You cant even do SAT in polynomial time with a quantum computer.
In fact, there really is no proof that quantum computers can do anything that (specialized) classical computers cant do.
I have seen hardware that uses content-addressable memory to data-search, or even sort, in O(1) time, and it doesnt use any quantum mumbo-jumbo.
Interestingly, even Shor's factoring algorithm may not be special. You see, nobody knows what the actually operating time of factoring on a determistic machine is; it is quite possible that an algorithm can be written which is just as "fast" using non-quantum hardware.
Ok, yes, quantum computers can make really good random numbers, but thats just a dongle on a classical computer stuck in a cup of tea.
How do you think this relates to quantum computers exactly? The very idea here is way way off base. Quantum computation is statistical. Solutions are built into a superposistion of answers, which must be clevery designed to positively interfere in order to create a signal that has a statistical chance of objectifying.
Take a good long look at Shor's algorithm to get an idea how this actually works.
I've seen several posts that imply that the its the job of the compiler to handle the parallelism. Quantum computers can only be exploited with highly specialized algorithms and as such, the compiler has no place. In fact people who study quantum computers today dont even use assembly language, they use gates.
It is clear to me that most people who think they have an idea of how quantum computers work don't. Now I'm not an expert, but I have studied up enough to know that they aren't just a happy parrallel abstraction. Most of the information you get on internet about quantum computers is completely bogus (as someone points out this paper appears to be).
Quantum computers are not universal; they cannot be used to do anything you want in "parallel universes".
I highly suggest people who even ponder quantum computers first get a reputable book on the subject.
The sharks and the marks. If you aren't the shark, then you are the mark. This thing will be stacked against gamers from the very start. Only a fool would put money on it.
What you say being true does not invalidate my comment. The wording of the statement is just not precise, which can be a factor in things of this nature.
If we choose to pick at it more, we see that the problem as stated does not actually assume the library is randomly sorted; what if it is is numberical order? In that case, locating shakespeare would be quite easy. In fact, there are many layouts of the library which could enable one to find information in less than order N time. It is a library after all, why would we assume it was randomly sorted? In fact, because the librarians have deduced that all combinations of books are present, then we must assume there IS a reducible logic that can locate books.
Yes, it does seem obvious at first that you could customize the recognizer for any particular person's speech pattern and achieve an agreeable solution.
However, the lessons of the parent posts must be taken into account; that which one can imagine to be so is not necessarily so, and using a particular technology instead of solving a particular problem gets the cart before the horse.
The point then is that in almost every case there is another solution *not using speech* that is a better solution than the best speech solution. The cost of customization for each user strongly outweighs simpler solutions that work better for a majority of accessibility challenged users.
Your "illustration" stating that a sufficiently complex system can't be emulated without some kind of understanding simply begs the question. It sits upon the foundation of it's own presumption. Arguing the Chinese room is like arguing the truth value of the statement "I am lying". It contains a self referential loop (the definition of semantics) that evades logical analysis.
The extra credit problem is flawed: using data compression it is possible to build an algorithm that does not "already contain shakespeare" that is able to perfectly identify shakespeare.
Just as an aside, I want to note that I did once investigate using speech for accessability [looking for a purpose for the technology] but was basically reprimanded by some accessibility Guru's that speech recognition is rarely a good solution. Basically a large number of people who are unable to use keyboard or mouse are also limited in their ability to speak clearly. Oh they had a litany of other reasons it wasn't as good as it sounds [familiar argument eh?].
What you are saying is in direct contradiction of several controlled studies I have participated in. My information comes from actual human studies and post-mortem analysis of the problem at hand; so I wonder what your basis is for comments like "voice command would make a wonderful optional addition." Could it be..pure imagination?
[Oh sure, there are a few people who appreciate techno-gadgets like "the clapper", but the majority of people do not, and in fact ridicule it.]
I was once paid to basically build "sci-fi" technology in order to demonstrate new research technologies.
My experience has been that sci-fi inspired technology rarely 'works' as dramatized on TV. What I mean by 'works' is that even with a perfect system [as simulated by Wizard of Oz experiements], humans will not be impressed by, nor even tolerate, those technolgies.
Here is an example of sci-fi meets reality.
One system we built was similar to the house in "Minority Report". You could talk to the lights and query the room about various information, that sort of thing. In the end, the idea was hopelessly misguided.
The reasons this particular demo sucks is because of cognitive load, cognitive dissonance, and limited human bandwidth. Cognitive load means your brain has to think more to get a task done; cognitive dissonance means your brain is uncomfortable doing the task, and bandwidth means mainly that human speech is slow.
For example, a "lights on" command requires concious thought in order to get lights, and some linguistic processing. The alternative light switch technology is less so, even automatic [you might notice this when the power goes out you still hit switches]. Also, humans are pre-programmed to talk to humans, talking to the wall is an unpleasant experience for most people. Finally, speech is really quite slow. Flicking a light switch is much faster than saying the words.
The point is that the dramatization of this technology is done in the imagination with all factors tuned optimally for dramatic effect; but the reality falls short of the fantasy. Real world factors not taken into account by the imagination destroys the appeal of the technology.
So what is a better model for driving innovation than the fantasy scenarios of fiction? Quite simply, it is the time tested process of real-world problem solving. Find a problem, look for a solution [as contrasted to find a technology look for a use]!
You are not correct! Read carefully: pre-emptive MSDOS multitasking. You could run multiple MSDOS apps at once and independantly. Probably if one crashed, the others would keep running. Though you should not think multitasking means protection; it did use protection, but not necessarily to protect anything:)
Why back in the middle ages get this..they used swords! Those fools! Why didn't they just use guns!
Programmers today have no clue what programming was like back in the early days of the PC. The system had to boot in 64k, which is equivilant to a few icons in todays world. The graphics technology was so primitive most programmers today would refuse to write code for it; the pixels weren't square and there was no screen read!
Yet the functionality was substantially similar to what we have today; networking, graphics, spreadsheets, word processors with fonts.
Put down the early days of windows all you want, twenty years from now you will be defending the "boneheaded code" you wrote in your youth and you may just get a taste of it; though not the full course meal since starting a billion dollar enterprise is much much more difficult than coat-tailing on one.
I was once enamored by the idea of building a house that would last a thousand years. While I came up with several ways to do it, I also came to the conclusion that to do so is a terrible idea.
The needs of today are not the needs of tomorrow. If you have ever visited some of those thousand year old towns in europe you know that the streets are too small, the heating sucks, power lines and pipes have no place to hide, drafty and damp. Not a happy place to be but for SCA fans. The castles of old are horrible places to spend any amount of time as well, not because they are old, but because they were designed with different priorities.
Thus, we can project that in the future, today's home of paradise might be quaint or gaudy to future eyes. But they won't be able to tear it down and building something good because it will be a historical landmark. A useless museum probably. And the children of tomorrow will be trapped inside buildings built by long dead peoples.
Perhaps in the future there will be no houses at all! Borrowing from Philo Farnsworth's ideas about the potential of fusion, maybe house of tomorrow will fit in your pocket when not in use, and be constructed entrirely of force fields. The old time houses will seem like caves!
What I want even more than that last bit of graphics speed is a driver that doesnt crash every few minutes. As everyone who plays 3d games knows, the driver that comes with the card is unusable. The only thing that will typically run under it is the benchmark. So the first thing you do when you get a new card OR a new game, is go to the board's manufacturer's site and get the latest driver (and pray).
My experience tells me that nVidia is ahead in this area. When a new game comes out, if there is a bug that stops it from running or causes random crashes, the fix will usually be released by the game's release date. ATI on the other hand tends to both have buggier drivers and lag weeks behind on bug fixes.
So the bottom line is, if you are planning on playing that hot new MMPORG on release day, you are probably better off going with nVidia since you are more likely to get a driver that works.
Based on random dice rolls. We can play as many times as you like. If you are one who believes in luck, lets see if you can beat me.
People who aren't good at statistics will get creamed. Its this way in real life too. Success is not dumb luck, its playing the odds. Its just an excuse when the loser blames the dice.
This article is not the way Bill Gates talks. Dollars to donuts this is a big hoax. It reads like something a slashdotter might say Bill Gates said, not something he would actually say.
Of course this can work. Just like paypal Customer creates and account and puts in a minimum amount via credit card. The card is charged. Just like paypal. The only difference is how the money is paid to the merchant; random digital peppercorns.
I remember back in the day when a certain OS used random to decide which thread got to run. That scheduler was based on similar logic to peppercoin's: It will all even out on the large scale. The problem was, some things required it to even out on the small scale and some important services were starving every once in a while (maybe days) and taking the system down.
Also, the article fails to explain why the payments are done randomly. And how the 'random' token is chosen. Without some rational and more precise information, this idea is nutty.
Thats about what I payed for my current card. I don't know I would say the average gamer spends this much, but there is a large minority that do. The number of hours some people put into games makes the investment trivial compared to the opportunity cost of playing a game for so many hours.
Why do so many game makers screw up the most basic part of the game: installing and paying. I have examined there site and it is a big confused jumble of gobblygook. A lot of games sites are like this, but most of them have a box and cd that gets you to the starting line.
Where is the starting line? Oh, I can figure it out no doubt, but will I bother?
Why not have the home page for this game be nice clean graphic with a signup and install button. Sure, have some tabs to the gobblygook. But in the main as a player, I don't want to have to surf around the site for the right voodoo doll to get thing up and running.
Time has not been proven to continue into infinity. Go read 'A Brief History of Time' for a good laymens introduction to cosmology.
Space and time are concepts deeply intertwined with energy and matter; they is not distinct from them. Thus, there is no 'before' the universe began, there is no time there, there is no there there either.
When I took physics not so very long ago, they never mentioned 'Dark Energy' as one of the forces of nature. They were real good at calling people who talked about alternative energy like 'radiant energy' whack-jobs. But reality is stranger than fiction; it turns out that the wackjobs were right about at least one thing: the universe is full (73%) of unseen energy forces which are beyond the ken of modern physics.
Slashdot Mirror
You cant even do SAT in polynomial time with a quantum computer.
In fact, there really is no proof that quantum computers can do anything that (specialized) classical computers cant do.
I have seen hardware that uses content-addressable memory to data-search, or even sort, in O(1) time, and it doesnt use any quantum mumbo-jumbo.
Interestingly, even Shor's factoring algorithm may not be special. You see, nobody knows what the actually operating time of factoring on a determistic machine is; it is quite possible that an algorithm can be written which is just as "fast" using non-quantum hardware.
Ok, yes, quantum computers can make really good random numbers, but thats just a dongle on a classical computer stuck in a cup of tea.
Unfortunately, though the concepts seem similar to quantum computing models, they are in fact different.
How do you think this relates to quantum computers exactly? The very idea here is way way off base. Quantum computation is statistical. Solutions are built into a superposistion of answers, which must be clevery designed to positively interfere in order to create a signal that has a statistical chance of objectifying.
Take a good long look at Shor's algorithm to get an idea how this actually works.
I've seen several posts that imply that the its the job of the compiler to handle the parallelism. Quantum computers can only be exploited with highly specialized algorithms and as such, the compiler has no place. In fact people who study quantum computers today dont even use assembly language, they use gates.
It is clear to me that most people who think they have an idea of how quantum computers work don't. Now I'm not an expert, but I have studied up enough to know that they aren't just a happy parrallel abstraction. Most of the information you get on internet about quantum computers is completely bogus (as someone points out this paper appears to be).
Quantum computers are not universal; they cannot be used to do anything you want in "parallel universes".
I highly suggest people who even ponder quantum computers first get a reputable book on the subject.
The sharks and the marks. If you aren't the shark, then you are the mark. This thing will be stacked against gamers from the very start. Only a fool would put money on it.
What you say being true does not invalidate my comment. The wording of the statement is just not precise, which can be a factor in things of this nature.
If we choose to pick at it more, we see that the problem as stated does not actually assume the library is randomly sorted; what if it is is numberical order? In that case, locating shakespeare would be quite easy. In fact, there are many layouts of the library which could enable one to find information in less than order N time. It is a library after all, why would we assume it was randomly sorted? In fact, because the librarians have deduced that all combinations of books are present, then we must assume there IS a reducible logic that can locate books.
Yes, it does seem obvious at first that you could customize the recognizer for any particular person's speech pattern and achieve an agreeable solution.
However, the lessons of the parent posts must be taken into account; that which one can imagine to be so is not necessarily so, and using a particular technology instead of solving a particular problem gets the cart before the horse.
The point then is that in almost every case there is another solution *not using speech* that is a better solution than the best speech solution. The cost of customization for each user strongly outweighs simpler solutions that work better for a majority of accessibility challenged users.
Your "illustration" stating that a sufficiently complex system can't be emulated without some kind of understanding simply begs the question. It sits upon the foundation of it's own presumption.
Arguing the Chinese room is like arguing the truth value of the statement "I am lying". It contains a self referential loop (the definition of semantics) that evades logical analysis.
The extra credit problem is flawed: using data compression it is possible to build an algorithm that does not "already contain shakespeare" that is able to perfectly identify shakespeare.
Just as an aside, I want to note that I did once investigate using speech for accessability [looking for a purpose for the technology] but was basically reprimanded by some accessibility Guru's that speech recognition is rarely a good solution. Basically a large number of people who are unable to use keyboard or mouse are also limited in their ability to speak clearly. Oh they had a litany of other reasons it wasn't as good as it sounds [familiar argument eh?].
What you are saying is in direct contradiction of several controlled studies I have participated in. My information comes from actual human studies and post-mortem analysis of the problem at hand; so I wonder what your basis is for comments like "voice command would make a wonderful optional addition." Could it be..pure imagination?
[Oh sure, there are a few people who appreciate techno-gadgets like "the clapper", but the majority of people do not, and in fact ridicule it.]
I was once paid to basically build "sci-fi" technology in order to demonstrate new research technologies.
My experience has been that sci-fi inspired technology rarely 'works' as dramatized on TV. What I mean by 'works' is that even with a perfect system [as simulated by Wizard of Oz experiements], humans will not be impressed by, nor even tolerate, those technolgies.
Here is an example of sci-fi meets reality.
One system we built was similar to the house in "Minority Report". You could talk to the lights and query the room about various information, that sort of thing. In the end, the idea was hopelessly misguided.
The reasons this particular demo sucks is because of cognitive load, cognitive dissonance, and limited human bandwidth. Cognitive load means your brain has to think more to get a task done; cognitive dissonance means your brain is uncomfortable doing the task, and bandwidth means mainly that human speech is slow.
For example, a "lights on" command requires concious thought in order to get lights, and some linguistic processing. The alternative light switch technology is less so, even automatic [you might notice this when the power goes out you still hit switches]. Also, humans are pre-programmed to talk to humans, talking to the wall is an unpleasant experience for most people. Finally, speech is really quite slow. Flicking a light switch is much faster than saying the words.
The point is that the dramatization of this technology is done in the imagination with all factors tuned optimally for dramatic effect; but the reality falls short of the fantasy. Real world factors not taken into account by the imagination destroys the appeal of the technology.
So what is a better model for driving innovation than the fantasy scenarios of fiction? Quite simply, it is the time tested process of real-world problem solving. Find a problem, look for a solution [as contrasted to find a technology look for a use]!
Those calls were only made..by 3rd party dos apps! And screw some book, I wrote some of that code :)
You are not correct! Read carefully: pre-emptive MSDOS multitasking. You could run multiple MSDOS apps at once and independantly. Probably if one crashed, the others would keep running. Though you should not think multitasking means protection; it did use protection, but not necessarily to protect anything :)
Why back in the middle ages get this..they used swords! Those fools! Why didn't they just use guns!
Programmers today have no clue what programming was like back in the early days of the PC. The system had to boot in 64k, which is equivilant to a few icons in todays world. The graphics technology was so primitive most programmers today would refuse to write code for it; the pixels weren't square and there was no screen read!
Yet the functionality was substantially similar to what we have today; networking, graphics, spreadsheets, word processors with fonts.
Put down the early days of windows all you want, twenty years from now you will be defending the "boneheaded code" you wrote in your youth and you may just get a taste of it; though not the full course meal since starting a billion dollar enterprise is much much more difficult than coat-tailing on one.
DOS thunking only occured to make 3rd party Terminate and Stay Resident (TSR) programs continue function as expected. No TSR's, no thunking.
Thus it was only for compatibility with old apps that this highly complex feature was added. Today they'd cry monopoly if that hadn't been done.
I was once enamored by the idea of building a house that would last a thousand years. While I came up with several ways to do it, I also came to the conclusion that to do so is a terrible idea.
The needs of today are not the needs of tomorrow. If you have ever visited some of those thousand year old towns in europe you know that the streets are too small, the heating sucks, power lines and pipes have no place to hide, drafty and damp. Not a happy place to be but for SCA fans.
The castles of old are horrible places to spend any amount of time as well, not because they are old, but because they were designed with different priorities.
Thus, we can project that in the future, today's home of paradise might be quaint or gaudy to future eyes. But they won't be able to tear it down and building something good because it will be a historical landmark. A useless museum probably. And the children of tomorrow will be trapped inside buildings built by long dead peoples.
Perhaps in the future there will be no houses at all! Borrowing from Philo Farnsworth's ideas about the potential of fusion, maybe house of tomorrow will fit in your pocket when not in use, and be constructed entrirely of force fields. The old time houses will seem like caves!
What I want even more than that last bit of graphics speed is a driver that doesnt crash every few minutes.
As everyone who plays 3d games knows, the driver that comes with the card is unusable. The only thing that will typically run under it is the benchmark. So the first thing you do when you get a new card OR a new game, is go to the board's manufacturer's site and get the latest driver (and pray).
My experience tells me that nVidia is ahead in this area. When a new game comes out, if there is a bug that stops it from running or causes random crashes, the fix will usually be released by the game's release date. ATI on the other hand tends to both have buggier drivers and lag weeks behind on bug fixes.
So the bottom line is, if you are planning on playing that hot new MMPORG on release day, you are probably better off going with nVidia since you are more likely to get a driver that works.
Based on random dice rolls. We can play as many times as you like. If you are one who believes in luck, lets see if you can beat me.
People who aren't good at statistics will get creamed. Its this way in real life too. Success is not dumb luck, its playing the odds. Its just an excuse when the loser blames the dice.
This article is not the way Bill Gates talks. Dollars to donuts this is a big hoax. It reads like something a slashdotter might say Bill Gates said, not something he would actually say.
Of course this can work. Just like paypal
Customer creates and account and puts in a minimum amount via credit card. The card is charged. Just like paypal.
The only difference is how the money is paid to the merchant; random digital peppercorns.
Don't ask me why this would be desirable.
I remember back in the day when a certain OS used random to decide which thread got to run. That scheduler was based on similar logic to peppercoin's: It will all even out on the large scale. The problem was, some things required it to even out on the small scale and some important services were starving every once in a while (maybe days) and taking the system down.
Also, the article fails to explain why the payments are done randomly. And how the 'random' token is chosen. Without some rational and more precise information, this idea is nutty.
Thats about what I payed for my current card. I don't know I would say the average gamer spends this much, but there is a large minority that do. The number of hours some people put into games makes the investment trivial compared to the opportunity cost of playing a game for so many hours.
Why do so many game makers screw up the most basic part of the game: installing and paying. I have examined there site and it is a big confused jumble of gobblygook. A lot of games sites are like this, but most of them have a box and cd that gets you to the starting line.
Where is the starting line? Oh, I can figure it out no doubt, but will I bother?
Why not have the home page for this game be nice clean graphic with a signup and install button. Sure, have some tabs to the gobblygook. But in the main as a player, I don't want to have to surf around the site for the right voodoo doll to get thing up and running.
Time has not been proven to continue into infinity. Go read 'A Brief History of Time' for a good laymens introduction to cosmology.
Space and time are concepts deeply intertwined with energy and matter; they is not distinct from them. Thus, there is no 'before' the universe began, there is no time there, there is no there there either.
When I took physics not so very long ago, they never mentioned 'Dark Energy' as one of the forces of nature. They were real good at calling people who talked about alternative energy like 'radiant energy' whack-jobs. But reality is stranger than fiction; it turns out that the wackjobs were right about at least one thing: the universe is full (73%) of unseen energy forces which are beyond the ken of modern physics.