You could ask the governor of Tokyo, Shintaro Ishihara, when he said the following:
In due course, the perpetrators were captured, and, just as had been suspected, the crime was one of revenge among Chinese criminals. There is fear–and not without cause–that it will not be long before the entire nature of Japanese society itself will be altered by the spread of this type of crime that is indicative of the ethnic DNA [of the Chinese].
No, of course we can study and discuss potential (and real) differences between arbitrary groups of people. And of course there's no need for secrecy or disclaimers.
The problem comes when Japanese see this work and say "Aha! Foreigners can't eat sushi! Therefore it's OK for me to ban foreigners from my sushi restaurant since they can't eat it anyway." And if that logic sounds ridiculous or impossible, you've obviously never lived in Japan.
My point is that people should be conscious of how their research will be used. In this case, it would have been better for the authors to do a more in-depth study with more subjects, particularly Koreans (who eat large amounts of seaweed) or other areas with heavily fish based cuisine.
From my experience in Japan, the 'natives' love it when I try and speak Japanese when communicating. I've never sensed an ounce of condescending tones when I struggle.
Certainly they like it when you struggle speaking Japanese, it reinforces the idea that foreigners can't speak the language. When you reach native fluency they're much less friendly (speaking from years of experience here).
Just what we need, more "Japanese are unique" idiocy to justify racism and discrimination in Japan. So far we've heard that "Japanese intestines are longer, so Japanese can't eat foreign beef", "Japanese brains are unique, so only Japanese people can speak the Japanese language." and so on, all of which are supported by pseudo-scientific studies such as this one.
This sort of incomplete research just feeds the view of racial uniqueness (and superiority) among Japanese and justifies their racism and discrimination against others.
This raises an interesting point. Maybe in the brain there's no way to reduce things to meaningful equations. Evolution didn't really have a set goal in mind, so what you get is a cobbled together mess of pieces that works. The reason it works is because the system can reform and grow itself in response to new input, not because it has a lot of pieces.
These large scale simulations are interesting from a neurobiological point of view. But if you want to build a "thinking computer brain", you might be better off understanding how the human brain grows and changes in response to stimuli, rather than trying to make a static copy at the cellular level. I have a feeling these supercomputer simulations are not going to be the path to conscious computers, nor even to mind uploading.
I'm guessing you don't live in Japan, because if you did you'd probably think otherwise. Unless you like hearing endless parades of megaphone blasting vans, with high pitched voice women screaming nothing but "Please vote for Tanaka! Thank you! Please vote for Tanaka! Thank you!".
And if you want to talk about judging politicians by their hair, former Prime Minister Koizumi was known as "Lionheart" because of his hairstyle. Most Japanese I've spoken to identify him merely by his hair, and can't name a single policy he enacted. In almost all respects, I'd say the Japanese system is a good example of what to avoid rather than emulate.
I speak Japanese fluently. I have lived in Japan for about 6 years, currently enrolled at one of the "Top 3" public universities. I can read a newspaper in Japanese. I have given lectures in Japanese on theoretical computer science. I pay my taxes, I renew my visa on time. I don't do any of the other illegal or questionable things on your list. Having said that, the grandparent is completely correct - Japan and Japanese are extremely racist, xenophobic and nationalistic. That's part of the reason why I'm leaving as soon as I finish my degree. You might think (as I did) if you learn the language and culture you'll be accepted - you're wrong. In Japan you will always be judged and discriminated against because you are not Japanese. It doesn't matter if it's in the city vs. countryside, company vs. university, public vs. private. If you can tolerate being treated as a second class citizen, being denied housing/employment/services because of your skin color, being treated as an idiot incapable of understanding even basic Japanese, then feel free to come to Japan. If not, then avoid it.
> For FFT, the Cell processor in PS3 performs 100 times faster than Pentium 4 in some tests, if properly configured.
This is exactly my point. Individually the processor may perform well, but when it's placed in the actual system, perform will undoubtedly drop. Right now, I'm doing performance tests on FFTs performed on GPUs (graphics cards). Theoretically, these should perform at the same "incredible" speed as the Cell processor (10 Gflops or better), but in reality bandwidth and cache constricts performance to half a Gflop.
You'll also note they say nothing about the system they tested on. Was it a PS3? I bet you it wasn't. Let's wait and see the actual performance of the PS3 before we get excited from in house tests done on an unspecified system by a company that is eager to boost impressions of its new chip.
I'm not entirely convinced the PS3 will be better than standard desktops, at least at non-gaming activities (I'm sure the PS3 will win as far as gaming goes). It seems if Sony could make a $300 console that outperforms a $1000 desktop, they'd be selling them as $1000 desktops instead of $300 consoles. But I guess we'll see when they come out.
At least with current platforms architectures. The author seems to do plenty of research on current distributed computing projects, but does none on how the consoles perform.
I know that SETI@home has been ported and tested at least on the XBox, and it performs miserably. These console gaming systems are designed to play games, not do radio signal analysis or other scientific calculation. For example, there's little need for fast memory writing when you're mostly reading textures from RAM, but there's an extreme need when you do millions of in-place Fourier transforms. Unless Microsoft and Sony change their architectures for some inexplicable reason, I can't imagine future architectures would perform much better.
This article smacks of ignorance on the part of the author, who clearly did no research into the actual performance of consoles in regard to standard scientific computing.
"The labor force of a country is different than the number of citizens."
My numbers for the US are presented in the chart I mentioned in the grandparent post. Granted these don't take into account migrant labor, but I'm pretty sure the percentage of migrant/illegal labor in the US is nowhere near comparable to Luxembourg (which is about 30%).
Basically, my point was that PCGDP does not equal productivity. I think that was apparent by my first or second post.
"For national productivity statistics, an obvious starting point is to..."
Meaning GDP is not the entire answer.
"...to measure labor input, such as the number of workers..."
Like I said before, the population of a country is not equal to the number of workers. This chart (which is actually explicitly labelled "Overall Productivity") takes this fact into account:
http://www.nationmaster.com/graph-T/eco_ove_pro_pp p
On this chart, Norway falls to #6 and the banking countries drop off the top 25. This graph (which is not PCGDP) is a much more accurate depiction of productivity.
"Labor Force: 200,000 (of whom 87,400 are foreign cross-border workers primarily from France, Belgium, and Germany) (2003)" (From CIA World Factbook)
The only reason Luxembourg is so high is because one third of their work force comes from France, Belgium and Germany. These people contribute to output (raising productivity), but are not counted in the Luxembourg work force (which would lower productivity).
Therefore, your original statement that the US is not the most productive country in the world is incorrect. Your original statement that productivity equals PCGDP is also incorrect. If it were correct, the above chart would have the same rankings as the PCGDP chart.
No, you're missing the point, and I could make an equally intelligent response by questioning your English reading ability.
The CIA never said that our productivity is not the best. The CIA said that our PCGDP is not the best (I won't argue with them there). But the CIA *never said* that PCGDP = productivity. That's what you said, and what I argued against.
In your original post you said that "The myth that American workers are the most productive (Per Capita GDP) persists". You equated productivity with PCGDP. I said your interpretation of PCGDP was wrong, not the figures for PCGDP.
Actually I was paying attention in Economics 101, which is why I asked you the question. And your quote does not support your argument.
From your own article: "Simply stated, productivity is output per unit of input." Where does PCGDP take input into account? Where does PCGDP take product accounts into account? Where is the labor and capital stock input in PCGDP? The labor force of a country is different than the number of citizens.
Basically, PCGDP is not a good measure of productivity. It fails to take into account labor vs. capital productivity levels, as well as skewing output levels. If the US government started massive spending on the "Dig a Big Hole In the Ground" project, PCGDP would rise but nobody would say the US was being more productive.
And you still haven't answered my question about Luxembourg, Sweden and the other banking countries. Do they have some secret productivity methods, or is their PCGDP just high as a result of investment banking (which is not related to productivity)?
I don't know where you got the idea that GDP = productivity. Can you provide any source where a noted economist says PCGDP and productivity are the same? GDP is defined as the sum of consumer, investment and government spending, plus exports minus imports. The reason that the Cayman Islands, Switzerland and Luxembourg (and probably others) are so high is that there's a lot of banking and investment money going into them.
Or am I just not aware of Luxembourg's world renoknowned productivity secrets?
Out of curiosity, what dependence on Middle Eastern oil?
If you look at the
Department of Energy statistics (older statistics also available
here),
you'll see that the Middle Eastern countries which the US imports crude oil from (Saudi Arabia, Kuwait, Algeria are the only ones in the top 15) comprise less than 20% of the US imports. Canada and Mexico together are over 30% of imports. Despite what most people think, the US imports oil from a wide variety of places. Please take a look at this before making more statements like that.
I live in Japan now. Going to work every day, I pass by several "auto graveyards". These are just empty spots (usually an area cut out of the forested hills) where people park their cars and leave them, rather than pay to have them taken away properly. Used cars are a liability that no one wants, and you can often get a used car for free (or less than $100) at a dealership. One can also see piles of old furniture, steel drums, electronics and other assorted odds and ends.
When it costs $50 to legally get rid of a TV, most people are going to just dump it somewhere instead. I see no reason why this won't happen with computers, and result in the various tasty bits (lead, etc) leeching into the groundwater.
This exploit really isn't as bad as people here like to make it out to be. In order to perform this buffer overrun, you would have to trick the S@H client to connect to a different server. Short of actually breaking into the host computer of the client, I believe this would prove extremely difficult (anyone know how to do this?).
And as was mentioned in the advisory, there has been no reported case of this actually being exploited (outside of proof of concept of course, where the discoverer changed the S@H server address in the client itself).
BOINC will allow for multiple applications running at once (processes rather than threads) so this solves your worker thread concern. As for your other question, SETI@home data are also being used in generating a neutral hydrogen map of the galaxy, as well as (eventually) in AstroPulse.
Other concerns mentioned here involved the autodownloading of executables in BOINC. We're taking security very seriously in BOINC, and are using MD5 and 1024 bit RSA encryption to protect against malicious attacks, as well as other general design techniques. Finally, the issue of optimization. Since BOINC is open source, you can optimize it however you want, but there won't be much gain since BOINC itself does very little processing. As far as I know there's still no decision on whether to optimize the SETI@home science.
I'm very curious where you got the belief that carbohydrate consumption causes obesity. I've heard this before, but never seen proof of it. In fact, a quick search on the web shows the National Institute of Health and American Obesity Association not mentioning a single thing about carbohydrates causing obesity, but mentioning plenty about fat intake and exercise. Also, otherwebsites in fact repudiate the claim that carbohydrates cause obesity. Could you please give us some scientific sources for this claim?
I'd just like to point out that Nike does not in fact *own* any of these sweatshops that people enjoy complaining about. They don't actually manufacture their products, they are purely a marketing and design company. Of course then we get into the topic of whether they should be held responsible for the actions of those they contract with. Should a company be held responsible if their business partner has questionable practices? Should a consumer be held responsible if they purchase products from a company that has questionable business practices?
One interesting application of Moore's Law is to distributed computers like SETI@home. One of the major problems with SETI work is that it requires a supercomputer to analyze all the radio signal data coming in. By the time that the computer has been constructed and running for a few months, however, it's already obsolete.
But when the system is constructed with thousands of individual PCs, which get upgraded anyway every few years, the entire computer gets a speed boost without having to be completely redesigned and rebuilt. I'd be curious to see how well the speed increase of SETI@home has matched Moore's Law over the two years since its conception, I bet it is pretty close.
First of all, optical SETI is much different than SETI@home, so there wouldn't really have to be a similar distributed computer to do the analysis. All the major signal analysis is done in hardware immediately at the telescope. Basically to search for these signals you need two or three detectors and some simple hardware.
Since the photon emissions from a star are in a random fashion with regard to time, photons from a star usually hit only one of the sensors at a time. The idea is that if a laser pulse was fired at Earth for a few nanoseconds, the detectors would simultaneously receive many photons in the same few nanoseconds, which would be out of the ordinary for normal stars.
The interesting thing about Optical SETI is that you don't necessarily need a massive telescope to do it. Since the HW only cares about multiple photons in the same time frame, it doesn't matter much how precise your telescope is, just how precise your HW and detectors are.
>about the time i hit 1000 data packets processed, i learned that they were accepting money from the co founder of microsoft.
>
>that's a little like taking maney from organized crime...so i pulled the plug/couldn't care less now.
Just a nitpick, that's the SETI Institute (Project Phoenix) which received the money from Allen. For a complete list of all of the *SETI@home* corporate/individual sponsors, check their front page and donor list. Project Phoenix and SETI@home are not the same.
Slashdot Mirror
You could ask the governor of Tokyo, Shintaro Ishihara, when he said the following:
In due course, the perpetrators were captured, and, just as had been suspected, the crime was one of revenge among Chinese criminals. There is fear–and not without cause–that it will not be long before the entire nature of Japanese society itself will be altered by the spread of this type of crime that is indicative of the ethnic DNA [of the Chinese].
And this guy has been re-elected twice.
No, of course we can study and discuss potential (and real) differences between arbitrary groups of people. And of course there's no need for secrecy or disclaimers.
The problem comes when Japanese see this work and say "Aha! Foreigners can't eat sushi! Therefore it's OK for me to ban foreigners from my sushi restaurant since they can't eat it anyway." And if that logic sounds ridiculous or impossible, you've obviously never lived in Japan.
My point is that people should be conscious of how their research will be used. In this case, it would have been better for the authors to do a more in-depth study with more subjects, particularly Koreans (who eat large amounts of seaweed) or other areas with heavily fish based cuisine.
From my experience in Japan, the 'natives' love it when I try and speak Japanese when communicating. I've never sensed an ounce of condescending tones when I struggle.
Certainly they like it when you struggle speaking Japanese, it reinforces the idea that foreigners can't speak the language. When you reach native fluency they're much less friendly (speaking from years of experience here).
Just what we need, more "Japanese are unique" idiocy to justify racism and discrimination in Japan. So far we've heard that "Japanese intestines are longer, so Japanese can't eat foreign beef", "Japanese brains are unique, so only Japanese people can speak the Japanese language." and so on, all of which are supported by pseudo-scientific studies such as this one.
This sort of incomplete research just feeds the view of racial uniqueness (and superiority) among Japanese and justifies their racism and discrimination against others.
This raises an interesting point. Maybe in the brain there's no way to reduce things to meaningful equations. Evolution didn't really have a set goal in mind, so what you get is a cobbled together mess of pieces that works. The reason it works is because the system can reform and grow itself in response to new input, not because it has a lot of pieces.
These large scale simulations are interesting from a neurobiological point of view. But if you want to build a "thinking computer brain", you might be better off understanding how the human brain grows and changes in response to stimuli, rather than trying to make a static copy at the cellular level. I have a feeling these supercomputer simulations are not going to be the path to conscious computers, nor even to mind uploading.
Disclaimer: I am not a neuroscientist.
Hahaha, I've had the exact same thoughts!
I'm guessing you don't live in Japan, because if you did you'd probably think otherwise. Unless you like hearing endless parades of megaphone blasting vans, with high pitched voice women screaming nothing but "Please vote for Tanaka! Thank you! Please vote for Tanaka! Thank you!".
And if you want to talk about judging politicians by their hair, former Prime Minister Koizumi was known as "Lionheart" because of his hairstyle. Most Japanese I've spoken to identify him merely by his hair, and can't name a single policy he enacted. In almost all respects, I'd say the Japanese system is a good example of what to avoid rather than emulate.
I speak Japanese fluently. I have lived in Japan for about 6 years, currently enrolled at one of the "Top 3" public universities. I can read a newspaper in Japanese. I have given lectures in Japanese on theoretical computer science. I pay my taxes, I renew my visa on time. I don't do any of the other illegal or questionable things on your list.
Having said that, the grandparent is completely correct - Japan and Japanese are extremely racist, xenophobic and nationalistic. That's part of the reason why I'm leaving as soon as I finish my degree. You might think (as I did) if you learn the language and culture you'll be accepted - you're wrong. In Japan you will always be judged and discriminated against because you are not Japanese. It doesn't matter if it's in the city vs. countryside, company vs. university, public vs. private.
If you can tolerate being treated as a second class citizen, being denied housing/employment/services because of your skin color, being treated as an idiot incapable of understanding even basic Japanese, then feel free to come to Japan. If not, then avoid it.
This is exactly my point. Individually the processor may perform well, but when it's placed in the actual system, perform will undoubtedly drop. Right now, I'm doing performance tests on FFTs performed on GPUs (graphics cards). Theoretically, these should perform at the same "incredible" speed as the Cell processor (10 Gflops or better), but in reality bandwidth and cache constricts performance to half a Gflop.
You'll also note they say nothing about the system they tested on. Was it a PS3? I bet you it wasn't. Let's wait and see the actual performance of the PS3 before we get excited from in house tests done on an unspecified system by a company that is eager to boost impressions of its new chip.
I'm not entirely convinced the PS3 will be better than standard desktops, at least at non-gaming activities (I'm sure the PS3 will win as far as gaming goes). It seems if Sony could make a $300 console that outperforms a $1000 desktop, they'd be selling them as $1000 desktops instead of $300 consoles. But I guess we'll see when they come out.
At least with current platforms architectures. The author seems to do plenty of research on current distributed computing projects, but does none on how the consoles perform.
I know that SETI@home has been ported and tested at least on the XBox, and it performs miserably. These console gaming systems are designed to play games, not do radio signal analysis or other scientific calculation. For example, there's little need for fast memory writing when you're mostly reading textures from RAM, but there's an extreme need when you do millions of in-place Fourier transforms. Unless Microsoft and Sony change their architectures for some inexplicable reason, I can't imagine future architectures would perform much better.
This article smacks of ignorance on the part of the author, who clearly did no research into the actual performance of consoles in regard to standard scientific computing.
I stated that in my second post when I said:
"The labor force of a country is different than the number of citizens."
My numbers for the US are presented in the chart I mentioned in the grandparent post. Granted these don't take into account migrant labor, but I'm pretty sure the percentage of migrant/illegal labor in the US is nowhere near comparable to Luxembourg (which is about 30%).
Basically, my point was that PCGDP does not equal productivity. I think that was apparent by my first or second post.
Meaning GDP is not the entire answer.
"...to measure labor input, such as the number of workers..."
Like I said before, the population of a country is not equal to the number of workers. This chart (which is actually explicitly labelled "Overall Productivity") takes this fact into account:
http://www.nationmaster.com/graph-T/eco_ove_pro_pp p
On this chart, Norway falls to #6 and the banking countries drop off the top 25. This graph (which is not PCGDP) is a much more accurate depiction of productivity.
"Labor Force: 200,000 (of whom 87,400 are foreign cross-border workers primarily from France, Belgium, and Germany) (2003)" (From CIA World Factbook)
The only reason Luxembourg is so high is because one third of their work force comes from France, Belgium and Germany. These people contribute to output (raising productivity), but are not counted in the Luxembourg work force (which would lower productivity).
Therefore, your original statement that the US is not the most productive country in the world is incorrect. Your original statement that productivity equals PCGDP is also incorrect. If it were correct, the above chart would have the same rankings as the PCGDP chart.
No, you're missing the point, and I could make an equally intelligent response by questioning your English reading ability.
The CIA never said that our productivity is not the best. The CIA said that our PCGDP is not the best (I won't argue with them there). But the CIA *never said* that PCGDP = productivity. That's what you said, and what I argued against.
In your original post you said that "The myth that American workers are the most productive (Per Capita GDP) persists". You equated productivity with PCGDP. I said your interpretation of PCGDP was wrong, not the figures for PCGDP.
Actually I was paying attention in Economics 101, which is why I asked you the question. And your quote does not support your argument.
From your own article: "Simply stated, productivity is output per unit of input." Where does PCGDP take input into account? Where does PCGDP take product accounts into account? Where is the labor and capital stock input in PCGDP? The labor force of a country is different than the number of citizens.
Basically, PCGDP is not a good measure of productivity. It fails to take into account labor vs. capital productivity levels, as well as skewing output levels. If the US government started massive spending on the "Dig a Big Hole In the Ground" project, PCGDP would rise but nobody would say the US was being more productive.
And you still haven't answered my question about Luxembourg, Sweden and the other banking countries. Do they have some secret productivity methods, or is their PCGDP just high as a result of investment banking (which is not related to productivity)?
I don't know where you got the idea that GDP = productivity. Can you provide any source where a noted economist says PCGDP and productivity are the same? GDP is defined as the sum of consumer, investment and government spending, plus exports minus imports. The reason that the Cayman Islands, Switzerland and Luxembourg (and probably others) are so high is that there's a lot of banking and investment money going into them.
Or am I just not aware of Luxembourg's world renoknowned productivity secrets?
If you look at the Department of Energy statistics (older statistics also available here), you'll see that the Middle Eastern countries which the US imports crude oil from (Saudi Arabia, Kuwait, Algeria are the only ones in the top 15) comprise less than 20% of the US imports. Canada and Mexico together are over 30% of imports. Despite what most people think, the US imports oil from a wide variety of places. Please take a look at this before making more statements like that.
... and here's why:
I live in Japan now. Going to work every day, I pass by several "auto graveyards". These are just empty spots (usually an area cut out of the forested hills) where people park their cars and leave them, rather than pay to have them taken away properly. Used cars are a liability that no one wants, and you can often get a used car for free (or less than $100) at a dealership. One can also see piles of old furniture, steel drums, electronics and other assorted odds and ends.
When it costs $50 to legally get rid of a TV, most people are going to just dump it somewhere instead. I see no reason why this won't happen with computers, and result in the various tasty bits (lead, etc) leeching into the groundwater.
This exploit really isn't as bad as people here like to make it out to be. In order to perform this buffer overrun, you would have to trick the S@H client to connect to a different server. Short of actually breaking into the host computer of the client, I believe this would prove extremely difficult (anyone know how to do this?).
And as was mentioned in the advisory, there has been no reported case of this actually being exploited (outside of proof of concept of course, where the discoverer changed the S@H server address in the client itself).
Other concerns mentioned here involved the autodownloading of executables in BOINC. We're taking security very seriously in BOINC, and are using MD5 and 1024 bit RSA encryption to protect against malicious attacks, as well as other general design techniques. Finally, the issue of optimization. Since BOINC is open source, you can optimize it however you want, but there won't be much gain since BOINC itself does very little processing. As far as I know there's still no decision on whether to optimize the SETI@home science.
For more information, you can check out the BOINC source and BOINC documentation
I'm very curious where you got the belief that carbohydrate consumption causes obesity. I've heard this before, but never seen proof of it. In fact, a quick search on the web shows the National Institute of Health and American Obesity Association not mentioning a single thing about carbohydrates causing obesity, but mentioning plenty about fat intake and exercise. Also, other websites in fact repudiate the claim that carbohydrates cause obesity. Could you please give us some scientific sources for this claim?
I'd just like to point out that Nike does not in fact *own* any of these sweatshops that people enjoy complaining about. They don't actually manufacture their products, they are purely a marketing and design company. Of course then we get into the topic of whether they should be held responsible for the actions of those they contract with. Should a company be held responsible if their business partner has questionable practices? Should a consumer be held responsible if they purchase products from a company that has questionable business practices?
/ nike010402.html)
(Source: http://abcnews.go.com/sections/business/DailyNews
But when the system is constructed with thousands of individual PCs, which get upgraded anyway every few years, the entire computer gets a speed boost without having to be completely redesigned and rebuilt. I'd be curious to see how well the speed increase of SETI@home has matched Moore's Law over the two years since its conception, I bet it is pretty close.
Since the photon emissions from a star are in a random fashion with regard to time, photons from a star usually hit only one of the sensors at a time. The idea is that if a laser pulse was fired at Earth for a few nanoseconds, the detectors would simultaneously receive many photons in the same few nanoseconds, which would be out of the ordinary for normal stars.
The interesting thing about Optical SETI is that you don't necessarily need a massive telescope to do it. Since the HW only cares about multiple photons in the same time frame, it doesn't matter much how precise your telescope is, just how precise your HW and detectors are.
>about the time i hit 1000 data packets processed, i learned that they were accepting money from the co founder of microsoft.
>
>that's a little like taking maney from organized crime...so i pulled the plug/couldn't care less now.
Just a nitpick, that's the SETI Institute (Project Phoenix) which received the money from Allen. For a complete list of all of the *SETI@home* corporate/individual sponsors, check their front page and donor list. Project Phoenix and SETI@home are not the same.