To have an idea of what could happen if the descendants of computers became many orders of magnitude more intelligent than humans, you can try and read some SF books on the matter, for example Ian M. Banks books on the Culture. Start with "Consider Phlebas".
You can make many readings of Banks' vision. Human life and endeavours may become worthless or they may not. People can choose to lead idle and comfortable lives, or they may decide to lead dangerous and still productive ones. Minds (as those computers are called) control most habitats and starships in a benevolent manner, but some Minds turn evil or at least manipulative, and some worlds are still free of their influence. In spite of their intelligence there are still things than Minds cannot do.
Things might work out. I would dearly like our beloved and respected leaders to be many orders of magnitude more intelligent and removed from the trappings of material life than they are right now.
Paul Allen has had cancer and survived. As a result he has re-organized his priorities, such as enjoying life a little bit more, giving away things he doesn't need, giving others a chance, etc.
Why does it takes cancer to come to this view of life is beyond me, but kudos all the same.
I'm not sure I understand what you mean. If you mean that BeOS makes things simple for the developers by allowing use of multi-processors transparently then I think this is a bit of an optimistic view.
Developers would still need to write their code taking multi-processors into account. This is not so easily done with standard languages such as C/C++ or even Java etc, which do have support for multithreading.
Support for threading and transparent threading usage are two different things. Usually the former is easy to implement, hard to use well for developers but when used well it does provide significant benefits for the end-user, and the latter is hard to implement, easy to use for the developer but does not usually deliver as much benefits as the hand-coded parallel-loop approach.
As an illustration, Intel provides a parallel C++ compiler for Linux and Windows. If you let the compiler do the parallelization it does work but the benefits are not great in general, unless the developers take great care in writing their code, essentially laying out cues to the compiler. If the developers do the parallelization by hand using OpenMP or threading libraries, then the code is even harder to write well but often it delivers better results: better scaling, better load balancing, etc.
The answer is that it doesn't make sense for a desktop machine.
Windows professional comes with a license for 1-2 CPUs. Above that you need to purchase one of the server edition, and it starts becoming *very* expensive.
Soon 2 CPUs will be for the masses, they already are with hyperthreading in a way. However 4 and above really are for servers; multi-user environments, etc.
Also while it is easy to exploit 2 CPUs in a desktop environment (roughly speaking 1 for the O/S, the other for the applications) there are diminishing returns for more than 2 CPUs, except for parallel applications, and there aren't that many of those.
This raises questions regarding stability and Windows.
While I find that multiprocs settings under Linux improve things to a significant degree (although there are still outstanding issuess with NVidia proprietary drivers and SMP), I found the opposite true for Windows.
The last time I tried, which was about 2-3 years ago, many drivers didn't seem to expect true concurency under Win2k and I was experiencing significantly more crashes on my dual P-III than when I forced the system to only use one of the CPUs. Yet it probably wasn't the hardware because that same machine was very stable with Linux.
With the advence of hyper-threading, have things improved markedly with WinXP?
Well the "I have nothing to hide or fear" argument is an old one, but it doesn't make it valid.
Currently to read my number plate more or less requires a human. A few systems are able to read my plate in a few selected places, but I don't really mind that.
With RFID it would finally become cheap and effective to track everybody everywhere. You could get a fine everytime you exceed the local speed limit, it would be all automated and reliable, which would be terribly annoying but maybe somewhat justifiable.
However how do we know that this situation will only be used for traffic regulation enforcement? Will it be tamper-proof? will you have the right to inspect your record? will the data be wiped out after a given time frame? Who will have the right to subpoeana the record?
The right to privacy is also extremely important. I don't want companies or agencies to look at my driving patterns and deduce any kind of behavioural conclusion without my approval.
Imagine for example that you are innocent but suspect of some kind of crime and that where you go and your surroundings are closely monitored. No one would like to go to your place or you to visit them for fear of being associated with you somehow and also questioned by police, say.
Other not-so-nice scenarios are not that hard to imagine.
All sorts of questions one would like answered before signing on to that sort of deal.
In pure dollar amounts you may be right, however it is still not a good trade because it will undermine incentives to develop IP and content at home. The Australian cinema you love will be even more challenged than it is now, and our beleagered research and IT sectors will suffer some more. Did you know that 96% of movies released in Australia come from Hollywood? The 4% remaining are for the rest of the world, including Australian cinema. Talk about cultural hegemony.
What this FTA is likely to do is play up our export strengths which are in the commodities business (primary resources) and undermine our weaknesses further (intellectual content). However commodities are relatively low-value, finite, volatile and not under our control (Australia doesn't get to set the price of coal for example). It might turn out OK, it may not ; conversely unique intellectual content is high-value, well protected, renewable and much more under control.
In the future we want to rely more on the former and less on the latter. The FTA is not going in this direction and this is dangerous.
Actually I wish this agreement would make us the 51st state, because then we would have some rights coming our way such as a proper constitution and a bill of right.
Instead this FTA takes away a lot of things, almost guarantees more litigation, invites for a larger trade deficit in favour of the US in the high-tech sector in return for some very meager returns in the Agricultural sector.
People say the balance will be somewhat even on the financial side all being considered, but in the meantime the rights of every Australians will be eroded with nothing to compensate.
Some content currently in the public domain in Australia will revert to being copyrighted. For example in Australia we have an event called Bloomsday, in honour of Ulysses, the book by James Joyce. This is a day of celebration of Joyce's memory, with new plays based on Joyce's book being peformed and things like that. This has only recently been possible because the estate of James Joyce generally forbids any adaptation of Joyce's novels for any purpose.
Of course the estate can do nothing about stuff that is in the public domain. But with the FTA, Ulysses and other works will return to the estate and we will have to wait another 15 years for the next bloomsday, or longer if yet another copyright extention is granted down the track.
I think this is mad. This sort of thing does not serve the public.
In Australia you may sometimes copy software for backup purposes, but you cannot backup music or video or any other copyrighted material even for personal purposes. Precisely what is purchased when buying a tape or CD or DVD is a license to play that content on approved harware using the medium provided only. If it breaks down you have to buy another one, sorry.
Photocopy of printed material is also very tightly controlled. At my place of work you cannot make a copy of anything yourself. You have to ask the librarian and fill a form. I am not joking. For example you are allowed to copy up to 10% or one article of one magazine or journal, whichever is less, for research or teaching purposes. Personal use is not allowed.
90% ? Where did you get this number? this is nonsense, sorry. The TOTAL contribution of ALL agricultural products to the Australian GDP is 3%. By comparison services amount to 71%.
As for exports, Australia primary sector (commodities such as coal, gold, meat, wool, alumina, iron ore, wheat, machinery and transport equipment) amount to 65% of all exports.
It's not as hard as it sounds, especially with a decent supervisor. The hard bit about science is not solving problem, it's knowing which one to solve. Every field is literally littered with unsolved questions and problems. Which ones are sufficiently hard to be interesting yet easy enough to be cracked in 3 years of work or less? It is the supervisor's work to come up with these, not the student's.
Microsoft may not use patents directly themselves, but they might sponsor other companies to do it for them. I think they will try a bit harder at least one more time even if SCO flounders.
At the moment the MSFT share price has stalled. They are looking for big markets and they are not finding them easy to break into. If Linux starts making inroads in the desktop in the next couple of years, I would expect Microsoft to start playing desperate tactics. It might be the beginning of the end but it would be nasty nonetheless.
If.NET really still is the cornerstone of Microsoft mid-term strategy (for when Longhorn is released), I would expect them to hit hardest there. Of course, who am I, I'm very likely to be completely wrong.
Servers are a tiny market compared with desktop, and are not software-bound. I don't believe they would use pseudo-interoperability as an argument to buy Microsoft. If they want they already have a portable environment for providing Unix-style servers, it used to be called Interix and it works very well. My interpretation of things as they stand is that.NET code still requires vast amounts of non-portable win32 code to run properly. There are even whole books devoted to this very topic. Microsoft is in the business of providing win32 solutions, not developing portable platforms.
Finally Microsoft gave away the specs of C# to ECMA for standardisation, but see above discussion about licensing terms. They are at least RAND, but no one seems to know for sure if they are royalty-free. In other words Microsoft hasn't given MSIL away, on the contrary, it could be a neat poison pill.
Microsoft will be able to take out Mono any time they choose.
The lion and the lamb shall lie down together, but the lamb won't get much sleep.
Microsoft now has a very strong research group located in various places around the globe: China, Europe, Australia, US, etc. These guys don't just innovate, they are in the process of inventing new and ambitious things. These guys are active in all the areas of computer science from computer vision to advanced databases.
The problem is that sometimes big things start small. Look at the PC industry. It really started with home hardware kits, and now look where it's at. This shouldn't be a problem, but Microsoft is not interested in small things. They only want to corner billion dollar markets: they don't want to defocus their interest into thousands of little projects, only a handful of which will make it into a multi-million-dollar industry, and potentially one or two would break the billion dollar threshold. They don't have the patience.
The reason for this is that they are the company with the largest market cap in history, and they need to deliver value for the shareholders. A new million-dollar Microsoft startup wouldn't even register on the investor's radar. Too many of these little companies failing would devalue the stock. The board would be accused of wasting the shareholders' money.
The associated problem is that there aren't so many multi-Billion (with a B) markets around that are somehow associated with Microsoft core business. Microsoft is not in energy distribution or transport, they would probably suck at managing that (Microsoft Air, who would get on board?). That is why they are now fighting with the Sony, Palm and Google of this world. They want those markets and they want them bad.
The problem is that they don't have a good customer relation image (they've worked very hard at appearing ruthless and uncaring, it seems) and they don't have a good track record of breaking into market where they don't have a monopoly. In these markets the traditional Microsoft tactics don't work (strongarm everybody). That is why they are currently losing lots of money in the console market.
It is interesting to watch that despite Bill Gates' team famed smarts and they billions in cash in the bank, Microsoft is still having a hard time owning the world.
It will be interesting to see how it all turns out. Can Microsoft break the conundrum of the big rich company hamstrung by its shareholders?
You are writing replies that are even longer than mine, quite a feat for/.
Yes ANN only generate computable outputs, at least the ones we have now. But is the same true for NNN (the Natural kind)? The answer is currently a matter of beliefs. If not, no amount of Moore's law is going to help.
I used to belive just like you that true AI was trivially provably possible by emulation, but now I'm not so sure.
Concerning a generation that doesn't die anymore, they might be stuck with interesting problems. What if your life was very very long but incredibly boring and frustrating, because the society around has ossified: no one can get promoted, new ideas are not wanted. Old men looking like they are 20 year olds are in power and want to remain so forever. It could in fact suck quite a bit.
When Niels Bohr was asked how he had managed to convince his physicist colleagues that his quantum ideas were correct, he replied that he never managed to do it. His ennemy belonged to the previous generation of physicists and simply eventually retired or died. They never accepted his work. In a society where no one dies, it could mean the end of many things.
I'm a believer of the incredible power of true science, but I've also been a witness of the effects of the results of science in the wrong hands, and so have been quite a few people.
You could be temporarily incapacitated. Hit by a bus and spend a few months in a coma. It would be nasty to wake up and find out your wife divorced you and your kids don't want to talk to you and you have no money left because of a few files.
I haven't missed your point, on the contrary. I know about the hippocampus story, and they have only emulated a tiny portion of the suborgan. The hippocampus is a little bit bigger than 18 neurons.
There are people who believe that simply putting together huge numbers of artificial neurons together will not be enough. They say that non-computable quantum effects occur in the brain and that we will need to emulate those to reproduce thinking. This is what most proeminently Roger Penrose belives. It is very difficult to simulate quantum effects. As you probably know, one can use quantum effects to do very quick calculations, for example factoring using Shor's algorithm. The converse of that is that it is very inefficient to reproduce quantum effects on traditional computers.
Results from people who work with artificial neural network seem to bear that out. In final analysis, an ANN doesn't perform anything widely different than what traditional statistical classification does, which is a normal deterministic computable task. In other words non-computable behaviours don't emerge spontaneously from ANN. Maybe it is a function of the size of the ANN, but people are beginning to doubt this.
Maybe Penrose is wrong and a straight normal emulation will work, but no one knows for sure. Also maybe we will have quantum computers at our disposal in a few decades, but then again maybe not, and will those that we will have help bringing about AI? no one knows.
Yes Moore's law has held out for a very long time, it doesn't mean that it will continue to do so forever. The end of Moore's law has been predicted many times, but at some point we will be hitting some pretty strong barriers: how many molecules can we use to make a switch? how small can we shrink a die so that speed-of-light delays don't induce problems (at 10GHz, a speed which is on the horizon, light only travels 3cm)?
The fact is we are still using the same basic techniques for building CPUs since the 1960s. Techniques have improved and dies have shrunk, but the basic ideas have remained the same (die, clock, gates, etc). To overcome the size barriers we will probably have to switch to clockless designs, massive parallel architectures, and certainly vastly more efficient cooling that we have now. Moore's law may have a few years or even decades to go, but I'm willing to bet it won't hold out another 100 years. In fact it is likely that soon, the cooling system in computers will cost more than the CPU. Pumps and cryogenic systems are not amenable to Moore's law.
If you look at the way you programmed a computer in 1970, it wasn't vastly different to what people are doing now. Learning to program clockless designs and parallel designs is going to require massive retraining, and not all problems are going to be amenable to these.
I think I agree with you that the next few decades are going to bring about some massive changes. Call it a singularity if you want. I think besides the computing issues, energy, environmental and political issues will need a total rethink. As we are now, we are close to a disaster of vast proportion in the Middle East. We survived the cold war, can we survive global terrorism? Are we going to start running out of oil in 10 years time or not? Some people show that the oil field discovery peaked at the end of the 80s and that inevitably the production peak will come. From then one whatever we do, we'll have less oil to work with. That's a sobering thought.
If the human race manages to not anihilate itself in the next 50 - 100 years, we might witness a number of totally extraordinary things, such as the first AI, self-reproducing benevolent nanotech, effective gene repairs, possibly a worldwide democracy, a fully explored solar system and vastly extended lifetimes. If we survive the next 100 years more or less intact, I'm more confident we can make it in the long run.
Possibly unlike you, I don't believe it will happen by itself. It will be hard work, and sometimes I worry we can't ma
Don't worry, there are plenty of anecdote of Jobs being an asshole too:
...One thing that Woz and agree on: the portrayal of Steve Jobs was good. In fact, Woz said that Jobs' tyrades and abuse of his employees was much worse than in the movie. The movie makes him out to be a real asshole with a messiah complex. Maybe it was all of the acid he dropped, I dunno....
Let's say they were at the right place at the right time with the IBM-PC deal. Also Gates did get a killer deal regarding MS-DOS, he could license it to whoever he wanted, not just to IBM. Without that PC clones would have been running DR-DOS or something like that.
After that Gates did a magnificent backstab with OS/2, that was masterful. Pretend to cooperate with an ailing IBM once again to develop the new "DOS", price the SDK too high for casual developers so no community develops, and pull out Windows-3.0 without warning, that could run on *every* PC, not just the 286s and above unlike OS/2.
Right about then they started being extremely nasty with competitors, like Borland, Lotus, etc.
Slashdot Mirror
So what you are saying is that the BSD licence is idealistic and the GPL is pragmatic but both tend towards the same goal?
If that's the only important difference then can we stop arguing over which one is better?
To have an idea of what could happen if the descendants of computers became many orders of magnitude more intelligent than humans, you can try and read some SF books on the matter, for example Ian M. Banks books on the Culture. Start with "Consider Phlebas".
You can make many readings of Banks' vision. Human life and endeavours may become worthless or they may not. People can choose to lead idle and comfortable lives, or they may decide to lead dangerous and still productive ones. Minds (as those computers are called) control most habitats and starships in a benevolent manner, but some Minds turn evil or at least manipulative, and some worlds are still free of their influence. In spite of their intelligence there are still things than Minds cannot do.
Things might work out. I would dearly like our beloved and respected leaders to be many orders of magnitude more intelligent and removed from the trappings of material life than they are right now.
Paul Allen has had cancer and survived. As a result he has re-organized his priorities, such as enjoying life a little bit more, giving away things he doesn't need, giving others a chance, etc.
Why does it takes cancer to come to this view of life is beyond me, but kudos all the same.
To the curious: Louis Armstrong's answer to the question "What is Jazz?"
In true civil desobedience fashion, the proper way to make decision makers understand that they are wasting their time is to:
1- purchase the CD
2- Optional: rip & copy it
3- return it and get a refund because it doesn't play on your equipment.
(2) is optional. The proper and law-abiding way is to not rip that CD.
If the return rate goes to around 10% or so I think the message will be pretty clear.
Two questions:
1- Which GUI toolkit did you use for the C++ application?
2- Which GUI toolkit did you use for the Java implementation?
Had you given any thoughts of using a modern C++ toolkit for the re-implementation?
Absolutely!
Good luck shooting sports using the LCD viewfinder.
I'm not sure I understand what you mean. If you mean that BeOS makes things simple for the developers by allowing use of multi-processors transparently then I think this is a bit of an optimistic view.
Developers would still need to write their code taking multi-processors into account. This is not so easily done with standard languages such as C/C++ or even Java etc, which do have support for multithreading.
Support for threading and transparent threading usage are two different things. Usually the former is easy to implement, hard to use well for developers but when used well it does provide significant benefits for the end-user, and the latter is hard to implement, easy to use for the developer but does not usually deliver as much benefits as the hand-coded parallel-loop approach.
As an illustration, Intel provides a parallel C++ compiler for Linux and Windows. If you let the compiler do the parallelization it does work but the benefits are not great in general, unless the developers take great care in writing their code, essentially laying out cues to the compiler. If the developers do the parallelization by hand using OpenMP or threading libraries, then the code is even harder to write well but often it delivers better results: better scaling, better load balancing, etc.
The answer is that it doesn't make sense for a desktop machine.
Windows professional comes with a license for 1-2 CPUs. Above that you need to purchase one of the server edition, and it starts becoming *very* expensive.
Soon 2 CPUs will be for the masses, they already are with hyperthreading in a way. However 4 and above really are for servers; multi-user environments, etc.
Also while it is easy to exploit 2 CPUs in a desktop environment (roughly speaking 1 for the O/S, the other for the applications) there are diminishing returns for more than 2 CPUs, except for parallel applications, and there aren't that many of those.
This raises questions regarding stability and Windows.
While I find that multiprocs settings under Linux improve things to a significant degree (although there are still outstanding issuess with NVidia proprietary drivers and SMP), I found the opposite true for Windows.
The last time I tried, which was about 2-3 years ago, many drivers didn't seem to expect true concurency under Win2k and I was experiencing significantly more crashes on my dual P-III than when I forced the system to only use one of the CPUs. Yet it probably wasn't the hardware because that same machine was very stable with Linux.
With the advence of hyper-threading, have things improved markedly with WinXP?
Well the "I have nothing to hide or fear" argument is an old one, but it doesn't make it valid.
Currently to read my number plate more or less requires a human. A few systems are able to read my plate in a few selected places, but I don't really mind that.
With RFID it would finally become cheap and effective to track everybody everywhere. You could get a fine everytime you exceed the local speed limit, it would be all automated and reliable, which would be terribly annoying but maybe somewhat justifiable.
However how do we know that this situation will only be used for traffic regulation enforcement? Will it be tamper-proof? will you have the right to inspect your record? will the data be wiped out after a given time frame? Who will have the right to subpoeana the record?
The right to privacy is also extremely important. I don't want companies or agencies to look at my driving patterns and deduce any kind of behavioural conclusion without my approval.
Imagine for example that you are innocent but suspect of some kind of crime and that where you go and your surroundings are closely monitored. No one would like to go to your place or you to visit them for fear of being associated with you somehow and also questioned by police, say.
Other not-so-nice scenarios are not that hard to imagine.
All sorts of questions one would like answered before signing on to that sort of deal.
If Saddam Hussein had been using DU weapons, for sure they would have been labelled as WMDs. Uranium! Radiations! eeek, run!
Just look how the single, 10 year old Sarin gas cannister that was found was labelled.
In pure dollar amounts you may be right, however it is still not a good trade because it will undermine incentives to develop IP and content at home. The Australian cinema you love will be even more challenged than it is now, and our beleagered research and IT sectors will suffer some more. Did you know that 96% of movies released in Australia come from Hollywood? The 4% remaining are for the rest of the world, including Australian cinema. Talk about cultural hegemony.
What this FTA is likely to do is play up our export strengths which are in the commodities business (primary resources) and undermine our weaknesses further (intellectual content). However commodities are relatively low-value, finite, volatile and not under our control (Australia doesn't get to set the price of coal for example). It might turn out OK, it may not ; conversely unique intellectual content is high-value, well protected, renewable and much more under control.
In the future we want to rely more on the former and less on the latter. The FTA is not going in this direction and this is dangerous.
Actually I wish this agreement would make us the 51st state, because then we would have some rights coming our way such as a proper constitution and a bill of right.
Instead this FTA takes away a lot of things, almost guarantees more litigation, invites for a larger trade deficit in favour of the US in the high-tech sector in return for some very meager returns in the Agricultural sector.
People say the balance will be somewhat even on the financial side all being considered, but in the meantime the rights of every Australians will be eroded with nothing to compensate.
Some content currently in the public domain in Australia will revert to being copyrighted. For example in Australia we have an event called Bloomsday, in honour of Ulysses, the book by James Joyce. This is a day of celebration of Joyce's memory, with new plays based on Joyce's book being peformed and things like that. This has only recently been possible because the estate of James Joyce generally forbids any adaptation of Joyce's novels for any purpose.
Of course the estate can do nothing about stuff that is in the public domain. But with the FTA, Ulysses and other works will return to the estate and we will have to wait another 15 years for the next bloomsday, or longer if yet another copyright extention is granted down the track.
I think this is mad. This sort of thing does not serve the public.
500h?
Given there are only 168 hours in a week I find this statement rather puzzling.
In Australia you may sometimes copy software for backup purposes, but you cannot backup music or video or any other copyrighted material even for personal purposes. Precisely what is purchased when buying a tape or CD or DVD is a license to play that content on approved harware using the medium provided only. If it breaks down you have to buy another one, sorry.
Photocopy of printed material is also very tightly controlled. At my place of work you cannot make a copy of anything yourself. You have to ask the librarian and fill a form. I am not joking. For example you are allowed to copy up to 10% or one article of one magazine or journal, whichever is less, for research or teaching purposes. Personal use is not allowed.
Fair use? don't know what that means.
90% ? Where did you get this number? this is nonsense, sorry. The TOTAL contribution of ALL agricultural products to the Australian GDP is 3%. By comparison services amount to 71%.
See Australian profile
As for exports, Australia primary sector (commodities such as coal, gold, meat, wool, alumina, iron ore, wheat, machinery and transport equipment) amount to 65% of all exports.
I don't know where you got this 90% from.
It's not as hard as it sounds, especially with a decent supervisor. The hard bit about science is not solving problem, it's knowing which one to solve. Every field is literally littered with unsolved questions and problems. Which ones are sufficiently hard to be interesting yet easy enough to be cracked in 3 years of work or less? It is the supervisor's work to come up with these, not the student's.
At the moment the MSFT share price has stalled. They are looking for big markets and they are not finding them easy to break into. If Linux starts making inroads in the desktop in the next couple of years, I would expect Microsoft to start playing desperate tactics. It might be the beginning of the end but it would be nasty nonetheless.
If
Servers are a tiny market compared with desktop, and are not software-bound. I don't believe they would use pseudo-interoperability as an argument to buy Microsoft. If they want they already have a portable environment for providing Unix-style servers, it used to be called Interix and it works very well. My interpretation of things as they stand is that
Finally Microsoft gave away the specs of C# to ECMA for standardisation, but see above discussion about licensing terms. They are at least RAND, but no one seems to know for sure if they are royalty-free. In other words Microsoft hasn't given MSIL away, on the contrary, it could be a neat poison pill.
Microsoft will be able to take out Mono any time they choose.
Microsoft now has a very strong research group located in various places around the globe: China, Europe, Australia, US, etc. These guys don't just innovate, they are in the process of inventing new and ambitious things. These guys are active in all the areas of computer science from computer vision to advanced databases.
The problem is that sometimes big things start small. Look at the PC industry. It really started with home hardware kits, and now look where it's at. This shouldn't be a problem, but Microsoft is not interested in small things. They only want to corner billion dollar markets: they don't want to defocus their interest into thousands of little projects, only a handful of which will make it into a multi-million-dollar industry, and potentially one or two would break the billion dollar threshold. They don't have the patience.
The reason for this is that they are the company with the largest market cap in history, and they need to deliver value for the shareholders. A new million-dollar Microsoft startup wouldn't even register on the investor's radar. Too many of these little companies failing would devalue the stock. The board would be accused of wasting the shareholders' money.
The associated problem is that there aren't so many multi-Billion (with a B) markets around that are somehow associated with Microsoft core business. Microsoft is not in energy distribution or transport, they would probably suck at managing that (Microsoft Air, who would get on board?). That is why they are now fighting with the Sony, Palm and Google of this world. They want those markets and they want them bad.
The problem is that they don't have a good customer relation image (they've worked very hard at appearing ruthless and uncaring, it seems) and they don't have a good track record of breaking into market where they don't have a monopoly. In these markets the traditional Microsoft tactics don't work (strongarm everybody). That is why they are currently losing lots of money in the console market.
It is interesting to watch that despite Bill Gates' team famed smarts and they billions in cash in the bank, Microsoft is still having a hard time owning the world.
It will be interesting to see how it all turns out. Can Microsoft break the conundrum of the big rich company hamstrung by its shareholders?
You are writing replies that are even longer than mine, quite a feat for /.
Yes ANN only generate computable outputs, at least the ones we have now. But is the same true for NNN (the Natural kind)? The answer is currently a matter of beliefs. If not, no amount of Moore's law is going to help.
I used to belive just like you that true AI was trivially provably possible by emulation, but now I'm not so sure.
Concerning a generation that doesn't die anymore, they might be stuck with interesting problems. What if your life was very very long but incredibly boring and frustrating, because the society around has ossified: no one can get promoted, new ideas are not wanted. Old men looking like they are 20 year olds are in power and want to remain so forever. It could in fact suck quite a bit.
When Niels Bohr was asked how he had managed to convince his physicist colleagues that his quantum ideas were correct, he replied that he never managed to do it. His ennemy belonged to the previous generation of physicists and simply eventually retired or died. They never accepted his work. In a society where no one dies, it could mean the end of many things.
I'm a believer of the incredible power of true science, but I've also been a witness of the effects of the results of science in the wrong hands, and so have been quite a few people.
Anyway, hopefully we'll see. All the best.
You could be temporarily incapacitated. Hit by a bus and spend a few months in a coma. It would be nasty to wake up and find out your wife divorced you and your kids don't want to talk to you and you have no money left because of a few files.
I haven't missed your point, on the contrary. I know about the hippocampus story, and they have only emulated a tiny portion of the suborgan. The hippocampus is a little bit bigger than 18 neurons.
There are people who believe that simply putting together huge numbers of artificial neurons together will not be enough. They say that non-computable quantum effects occur in the brain and that we will need to emulate those to reproduce thinking. This is what most proeminently Roger Penrose belives. It is very difficult to simulate quantum effects. As you probably know, one can use quantum effects to do very quick calculations, for example factoring using Shor's algorithm. The converse of that is that it is very inefficient to reproduce quantum effects on traditional computers.
Results from people who work with artificial neural network seem to bear that out. In final analysis, an ANN doesn't perform anything widely different than what traditional statistical classification does, which is a normal deterministic computable task. In other words non-computable behaviours don't emerge spontaneously from ANN. Maybe it is a function of the size of the ANN, but people are beginning to doubt this.
Maybe Penrose is wrong and a straight normal emulation will work, but no one knows for sure. Also maybe we will have quantum computers at our disposal in a few decades, but then again maybe not, and will those that we will have help bringing about AI? no one knows.
Yes Moore's law has held out for a very long time, it doesn't mean that it will continue to do so forever. The end of Moore's law has been predicted many times, but at some point we will be hitting some pretty strong barriers: how many molecules can we use to make a switch? how small can we shrink a die so that speed-of-light delays don't induce problems (at 10GHz, a speed which is on the horizon, light only travels 3cm)?
The fact is we are still using the same basic techniques for building CPUs since the 1960s. Techniques have improved and dies have shrunk, but the basic ideas have remained the same (die, clock, gates, etc). To overcome the size barriers we will probably have to switch to clockless designs, massive parallel architectures, and certainly vastly more efficient cooling that we have now. Moore's law may have a few years or even decades to go, but I'm willing to bet it won't hold out another 100 years. In fact it is likely that soon, the cooling system in computers will cost more than the CPU. Pumps and cryogenic systems are not amenable to Moore's law.
If you look at the way you programmed a computer in 1970, it wasn't vastly different to what people are doing now. Learning to program clockless designs and parallel designs is going to require massive retraining, and not all problems are going to be amenable to these.
I think I agree with you that the next few decades are going to bring about some massive changes. Call it a singularity if you want. I think besides the computing issues, energy, environmental and political issues will need a total rethink. As we are now, we are close to a disaster of vast proportion in the Middle East. We survived the cold war, can we survive global terrorism? Are we going to start running out of oil in 10 years time or not? Some people show that the oil field discovery peaked at the end of the 80s and that inevitably the production peak will come. From then one whatever we do, we'll have less oil to work with. That's a sobering thought.
If the human race manages to not anihilate itself in the next 50 - 100 years, we might witness a number of totally extraordinary things, such as the first AI, self-reproducing benevolent nanotech, effective gene repairs, possibly a worldwide democracy, a fully explored solar system and vastly extended lifetimes. If we survive the next 100 years more or less intact, I'm more confident we can make it in the long run.
Possibly unlike you, I don't believe it will happen by itself. It will be hard work, and sometimes I worry we can't ma
reference
Let's say they were at the right place at the right time with the IBM-PC deal. Also Gates did get a killer deal regarding MS-DOS, he could license it to whoever he wanted, not just to IBM. Without that PC clones would have been running DR-DOS or something like that.
After that Gates did a magnificent backstab with OS/2, that was masterful. Pretend to cooperate with an ailing IBM once again to develop the new "DOS", price the SDK too high for casual developers so no community develops, and pull out Windows-3.0 without warning, that could run on *every* PC, not just the 286s and above unlike OS/2.
Right about then they started being extremely nasty with competitors, like Borland, Lotus, etc.