Educated adults are dismissive of children's culture because educated adults know much more than children, and understand the world much better than children.
So, an educated adult (as opposed to an overgrown child) will rightly be dismissive of mass culture targeted at children. Such culture is almost always very superficial, and devoid of deeper, symbolic meaning such as one routinely finds in the culture of educated adults.
This is one reason that the earlier Disney films were so successful - they were *not* merely mass kid culture, but were based on myths/fairy tales (Sleeping Beauty, Beauty and the Beast, Snow White) that are actually deeper symbolic tales that are kept alive by re-telling them to children each generation; or based on books that were only *seemingly* children's stories (The Jungle Book, Alice in Wonderland) but are actually very adult tales disguised as works for children.
" The human psychology seems to prefer working a full day, so while we really could probably be working only 1 hour a day and still maintaining sustenance levels of production, we are still working 8 (or more!) hours for a nice lifestyle."
No. People prefer to work 2 - 4 hours per day, not 8. We know this because there are still societies where people do live by subsistence, and, when they have the choice to do so, they will work a few hours a day (2-4), and spend the rest of their time in other pursuits (decorative crafts, gossiping, dance, etc.).
People only work 8 hours/day in the industrialized world because, at the low end of the pay scale, one actually needs to work *more* than 8 hours/day in order to make a living. Higher up the pay scale, there are no real options for high paying jobs where your employer will let you work just 1 hour a day.
Not that I believe that the economy is a zero-sum-game, but be aware that commonly used models do not account for the real cost of the use of non-renewable resources.
Some examples:
Analyses of the petroleum industry do not account the cost of depleting a non-renewable resource, just the cost of its extraction. This is a false accounting. In the future, we will pay the cost of exhausting this resource. That cost may be great (global economic dislocation, war, famine, disease) or it may be small (we find a completely substitutable replacement for petroleum), but our economics almost never does a real accounting, because certain costs are simply not considered.
Analyses of the fishing industry do not account the cost of driving certain food species to levels where it is no longer practical to fish them. Again, these costs may be great, or small, but they do not figure in economic analyses.
Even where we can make some estimate of the cost, such as the cost in terms of health care of air pollution, economists do not count these costs against the industries that generate the pollution.
Don't pretend that economics is anything like mathematics in its completeness or rigor. It is simply an elaboration of book keeper's accounting, which only considers costs that the enterprise itself must pay. We know, of course, that the real costs are often far greater than this, but economists have not been trained to model costs in anything approaching a complete manner.
Nice try, but you forget some basic requirements of physics. Once built, the robot will not "be able to produce essentially free labor for an indefinite amount of time." Rather, it will produce labor which will cost whatever the energy to run the robot costs, plus whatever consumables it requires (lubrication, replacement parts), plus maintenance costs.
So, if
(purchase price of robot + lifetime energy cost + lifetime consumables cost + lifetime maintenance costs) / lifetime of robot (in hours)
is greater than the hourly wage of the laborer it would replace, then the robot is more expensive than the laborer.
No, they're pulled out of his ass. Kurzweil makes the same conflation of CPU power with understanding that this article does.
It really doesn't matter if you have a CPU with the same processing power as a human brain, because now you have to have software to run on this CPU - software that is as complex, error tolerant, and open-ended as a human mind.
It will be no trick to build a CPU with the raw processing power of a human brain within the time frame postulated by Kurzweil.
It will, however, be an incredibly difficult problem to understand the complexity of the interconnections of the brain, and how these connections *implicity* define a myriad of complex, real-time, interacting, fault tolerant, algorithms that we call a mind, capable of sensation, perception, emotion, thought, and learning.
Large software projects to do diddly things like airline reservation systems routinely run years behind schedule and millions over budget. And we're going to hack out software to emulate a human mind in 40 odd years? Yeah, right.
No, you forgot choice number 3, which is to burn excess protein for energy. This is an inherently inefficient process (for human metabolism anyway) and this is part of why one loses weight when eating a high protein, low carbohydrate diet - your body can't burn protein for energy as efficiently as it can use carbs, so the net result is that body fat, as well as lean body tissue (muscle) is burned to make up the energy deficit.
No, actually, this is not what the studies show. Studies show that students who get fewer than 7 hours of sleep the night before a test do decidedly worse than those who get 7 or more hours of sleep. This is especially true of younger students (read, high school age).
Long term memories -which is what you need to lay down for test taking - are apparently finalized in REM sleep, much of which takes place in the last sleep cycle before waking - i.e., the last 90 minutes of a 7-8 hour night of sleep. Most of early sleep is deep sleep, when body repair takes place. Most of late sleep is REM sleep, where dreaming takes place, and, apparently, long term memories from the previous day's experiences are laid down. This division is true both of the individual sleep cycles - deep first, REM last - and of the course of a night's sleep as a whole. Early sleep cycles consist of mostly deep sleep, and later sleep cycles consist of proportionately more REM sleep.
So, if you miss that last hour of needed sleep, your body is repaired, but your memory will suffer. Remember this the next time you take a test. If you need to get up at 8:00, when midnight rolls around, you'd be *much* better off going to sleep, than studying to 1:30 am, and trying to make do on 6 1/2 hours of sleep.
Re:your comment was 'easily breakable'
on
Inkblot Passwords
·
· Score: 3, Insightful
No, all a cracker would need to do is to test the permutations of the most likely variant responses *first*. The cracker would need to know *nothing* about the individual user, just what responses were most common statistically. Even if such knowledge consisted entirely of what words people use most often in short descriptive phrases (independent of ink blots), it would shrink the search space dramatically.
Combined with the fact that the cracker is dealing only with alphabetic characters, you end up with a highly structured system, with an obvious, and likely quite fruitful, means of attack.
Structural regularity leads to easy line of attack
on
Inkblot Passwords
·
· Score: 2, Interesting
Read the article - they use the first *and* last letter, so the line you quoted from Macbeth becomes:
wnslwetemtanintrlgorrn
Which points up a flaw in the system that a previous poster alluded to, namely, that you end up with only alphanumeric character passwords, so a cracker program would only need to run permutations of first/last letter pairs from a dictionary to crack these passwords.
Moreover, there are undoubtedly some first/last letter combinations that are more common than others in english, even for multi-word phrases, so the crackers would try these first in their search.
In other words, their very structural regularity leads to an easy line of attack.
OS X is Mach with a BSD "personality," i.e., it has a BSD layer, but Mac OS X GUI software, like ManOpen, is not required to, not likely to, rely entirely on the BSD layer. This means that Mac OS X GUI software will rarely, if ever, "just run" on some other *nix, or *bsd. The only exceptions are Mac OS X applications that are intended to run under a Mac OS X X Window server.
From the ReadMe that comes with ManOpen: "ManOpen should run on MacOS X 10.x, MacOS X Server 1.2, and OPENSTEP/Mach machines. Source code is available."
So you could try porting it to the X Window system on you favorite *nix if you like.
The Apple Music Store does something very like this. It looks at what you listen to (30 second excerpts) and buy, and it puts a header on the search page with suggestions for other artists and albums you might like, based on your search results, and the listening/buying habits of the entire web of listeners (several million).
This scheme is easily broken, and everyone sharing identified. All the RIAA would have to do is get a single paid informant into the web of sharers. The RIAA installs the necessary IP logging software on his machine, downloads a bunch of files, and everyone in this "secret" interconnected in-group is sued by the RIAA.
... I mean, 2 words: Lighten up, buddy...... I mean, 3 words: Lighten up some, buddy...... I mean, 4 words: Amongst our weapons are, fear, surprise, and an almost fanatical devotion to the Pope!
No, actually, Metroworks CodeWarrior produces significantly faster code (i.e, the generated code runs faster).
But Apple wanted to do a test on a *level* field. That means both CPUs running the same code, to see which CPU is faster. The closest you can come to this is to use:
1. the same source code (spec) 2. the same compiler (gcc 3.3)
Since Intel's compiler is not available for the PPC, Apple used gcc.
Interestingly, Metrowerks/Motorola's compiler (i.e., CodeWarrior - Motorola bought Metrowerks a year or two ago) *is* available for both PPC and Intel, but Metroworks/Motorola may not have completed their G5 specific back end yet.
spec tests processors, and processors only, with other factors (like compilers) held constant. If you want to test real world performance (as in your analogy of the cars and fuels) you should test real applications, running on both platforms, side by side... Wait, Apple already did that, and the G5 smoked both the P4 and the dual xeon boxes.
Multiple processors will only increase latency when they get in each other's way. The whole point of the G5 is that it is an entirely new architecture (for Apple, not IBM) in which *each* of the Dual procs has its own 1/2 speed (1GHz for the dual 2GHz machine) bus.
Which would be true if you really wiped your drive, and *never* dual booted the machine into Windows, but how many Linux/*BSD users *really* do that for a desktop machine (i.e., *not* a server)? Not many, I'll warrant.
Oh, please. Methinks the Intel zealot doth protest too much. If you *really* didn't give a shit, you wouldn't be posting about the issue here on Slashdot, would you?
Be real. You obviously care. You have much time, money, and expertise bound to the intel compatible platform. Like everyone else with a brain, you know that Mac OS X is a superior OS to either Windows or the various other OSs that run on intel (some of which also run on the Mac anyway).
So it is a bit frightening to contemplate the possibility that you have spent all those many hours backing the wrong horse, i.e., the one that can run the dog OS, Windows, but can't run the sweet OS, Mac OS X.
Well, when we talk about "all optimizations on" we're really talking about real world application performance, since software vendors determine what "all optimizations on" means in practice for the software we actually use most of the time.
In this case, that means an independent comparison of Photoshop, 3D rendering, etc.
Apple's Photoshop tests suggests that the two platforms are at least about on par now, which is all that matters. This simply means that one can now choose a platform based on OS, not on performance.
That being the case, I would have no hesitation to buy a high end G5, whether or not it is slightly faster, or slightly slower than the current top of the line intel box, because I'd much rather be using Mac OS X than Windows (Linux is not an issue since it runs on both platforms).
There is if you want the benchmark to be a measure of processor performance rather than a measure of compiler performance.
Didn't you people learn anything about controlling variables in school? You have to fix the others (compiler, RAM) so that you can have a meaningful comparison of the one in question - CPU.
I'm glad I don't have you doing lab work for me. The results from one test to the next would be completely meaningless.
No, we wouldn't, because we can't control Intel's compiler, or Motorola's, so we don't know if each has spec specific optimizaitons which render the benchmarks worthless.
However, we *do* know who wins an unbiased comparison, using the same, open source compiler. And we also *do* know who wins real world performance tests with BLAST and Photoshop - the G5.
No, Apple quotes comparable runs, meaning, runs compiled with the same open source compiler, not some proprietary benchmark generator with specific optimizations for spec.
The whole idea of benchmarking is to make everything as much the same as possible (i.e., controlling variables) so that the variable of interest, here, the CPUs, can be compared on a level field. That's exactly what Apple have done.
If you want instead to look at real world performance, not useless benchmarking, see the Photoshop comparisons, or the BLAST comparisons, where the G5 also handily trounces the Wintel boxen.
Slashdot Mirror
Educated adults are dismissive of children's culture because educated adults know much more than children, and understand the world much better than children.
So, an educated adult (as opposed to an overgrown child) will rightly be dismissive of mass culture targeted at children. Such culture is almost always very superficial, and devoid of deeper, symbolic meaning such as one routinely finds in the culture of educated adults.
This is one reason that the earlier Disney films were so successful - they were *not* merely mass kid culture, but were based on myths/fairy tales (Sleeping Beauty, Beauty and the Beast, Snow White) that are actually deeper symbolic tales that are kept alive by re-telling them to children each generation; or based on books that were only *seemingly* children's stories (The Jungle Book, Alice in Wonderland) but are actually very adult tales disguised as works for children.
" The human psychology seems to prefer working a full day, so while we really could probably be working only 1 hour a day and still maintaining sustenance levels of production, we are still working 8 (or more!) hours for a nice lifestyle."
No. People prefer to work 2 - 4 hours per day, not 8. We know this because there are still societies where people do live by subsistence, and, when they have the choice to do so, they will work a few hours a day (2-4), and spend the rest of their time in other pursuits (decorative crafts, gossiping, dance, etc.).
People only work 8 hours/day in the industrialized world because, at the low end of the pay scale, one actually needs to work *more* than 8 hours/day in order to make a living. Higher up the pay scale, there are no real options for high paying jobs where your employer will let you work just 1 hour a day.
Not that I believe that the economy is a zero-sum-game, but be aware that commonly used models do not account for the real cost of the use of non-renewable resources.
Some examples:
Analyses of the petroleum industry do not account the cost of depleting a non-renewable resource, just the cost of its extraction. This is a false accounting. In the future, we will pay the cost of exhausting this resource. That cost may be great (global economic dislocation, war, famine, disease) or it may be small (we find a completely substitutable replacement for petroleum), but our economics almost never does a real accounting, because certain costs are simply not considered.
Analyses of the fishing industry do not account the cost of driving certain food species to levels where it is no longer practical to fish them. Again, these costs may be great, or small, but they do not figure in economic analyses.
Even where we can make some estimate of the cost, such as the cost in terms of health care of air pollution, economists do not count these costs against the industries that generate the pollution.
Don't pretend that economics is anything like mathematics in its completeness or rigor. It is simply an elaboration of book keeper's accounting, which only considers costs that the enterprise itself must pay. We know, of course, that the real costs are often far greater than this, but economists have not been trained to model costs in anything approaching a complete manner.
Nice try, but you forget some basic requirements of physics. Once built, the robot will not "be able to produce essentially free labor for an indefinite amount of time." Rather, it will produce labor which will cost whatever the energy to run the robot costs, plus whatever consumables it requires (lubrication, replacement parts), plus maintenance costs.
So, if
(purchase price of robot + lifetime energy cost + lifetime consumables cost + lifetime maintenance costs) / lifetime of robot (in hours)
is greater than the hourly wage of the laborer it would replace, then the robot is more expensive than the laborer.
No, they're pulled out of his ass. Kurzweil makes the same conflation of CPU power with understanding that this article does.
It really doesn't matter if you have a CPU with the same processing power as a human brain, because now you have to have software to run on this CPU - software that is as complex, error tolerant, and open-ended as a human mind.
It will be no trick to build a CPU with the raw processing power of a human brain within the time frame postulated by Kurzweil.
It will, however, be an incredibly difficult problem to understand the complexity of the interconnections of the brain, and how these connections *implicity* define a myriad of complex, real-time, interacting, fault tolerant, algorithms that we call a mind, capable of sensation, perception, emotion, thought, and learning.
Large software projects to do diddly things like airline reservation systems routinely run years behind schedule and millions over budget. And we're going to hack out software to emulate a human mind in 40 odd years? Yeah, right.
No, you forgot choice number 3, which is to burn excess protein for energy. This is an inherently inefficient process (for human metabolism anyway) and this is part of why one loses weight when eating a high protein, low carbohydrate diet - your body can't burn protein for energy as efficiently as it can use carbs, so the net result is that body fat, as well as lean body tissue (muscle) is burned to make up the energy deficit.
No, actually, this is not what the studies show. Studies show that students who get fewer than 7 hours of sleep the night before a test do decidedly worse than those who get 7 or more hours of sleep. This is especially true of younger students (read, high school age).
Long term memories -which is what you need to lay down for test taking - are apparently finalized in REM sleep, much of which takes place in the last sleep cycle before waking - i.e., the last 90 minutes of a 7-8 hour night of sleep. Most of early sleep is deep sleep, when body repair takes place. Most of late sleep is REM sleep, where dreaming takes place, and, apparently, long term memories from the previous day's experiences are laid down. This division is true both of the individual sleep cycles - deep first, REM last - and of the course of a night's sleep as a whole. Early sleep cycles consist of mostly deep sleep, and later sleep cycles consist of proportionately more REM sleep.
So, if you miss that last hour of needed sleep, your body is repaired, but your memory will suffer. Remember this the next time you take a test. If you need to get up at 8:00, when midnight rolls around, you'd be *much* better off going to sleep, than studying to 1:30 am, and trying to make do on 6 1/2 hours of sleep.
No, all a cracker would need to do is to test the permutations of the most likely variant responses *first*. The cracker would need to know *nothing* about the individual user, just what responses were most common statistically. Even if such knowledge consisted entirely of what words people use most often in short descriptive phrases (independent of ink blots), it would shrink the search space dramatically.
Combined with the fact that the cracker is dealing only with alphabetic characters, you end up with a highly structured system, with an obvious, and likely quite fruitful, means of attack.
Read the article - they use the first *and* last letter, so the line you quoted from Macbeth becomes:
wnslwetemtanintrlgorrn
Which points up a flaw in the system that a previous poster alluded to, namely, that you end up with only alphanumeric character passwords, so a cracker program would only need to run permutations of first/last letter pairs from a dictionary to crack these passwords.
Moreover, there are undoubtedly some first/last letter combinations that are more common than others in english, even for multi-word phrases, so the crackers would try these first in their search.
In other words, their very structural regularity leads to an easy line of attack.
No, species follows genus, as the grandparent humorous post suggested.
Kingdom Phylum Class Order Family Genus Species
("King Phillip Crossed Over From Grant Street" is the usual mnemonic).
OS X is Mach with a BSD "personality," i.e., it has a BSD layer, but Mac OS X GUI software, like ManOpen, is not required to, not likely to, rely entirely on the BSD layer. This means that Mac OS X GUI software will rarely, if ever, "just run" on some other *nix, or *bsd. The only exceptions are Mac OS X applications that are intended to run under a Mac OS X X Window server.
From the ReadMe that comes with ManOpen:
"ManOpen should run on MacOS X 10.x, MacOS X Server 1.2, and OPENSTEP/Mach machines. Source code is available."
So you could try porting it to the X Window system on you favorite *nix if you like.
The Apple Music Store does something very like this. It looks at what you listen to (30 second excerpts) and buy, and it puts a header on the search page with suggestions for other artists and albums you might like, based on your search results, and the listening/buying habits of the entire web of listeners (several million).
This scheme is easily broken, and everyone sharing identified. All the RIAA would have to do is get a single paid informant into the web of sharers. The RIAA installs the necessary IP logging software on his machine, downloads a bunch of files, and everyone in this "secret" interconnected in-group is sued by the RIAA.
... I mean, 2 words: Lighten up, buddy... ... I mean, 3 words: Lighten up some, buddy... ... I mean, 4 words: Amongst our weapons are, fear, surprise, and an almost fanatical devotion to the Pope!
... Steve is the square root of Greg.
No, actually, Metroworks CodeWarrior produces significantly faster code (i.e, the generated code runs faster).
But Apple wanted to do a test on a *level* field. That means both CPUs running the same code, to see which CPU is faster. The closest you can come to this is to use:
1. the same source code (spec)
2. the same compiler (gcc 3.3)
Since Intel's compiler is not available for the PPC, Apple used gcc.
Interestingly, Metrowerks/Motorola's compiler (i.e., CodeWarrior - Motorola bought Metrowerks a year or two ago) *is* available for both PPC and Intel, but Metroworks/Motorola may not have completed their G5 specific back end yet.
spec tests processors, and processors only, with other factors (like compilers) held constant. If you want to test real world performance (as in your analogy of the cars and fuels) you should test real applications, running on both platforms, side by side...
Wait, Apple already did that, and the G5 smoked both the P4 and the dual xeon boxes.
Multiple processors will only increase latency when they get in each other's way. The whole point of the G5 is that it is an entirely new architecture (for Apple, not IBM) in which *each* of the Dual procs has its own 1/2 speed (1GHz for the dual 2GHz machine) bus.
Which would be true if you really wiped your drive, and *never* dual booted the machine into Windows, but how many Linux/*BSD users *really* do that for a desktop machine (i.e., *not* a server)? Not many, I'll warrant.
Oh, please. Methinks the Intel zealot doth protest too much. If you *really* didn't give a shit, you wouldn't be posting about the issue here on Slashdot, would you?
Be real. You obviously care. You have much time, money, and expertise bound to the intel compatible platform. Like everyone else with a brain, you know that Mac OS X is a superior OS to either Windows or the various other OSs that run on intel (some of which also run on the Mac anyway).
So it is a bit frightening to contemplate the possibility that you have spent all those many hours backing the wrong horse, i.e., the one that can run the dog OS, Windows, but can't run the sweet OS, Mac OS X.
Well, when we talk about "all optimizations on" we're really talking about real world application performance, since software vendors determine what "all optimizations on" means in practice for the software we actually use most of the time.
In this case, that means an independent comparison of Photoshop, 3D rendering, etc.
Apple's Photoshop tests suggests that the two platforms are at least about on par now, which is all that matters. This simply means that one can now choose a platform based on OS, not on performance.
That being the case, I would have no hesitation to buy a high end G5, whether or not it is slightly faster, or slightly slower than the current top of the line intel box, because I'd much rather be using Mac OS X than Windows (Linux is not an issue since it runs on both platforms).
The spec scores you cite from the spec site use a different compiler which is only availible for intel, and probably optimzed for the spec benchmark.
Apple have used an open source compiler - GCC 3.3 - available for both platforms.
IOW, Apple's benchmark is a fair comparison of CPU speed, not a biased comparison of proprietary, benchmark-optimzed compilers.
And in the real world, the G5 spanks the fastest Wintel PCs in Photoshop and BLAST.
Parent post is a troll. There's no way on earth that a 17 Meg file takes 20 minutes to copy. That's simply a troll's lie.
There is if you want the benchmark to be a measure of processor performance rather than a measure of compiler performance.
Didn't you people learn anything about controlling variables in school? You have to fix the others (compiler, RAM) so that you can have a meaningful comparison of the one in question - CPU.
I'm glad I don't have you doing lab work for me. The results from one test to the next would be completely meaningless.
No, we wouldn't, because we can't control Intel's compiler, or Motorola's, so we don't know if each has spec specific optimizaitons which render the benchmarks worthless.
However, we *do* know who wins an unbiased comparison, using the same, open source compiler. And we also *do* know who wins real world performance tests with BLAST and Photoshop - the G5.
No, Apple quotes comparable runs, meaning, runs compiled with the same open source compiler, not some proprietary benchmark generator with specific optimizations for spec.
The whole idea of benchmarking is to make everything as much the same as possible (i.e., controlling variables) so that the variable of interest, here, the CPUs, can be compared on a level field. That's exactly what Apple have done.
If you want instead to look at real world performance, not useless benchmarking, see the Photoshop comparisons, or the BLAST comparisons, where the G5 also handily trounces the Wintel boxen.