are the customers of choice.
If we go down that sort of a road, with spammers and crackers controlling research and development, humans will cease to be a factor in decisions at all. Well, not that we're really making many decisions as it is. Planes are all fly-by-wire, car engines and car navigation is all done by computers, human resources in corporations use computers to automatically select candidates, AIs are now even being used to derive and perform scientific research and make deductions from the results.
If we eliminate the need for humans to process e-mail, then there'll be nothing left for us to do.
Have you ever seen whelk sushi? No? I'll tell you why not. They're vicious bastards that'd kill you as soon as look at you. A whelk once beat Chuck Norris in a fight. Whelks are secretly in control of the Illuminati and the US military.
Yes, installed both from Ubuntu's official site, same distro version, nothing fancy.:) Burned myself with mixing packages with vanilla installers and external repos a few times ion the distant past, so this time did things "properly". Clean install, official methods of doing everything, official packages, the lot. Still barfed.
I've had problems with Firefox 3.0.8 and 3.5 installed at the same time, but that's not strictly a Ubuntu problem unless you consider the failure to de-install the old version when there's a clash a distro issue.
There have been a few other minor glitches here and there, but nothing I'd call substantial. Certainly nothing unexpected for a pre-release.
Some packages long-overdue for updating still haven't been. (A trivial example: ATLAS is at 3.6.0, the official stable version is 3.8.3 and the development version is 3.9.11.) However, that's not going to happen at this stage in the game and it's largely downstream packagers not packaging that's to blame.
Some bugs in the beta (including an odd one that caused the System menu option to sometimes think it was Firefox) have been ironed out and it looks pretty decent.
I think that hinges on why the judges went from "understanding" to "not understanding". If it was because of behind-the-scenes pressure or incentives, then the answer would be no. Tenebaum couldn't win. There's not a whelk's chance in a supernova that anyone could out-bribe the RIAA, and the chances are extremely high they'd be caught, which would fry their chances of winning the "real" case.
If, on the other hand, it's because the judges never understood the legal issues in the first place, then yes. An appeal probably could win. Likewise if the judges deliberately sabotaged their ruling by making it illegal to follow. (One reason a judge might do that is if they are being pressured to render a judgement they disagree with but if they don't they'll be as screwed over as the defendant.)
If it's more that it's completely new territory and the judges are terrified of setting precedents (judges hate doing that), an appeal might go any which way.
Is the judge even required to adhere to an advisory? In other words, by making a ruling that NYCL says isn't legally recognized and is explicitly only "advice", can the judge merely treat it like any "friend of the court" filing rather than a court ruling?
Possibly nothing. "Advisory" would imply that it isn't actually something the court is ruling on but merely offering an opinion on. This is reinforced by NYCL's assertion that "advisories" aren't permitted from a Federal court, suggesting the original court would not be authorized to comply.
On the other hand, possibly everything. If the judges in the appeals court did indeed understand the case and then suddenly lose that understanding, they may have been "leaned on" or were taking backhanders. (I seem to recall a judge pleading guilty to taking bribes from a juvenile detention centre to convict kids just recently. I doubt it's an isolated case.)
There again, since the appeals court acknowledged some dubious elements to the appeal, there may be grounds to take it further, in which case it might mean anything the next lot of judges want it to mean.
In order to maintain parallelism, you could put the acceleration hardware on the ethernet card. If it's in the CPU, unless it's a parallel core (in the same way that the IBM Cell operates), you don't gain any offload advantage.
The problem is, the cats in the universe where you put them live into the box have developed wormhole technology and are going to burst through into this universe and slaughter everyone.
If the cab is out on the roads doing no productive work, merely getting to the state where productive work is possible, then it's generating pollutants that generate smog for no return. It's waste driving. You are bound to have waste driving on any system other than mass transit, but the less waste driving you have, the less pollution you generate. (You can't move more people because that's limited by the number of cabs.)
What's more, the less time spent on the roads, the less the cabs add to traffic congestion, and the biggest fuel consumption killer is stop/start traffic. Reducing stop/start by even a very small amount will reduce the total impact on the environment by a considerably larger amount.
However, this is still peanuts compared to the impact a better mass transit system would have. Manchester in England added a light rail system to the derision of the then-Conservative Government, who called it a total waste of public money. It slashed road usage in the areas it served by a third. That's the kind of waste we could do with more of.
The other major environmental system introduced into Britain was regenerative braking, where wheels are turned into dynamos. Very very effective on the intercity trains, which are hardly slow, and the electricity they put back into the grid does cut the energy needs of the transit system a fair bit.
In electrical cars, it would be trivial to use the same system to put power back into the battery. It won't save a hell of a lot, but it will save some. And the less often you need to recharge the battery, the less energy is needed from the grid, and power stations are themselves polluters.
How about requiring electric and/or hybrid cars (particularly if cabs) be fitted with regenerative braking systems rather than traditional friction-based brakes?
Another way to reduce smog would be to reduce the distance cabs travel. An optimizer that fed the dispatchers with the best way to place unused cabs to minimize probable call times and got closest (rather than random) cabs to calls would help.
If the TARDIS couldn't translate the language of the Beast on the Impossible Planet, what makes you think any human-designed translator could manage the feat?
Better yet, it'd be nice to see SSL used on all pages on all web sites. One of the first rules of security is that context can tell you a lot about what is being encrypted and can potentially weaken that encryption. It also allows attackers to distinguish packets of interest from context.
Using SSL for only critical stuff is like using encryption for only shell passwords. It's better than nothing, but exposes far far too much.
(One might argue that there's so much valuable data placed on computers in corporate DMZ's that further security is pointless until that is fixed. That's true, but one reason corporations don't bother with security is that customers don't demand it. One reason customers don't demand it is that SSL is slow, so sites that don't have good security give a better response, which is what the customer thinks they want. If the response was fixed, customers might start considering sites with competent security preferable to those that effectively hand out bank details to any cracker that asks.)
One thing to consider is whether you need session-based fail-over. If you're going to only treat one as live at a time, then send the spare computer the same packets you are sending the live computer, but drop the responses. If the live computer stops responding, allow the responses from the spare to go through.
The problem will be getting the machines back in sync once the first machine is rebooted. If you assume that the time between the two machines failing is going to be great enough, forward new connections only to the rebooted machine on the basis that all older sessions will terminate before the second machine fails.
...you'd offload the entire TCP/IP stack (Linux' networking isn't the fastest) as well as the SSL. Preferably get the IPSEC in there as well. It shouldn't be too hard to build a card that does the lot. You could then use VCHAN or some other kernel bypass method to forward the data as though Linux had just processed the packets within its own networking stack. The software doesn't need to know where the operation is taking place, so long as the API is the same.
However, just getting the SSL onto a card is a definite advantage, as SSL is a heavy processor consumer and is used frequently-enough that it's a drag on systems.
There are many encryption chips out there (Freescale's S1, for example) and there are projects on OpenCores that you can download right into a low-cost FPGA, so you can get pretty much whatever speed you want at whatever budget you're prepared to set aside.
Slashdot Mirror
But in a virtual reality, there is as much isolation or non-isolation as the simulation cares to provide at that given instance. It is entirely controllable - and entirely measurable. In the physical world, you cannot readily control such data, let alone measure it, and therefore cannot quantify its effects.
The first problem with just accepting "this is how it works" is that classical Hard AI solutions haven't worked, and they all rely on a direct connection.
The second problem with just accepting "this is how it works" is that autism is a consequence of the breakdown of the mechanism you describe, with the severity of the autism being a function of the severity of the breakdown. Autism has direct impact on the brain's capacity to function - more so than just about anything else. There can be quite a bit of brain damage without altering logical thought in the slightest, but even a few haywire mirror neurons can radically alter every thinking process.
Conclusion: The ability to symbolically represent and manipulate a synthetic reality is paramount the intelligence of humans. The ability to process that data is entirely secondary.
Chaos theory is 100% deterministic and 0% predictable. Chaos theory is 100% representable in pure mathematics and requires no emotional component. Chaos theory rules the marketplace. (Always has. It's how fractals were discovered by Mandelbrot, after all.)
Most Slashdot posters are already using the random research paper generator for the articles and the abstract for the replies.
The spammers would add in backdoors that let through the spam they themselves generate.
Roundabouts are superior to traffic lights, in many respects. You don't hold up traffic at all, provided streaming is done right. The biggest problem is when they're used at small, infrequently-used intersections.
Neural nets can do nothing that is non-computable and are not suited to all kinds of problems. Petri nets are also quite interesting, but again have a very specific role in AI.
It has been shown that a single neuron from a physical brain can perform extremely complex operations. How is, as far as I know, unknown. However, it means that the most advanced computer neural nets are not yet as advanced as single organic neurons, and organic brains that have anything approximating intelligence have billions of organic neurons.
In other words, if you gave everyone on the planet an Origin 3000 and a terabit pipe, and ran all those computers as a single distributed neural net program, you would have a computer with the mental capacity of an African Grey parrot.
Now, clearly that much compute power is capable of doing FAR more work and is capable of vastly superior intellect, from which we can conclude that the problem lies not in the computers but in the techniques used. We simply don't have the software tools necessary for quality AI. At least, not by using any approach so far adopted.
My personal theory is that AI will evolve out of virtual reality because the human brain does not interact directly with the senses or muscles but via an internal simulation. All the physical world stuff is handled by external plugins to that internal VR. That seems simple to reproduce.
Since intelligence is merely an evolutionary byproduct of how that internal VR is handled, and as computers can run as many parallel handlers as there are processors thrown at the problem, I don't see that it's necessary for humans to ever solve how to program AI. Give it a self-contained MMORG and let it herustically develop itself.
Since AIs are, almost by definition, more predictable than humans, it is self-evident that AI customers can be more cost-effectively be tailored for and more easily swayed. Since limited intelligences already handle most financial decisions (virtually no humans actually play the stock markets these days), it is the AIs who have the serious money and therefore are the customers of choice.
If we go down that sort of a road, with spammers and crackers controlling research and development, humans will cease to be a factor in decisions at all. Well, not that we're really making many decisions as it is. Planes are all fly-by-wire, car engines and car navigation is all done by computers, human resources in corporations use computers to automatically select candidates, AIs are now even being used to derive and perform scientific research and make deductions from the results.
If we eliminate the need for humans to process e-mail, then there'll be nothing left for us to do.
Have you ever seen whelk sushi? No? I'll tell you why not. They're vicious bastards that'd kill you as soon as look at you. A whelk once beat Chuck Norris in a fight. Whelks are secretly in control of the Illuminati and the US military.
Yes, installed both from Ubuntu's official site, same distro version, nothing fancy. :) Burned myself with mixing packages with vanilla installers and external repos a few times ion the distant past, so this time did things "properly". Clean install, official methods of doing everything, official packages, the lot. Still barfed.
I've had problems with Firefox 3.0.8 and 3.5 installed at the same time, but that's not strictly a Ubuntu problem unless you consider the failure to de-install the old version when there's a clash a distro issue.
There have been a few other minor glitches here and there, but nothing I'd call substantial. Certainly nothing unexpected for a pre-release.
Some packages long-overdue for updating still haven't been. (A trivial example: ATLAS is at 3.6.0, the official stable version is 3.8.3 and the development version is 3.9.11.) However, that's not going to happen at this stage in the game and it's largely downstream packagers not packaging that's to blame.
Some bugs in the beta (including an odd one that caused the System menu option to sometimes think it was Firefox) have been ironed out and it looks pretty decent.
I think that hinges on why the judges went from "understanding" to "not understanding". If it was because of behind-the-scenes pressure or incentives, then the answer would be no. Tenebaum couldn't win. There's not a whelk's chance in a supernova that anyone could out-bribe the RIAA, and the chances are extremely high they'd be caught, which would fry their chances of winning the "real" case.
If, on the other hand, it's because the judges never understood the legal issues in the first place, then yes. An appeal probably could win. Likewise if the judges deliberately sabotaged their ruling by making it illegal to follow. (One reason a judge might do that is if they are being pressured to render a judgement they disagree with but if they don't they'll be as screwed over as the defendant.)
If it's more that it's completely new territory and the judges are terrified of setting precedents (judges hate doing that), an appeal might go any which way.
Is the judge even required to adhere to an advisory? In other words, by making a ruling that NYCL says isn't legally recognized and is explicitly only "advice", can the judge merely treat it like any "friend of the court" filing rather than a court ruling?
Possibly nothing. "Advisory" would imply that it isn't actually something the court is ruling on but merely offering an opinion on. This is reinforced by NYCL's assertion that "advisories" aren't permitted from a Federal court, suggesting the original court would not be authorized to comply.
On the other hand, possibly everything. If the judges in the appeals court did indeed understand the case and then suddenly lose that understanding, they may have been "leaned on" or were taking backhanders. (I seem to recall a judge pleading guilty to taking bribes from a juvenile detention centre to convict kids just recently. I doubt it's an isolated case.)
There again, since the appeals court acknowledged some dubious elements to the appeal, there may be grounds to take it further, in which case it might mean anything the next lot of judges want it to mean.
In order to maintain parallelism, you could put the acceleration hardware on the ethernet card. If it's in the CPU, unless it's a parallel core (in the same way that the IBM Cell operates), you don't gain any offload advantage.
It fell behind the sofa.
Poyntee ends can eggsplayn anyfing tu anywun beri beri eezily. Paynfully but eezily.
The problem is, the cats in the universe where you put them live into the box have developed wormhole technology and are going to burst through into this universe and slaughter everyone.
In quantum mechanics, you cannot be both wishful and thinking at the same time.
In yet other news: A recent study by the Popular Mechanics Readers' Society, research into quantum mechanics causes cancer in rats.
If the cab is out on the roads doing no productive work, merely getting to the state where productive work is possible, then it's generating pollutants that generate smog for no return. It's waste driving. You are bound to have waste driving on any system other than mass transit, but the less waste driving you have, the less pollution you generate. (You can't move more people because that's limited by the number of cabs.)
What's more, the less time spent on the roads, the less the cabs add to traffic congestion, and the biggest fuel consumption killer is stop/start traffic. Reducing stop/start by even a very small amount will reduce the total impact on the environment by a considerably larger amount.
However, this is still peanuts compared to the impact a better mass transit system would have. Manchester in England added a light rail system to the derision of the then-Conservative Government, who called it a total waste of public money. It slashed road usage in the areas it served by a third. That's the kind of waste we could do with more of.
The other major environmental system introduced into Britain was regenerative braking, where wheels are turned into dynamos. Very very effective on the intercity trains, which are hardly slow, and the electricity they put back into the grid does cut the energy needs of the transit system a fair bit.
In electrical cars, it would be trivial to use the same system to put power back into the battery. It won't save a hell of a lot, but it will save some. And the less often you need to recharge the battery, the less energy is needed from the grid, and power stations are themselves polluters.
How about requiring electric and/or hybrid cars (particularly if cabs) be fitted with regenerative braking systems rather than traditional friction-based brakes?
Another way to reduce smog would be to reduce the distance cabs travel. An optimizer that fed the dispatchers with the best way to place unused cabs to minimize probable call times and got closest (rather than random) cabs to calls would help.
If the TARDIS couldn't translate the language of the Beast on the Impossible Planet, what makes you think any human-designed translator could manage the feat?
Better yet, it'd be nice to see SSL used on all pages on all web sites. One of the first rules of security is that context can tell you a lot about what is being encrypted and can potentially weaken that encryption. It also allows attackers to distinguish packets of interest from context.
Using SSL for only critical stuff is like using encryption for only shell passwords. It's better than nothing, but exposes far far too much.
(One might argue that there's so much valuable data placed on computers in corporate DMZ's that further security is pointless until that is fixed. That's true, but one reason corporations don't bother with security is that customers don't demand it. One reason customers don't demand it is that SSL is slow, so sites that don't have good security give a better response, which is what the customer thinks they want. If the response was fixed, customers might start considering sites with competent security preferable to those that effectively hand out bank details to any cracker that asks.)
One thing to consider is whether you need session-based fail-over. If you're going to only treat one as live at a time, then send the spare computer the same packets you are sending the live computer, but drop the responses. If the live computer stops responding, allow the responses from the spare to go through.
The problem will be getting the machines back in sync once the first machine is rebooted. If you assume that the time between the two machines failing is going to be great enough, forward new connections only to the rebooted machine on the basis that all older sessions will terminate before the second machine fails.
We did the particle accelerator some time back. Do a search for "scotch tape" and "x-rays".
...you'd offload the entire TCP/IP stack (Linux' networking isn't the fastest) as well as the SSL. Preferably get the IPSEC in there as well. It shouldn't be too hard to build a card that does the lot. You could then use VCHAN or some other kernel bypass method to forward the data as though Linux had just processed the packets within its own networking stack. The software doesn't need to know where the operation is taking place, so long as the API is the same.
However, just getting the SSL onto a card is a definite advantage, as SSL is a heavy processor consumer and is used frequently-enough that it's a drag on systems.
There are many encryption chips out there (Freescale's S1, for example) and there are projects on OpenCores that you can download right into a low-cost FPGA, so you can get pretty much whatever speed you want at whatever budget you're prepared to set aside.