That often-repeated claim fails to explain why progress in AI has been so abysmally slow. Are people really that stupid?
I think we are that stupid. We've wasted more than fifty years and billions of dollars on the GOFAI symbolic approach.
Have we failed to see the answer lying in plain sight for so long?
I think that, when we finally find the answer, we will kick ourselves in the ass for having been so stupid.
We don't need more computing power exactly, we need a different type of computing power. The processing needs to be closer to the memory and massively parallel, but not necessarily very fast. Simulating this type of system on a von Neumann architecture just doesn't work very well.
I completely disagree. We can emulate neural networks and other parallel systems perfectly well with our von Neumann systems. If we need more computing power, we can always use clustering technology. That is not the problem. The problem is that we don't yet know what the principles of intelligence are. We will, soon enough, IMO. Only because a lot of people are now trying to solve the right problems, having learned the lessons of the past.
If mimicking the brain is your goal, then a completely different computer architecture is needed. You need a large memory with embedded massively parallel processing units. You need a non-von Neumann architecture to eliminate the von Neumann bottleneck. Only after this brain-like computer architecture is developed will we be able to implement true AI.
I think it will require a lot less processing power than most people would think. It is known that the brain can focus on a very narrow subject/concept at a time. This ability alone can do wonders for limiting performance requirements. As far as DARPA's GC is concerned, I think it could be done with a brain a lot less powerful than our own. Consider that a honeybee has only about 1 million neurons and yet, its behavior is orders of magnitude more sophisticated than any robot or AI program in existence, including any of the robots participating in the challenge.
I think that a brain with a few million neurons and true AI could easily win the Grand Challenge. It does not have to understand English, talk, mate, etc... It just needs to know how to navigate a course in the desert. For example, it does not even need to have human visual acuity. An eye with 16,000 pixel/detector resolution and 8 shades of grey would do fine. It just needs to be able to learn from experience and adapt via reinforcement.
The often repeated claim that we need more computer power to do real AI is a copout, IMO. It's mostly coming from people at CMU such as Hans Moravec. We could do amazing things with what we have, if only we knew how.
If "true AI" is the best way to achieve results, then the people who use it will win. If it is not, then requiring it would be counterproductive.
I see what you mean but I have to disagree. By not requiring learning systems, DARPA is not encouraging progress in AI. In fact, it is discouraging it because robot people love to tinker with their robots by progamming the hehaviors themselves instead of giving the machines the ability to acquire their own behavior through trial and error. The US defence department would sell its soul for a truly intelligent system and that's what we should be after. DARPA's GC is not going to give it to them.
Instead of spending $100 billion to go back to the moon and send people to Mars, part of the money should be given to DARPA. They should increase the Grand Challenge Prize to $10 Billion, change the rules to prohibit non-learning systems and include big-city driving in the challenge. It would be the AI X-Prize, if you will. That would make it much more interesting and would advance the art tremendously , IMO. As it stands, all we're gonna get is clever engineering which we already know we're good at, but not good enough.
Driving across 150 miles of roadless, obstacle-ridden desert is not something most humans do, or even attempt. Don't be so sure that "even severely IQ handicapped humans" could handle it routinely.
I think that driving around a big city (New York, London, Paris, etc...) is much harder than driving around the desert, orders of magnitude harder, IMO. Especially during rush hour.
Yes, because being able to take two dimensional sensory input and use it to construct an acccurate three-dimensional representation of the local surroundings, and then plan a viable route through those surroundings, is not a trivial task.
That's the CMU team's approach. Rodney Brooks (MIT AI Lab Director) has shown with his subsumption architecture that this is precisely how not to do it. The coupling between sensors and effectors should be as short as possible, especially when your processors (neurons) are very slow. In fact, as Brooks pointed out in an interview with Edge.org, the connectivity diameter of the brain is no more than 5 or 6 neurons, that is, fron sensor layer to motor layer. Not nearly enough time to do the sort of serial processing needed to construct a 3-D model of the environment in real time. The brain is a reactive system, more than anything else. It learns to react appropriately from experience.
Did AI research implode for lack of funding, or is it really that hard?
None of the competitors are doing true AI. They are not using learning systems as far as I know. This is just good old fashioned programming where the designers/programmers try to think of all possibilities in advance. I don't see how this contest is advancing our understanding of intelligence. I think that the qualifying rules should have been more stringent and should have prohibited non-learning systems. Otherwise it's the same old traditional stuff.
No, the White House policy effectively prohibits using embryonic stem cells.
This is very much in doubt since it is legal to conduct embryonic stem cell research in the US. Some states (e.g., California) provide their own funding for embryonic stem cell research, in effect circumventing the federal ban on funding.
However, giving them the benefit of the doubt, it is too bad the field of stem cell research in the US has been badly damaged by policies the current Whitehouse administration have put into place.
The Korean researchers used umbilical stem cells, not embryonic cells from a fetus. It is a lie to insinuate that the white house forbids stem cell research. It only forbids federal grants to researchers using cells from aborted fetuses. Why be so disingenuous?
Software is in the same disconnected state. We build modules but we don't build their connections, on the relationships between these modules.
Yes indeed. But it goes deeper than the module level. The problem also exists at the roots of software, i.e., at the operations level. Modifying a variable should be immediately broadcasted to every part of the program that depends on the variable. This capability must be an inherent part of the software construction tools and should not be left to the programmer. Ineffective communication is 90% of the software reliability problem. It is especially serious in legacy systems after the old programmers are long gone.
It sounds like you have never heard of object-oriented programming - or even event-driven programming.
Object-oriented programming, while useful for software composition, does not solve the data dependency problem which is the biggest contributor to defective software. I have nothing against OOP. Event-driven programming is not strictly synchronous and does not go far enough. To solve the data-dependency problem at the program level, every effector operation (an operation that can modify data) must generate an event. I suggest you take a closer look at synchronous reactive programming (do a search on Google) so you can become more familiar with signal-based software in general.
Every piece of software starts with a clean, elegant structure - in the mind of whoever created it. Over time some of their assumptions prove false, and more importantly, many of the "true believers" who originally engineered the system move on.
FTFA:...it became harder to strap new features onto the software since new code could affect everything else in unpredictable ways.
The problem is a communication problem, not between programmers (nothing can really be done about that since they come and go) but between different parts of a complex software system. It has to do with data dependencies, not only at the program level, but also at the system level. It's a matter of the left hand not knowing what the right hand is doing. The problem is proportional with complexity and it affects the entire software development industry, not just Microsoft. But is does not have to be that way. There is a solution, one which, unfortunately will require a fundamental rethinking of software construction. It's never too late to retrace one's steps. See my site for more.
VCR's, and DVD players have deep market penetration already - there is continuing growth, but its not what it was, and its much lower margin than when the technology is new and booming. Same with PS and PS2 - most everyone who wanted one has one by now.
In my opinion, Sony has fallen on hard times for the same reason that Zerox did several years ago. Their old patents expired and the competition has begun to compete on a level playing field. Serves them right, though. Rather than putting more money into research and development, they decided to get into the movie business instead. They only have themselves to blame.
There hasn't been a single article from either side saying 'cut off funding for the other'. All the scientists agree that 'more research, more funding, more computer models, etc. ..' are needed.
Are you saying that scientists are human just like lawyers? I'm also reminded of doctors and mechanics. Who would benefit if nobody got sick and cars fixed themselves?
In reality, languages have rational syntax in the sense that people can understand them, but none have the sort of rational syntax that a computer is good at understanding.
True. This means that computers should be more like people. That is to say, they should be intelligent. So, short of having a truly intelligent machine that can learn a language the way children do, a really good grammar checker is still in the future. But who knows, true AI may be just around the corner.
The whole IP stuff needs to be reviewed and changed, but as long as it's in place, the only option is to pick up the sword and defend yourself.
You're right, of course. Asking a company not to use IP laws would be like a prisonner not to eat prison food. One is forced to use the system currently in place. My beef is that none of the big players are in a hurrey to see the system changed. Why? Because the system allows them to make a shitload of money for comparatively little work. Patent a simple little algorithm, wait for some other guy to start making money with a product and sue him for a pice of the action. Some companies don't even make any products to sell. All they do is accumulate patents. Unmitigated greed will eventually destroy society as we know it.
Yes they can. They just need the right law firm. Google has a lot of money and everybody wants a piece of it. The IP hoarders deserve to be constantly at each other's throats, IMO. If you live by the sword, you will perish by the sword. Well, you will at least get hurt. Didn't Oracle have to pay 100 million US dollars or so to settle an IP-related suit just recently? Companies are like kids fighting over who the toys belong to. When will they wake up and grow up? It is obvious that the IP laws are stupid. So isn't something done about it? Answer: Greed.
I say let them rip out one another's throats. Problem is, the lawyers benefit immensely from this crap. And as long as lawyers are making both the laws and a shitload of money, we (the public) will continue to pay the price. In the meantime, the rich gets richer and the poor gets shafted. Sorry for being so cynical.
I look forward to the day when we relinquish all control of our cars once we enter the freeway.
Yeah. Problem is, manufacturers cannot guarantee that their software is bug-free. This is the reason that automated cars are legally banned in most of Europe and other parts of the world.
Software unreliability is a huge problem, one which is preventing us from achieving the full promise of the computer age. The more complex our systems the more unreliable they are. The reason is that there is something fundamentally wrong with the way we create software. Contrary to conventional wisdom, unreliability is not an essential characteristic of complex software programs.
The solution will require a radical change in the way we program our computers. In my opinion, the main reason that software is so unreliable has to do with a custom that is as old as the computer: the practice of using the algorithm as the basis of software construction Switching to a signal-based, synchronous software model will not only result in an improvement of several orders of magnitude in productivity, but also in programs that are guaranteed free of defects, regardless of their complexity.
Cringely is a Slashdot Karma Whore
on
Has Google Peaked?
·
· Score: 2, Interesting
Why? Because he hates Microsoft and loves Apple. Both MS and Google have a shitload of money to invest and both have tons of stuff they can do with their money. But if Google really wants to take on both MS and Apple, and even Intel, I think I got an excellent suggestion for them.
In my opinion, Intel and the rest of the big processor vendors can only come up with so many incremental improvements before they bore the market to death. Microsoft is mired in buggy code that they'll never be able to fix. Apple is playing second fiddle in the market. So what comes next?
I suggest that Google starts working on the biggest problem facing the computer industry today: unreliable software. It's costing us billions of dollars and even human lives. Consider that the basic architecture of the processor has not changed in more than 150 years, when a guy named Babbage and his girlfriend Ada built their mechanical computer around the "table of instructions". All processor architectures have been based on and optimized for the algorithm ever since.
A truly innovative architecture would abandon the algorithmic model altogether and embrace a non-algorithmic, signal-based synchronous software model. It would not only revolutionize the computer industry, it would solve its nastiest problem: software unreliability.
But can we really expect the big guys (Intel, AMD, IBM, etc...) to be truly innovative at this stage of the game? Their approach is evolutionary, not revolutionary; and they are doing just fine as it is. They have no great incentive to change. Hopefully, a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it's too late. The message is simple: There is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. The advantage is too good to ignore: 100% software reliability! Guaranteed!
This is the stuff that revolutions and great companies are made of. After a century and a half, I think it's time for a change. He who has an ear (and the venture capital) let him hear!
The reason chips have so few errors is because chip problems are *incredibly* expensive to fix (have to rip them off the circuit boards and solder on new ones), so chip companies take a *lot* of care to eliminate errors before shipping, or they go out of business. Software companies know they can let users find errors and send out a patch later for next to nothing, so they don't spend as much on QA. Plus software consumers have low expectations for quality, and high expectations for release schedules - they want the program now, even if it's not quite done.
The reason that it is so expensive to design and build harware has little to do with hardware logic. It has to do with the physical aspects of circuit layout and fabrication. Hardware failures are almost always physical failures and almost never logic failures. Logic failures are nipped in the bud because they are easy to find. In fact, they are found during simulation. This is because of the signal-based synchronous nature of logic. A computer system consists of many different chips interacting in very complex ways and yet they almost never fail due to logic defects. This is true regardless of complexity.
Sorry to be the one to tell you this, but after checking out your website, it looks like you've wasted a lot of your life on a crackpot theory. Sucks to be you!
One man's opinion, of course. A lot of people disagree with your point of view. I know, they tell me so. Peace and love to you.
In my opinion, Intel and the rest of the big processor vendors are running out of ideas. They can only come up with so many incremental improvements before they bore the market to death. So what comes next?
I suggest that they start working on the biggest problem facing the computer industry today: unreliable software. It's costing us billions of dollars and even human lives. Consider that the basic architecture of the processor has not change in more than 150 years, ever since a guy named Babbage and his girlfriend Ada built their mechanical computer around the "table of instructions". All processor architectures have benn based on and optimized for the algorithm ever since.
A truly innovative architecture would abandon the algorithm and embrace a non-algorithmic, signal-based synchronous software model. It would not only revolutionize the computer industry, it would solve its nastiest problem: software unreliability.
But can we really expect the big guys (Intel, AMD, IBM, etc...) to be truly innovative at this stage of the game? Their approach is evolutionary, not revolutionary and they are doing just fine as it is. They have no great incentive to change. Hopefully, a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it's too late. The message is simple: There is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. The advantage is too good to ignore: 100% software reliability! Guaranteed!
This is the stuff that revolutions and great companies are made of. After a century and a half, I think it's time for a change. He who has an ear (and the venture capital) let him hear!
Hopefully this will give nex-gen AMD chips a fresh design and hopefully push them to a significant majority over Intel.
This will not happen. Intel's marketing prowess is much better than its competition. What would scare Intel (and the others) is a revolutionary new chip that solves a major problem in the industry. Consider that all processor architectures are based on and optimized for the algorithm, a custom started by a guy named Babbage more than 150 years ago. Progress has only been incremental since.
A really new architecture should abandon the algorithmic model and adopt a non-algorithmic, signal-based synchronous software model. It would revolutionize computing and solve the nastiest problem in the computer industry: software unreliability.
But we cannot expect big companies like Intel, AMD and IBM to be truly innovative. Their approach is evolutionary, not revolutionary. Hopefully a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it is too late.
The message is that there is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. But the advantage is too good to ignore: 100% software reliability! Guaranteed!
I don't think so. The Cell Processor is really a multicore processor. It encapsulates a number small bundles of algorithmic code and assigns them to multiple cores for concurrent processing. It's an interesting processor but I would not call it non-algorithmic. A truly synchronous, signal-based processor would be optimized for discrete signal communication between elementary cells/operations right out of the gate.
and it comes from some mighty big players.
Yep. I was surprised. But it does not go far enough. The revolution is still out there.
All processor architectures are based on the algorithm, a custom started by a guy named Babbage some 150 years ago.
A really new architecture should abandon the algorithmic model and adopt a non-algorithmic, signal-based synchronous software model. It would revolutionize computing and solve the nastiest problem in the computer industry: software unreliability.
But we cannot expect a big company like Intel to be truly innovative. Hopefully a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. The won't know what hit them until it's too late.
Compare that to CCS or CSP which actually provides an algebra via which you can logially reason about the system and hence formally mathematically prove things about the system... I don't see how COSA offers anything comparable - unless I've missed something.
Math has nothing to do with it. COSA solves the nastiest problem of complex software programs: blind code or unresolved data dependencies. That is, something is modified in one part of the program unbeknownst to another. This makes it almost impossible to modify complex code without introducing dangerous and unforeseen side effects. The second greatest cause of unreliability is non-deterministic timing. COSA solves this problem through synchronicity.
All other programming problems are, as Fred Brooks put it, are accidental and can be easily dealt with using traditional methods.
COSA is revolutionary because it guarantees 100% reliability. No other model comes close.
As an aside: I suggest that you spend some time looking into concurrency theories such as CSP, CCS, and the pi-calculus.
Academia has turned software construction into a complete babelized mess over the last fifty years. Software engineering will not come of age until a single software construction model is universally adopted. Academics love to complicate the hell out of everything because they have to obtain grants and write their theses.
COSA is trying to be as simple as possible. The COSA motto is: If it is not simple, it's wrong.
Slashdot Mirror
That often-repeated claim fails to explain why progress in AI has been so abysmally slow. Are people really that stupid?
I think we are that stupid. We've wasted more than fifty years and billions of dollars on the GOFAI symbolic approach.
Have we failed to see the answer lying in plain sight for so long?
I think that, when we finally find the answer, we will kick ourselves in the ass for having been so stupid.
We don't need more computing power exactly, we need a different type of computing power. The processing needs to be closer to the memory and massively parallel, but not necessarily very fast. Simulating this type of system on a von Neumann architecture just doesn't work very well.
I completely disagree. We can emulate neural networks and other parallel systems perfectly well with our von Neumann systems. If we need more computing power, we can always use clustering technology. That is not the problem. The problem is that we don't yet know what the principles of intelligence are. We will, soon enough, IMO. Only because a lot of people are now trying to solve the right problems, having learned the lessons of the past.
If mimicking the brain is your goal, then a completely different computer architecture is needed. You need a large memory with embedded massively parallel processing units. You need a non-von Neumann architecture to eliminate the von Neumann bottleneck. Only after this brain-like computer architecture is developed will we be able to implement true AI.
I think it will require a lot less processing power than most people would think. It is known that the brain can focus on a very narrow subject/concept at a time. This ability alone can do wonders for limiting performance requirements. As far as DARPA's GC is concerned, I think it could be done with a brain a lot less powerful than our own. Consider that a honeybee has only about 1 million neurons and yet, its behavior is orders of magnitude more sophisticated than any robot or AI program in existence, including any of the robots participating in the challenge.
I think that a brain with a few million neurons and true AI could easily win the Grand Challenge. It does not have to understand English, talk, mate, etc... It just needs to know how to navigate a course in the desert. For example, it does not even need to have human visual acuity. An eye with 16,000 pixel/detector resolution and 8 shades of grey would do fine. It just needs to be able to learn from experience and adapt via reinforcement.
The often repeated claim that we need more computer power to do real AI is a copout, IMO. It's mostly coming from people at CMU such as Hans Moravec. We could do amazing things with what we have, if only we knew how.
And, no, a soccer mom with a big mac on her knee talking on a cell phone couldn't traverse the course that they'll be on.
IMO, a soccer mom could do MUCH better than that, after proper training and sufficient practice in desert terrains.
If "true AI" is the best way to achieve results, then the people who use it will win. If it is not, then requiring it would be counterproductive.
I see what you mean but I have to disagree. By not requiring learning systems, DARPA is not encouraging progress in AI. In fact, it is discouraging it because robot people love to tinker with their robots by progamming the hehaviors themselves instead of giving the machines the ability to acquire their own behavior through trial and error. The US defence department would sell its soul for a truly intelligent system and that's what we should be after. DARPA's GC is not going to give it to them.
Instead of spending $100 billion to go back to the moon and send people to Mars, part of the money should be given to DARPA. They should increase the Grand Challenge Prize to $10 Billion, change the rules to prohibit non-learning systems and include big-city driving in the challenge. It would be the AI X-Prize, if you will. That would make it much more interesting and would advance the art tremendously , IMO. As it stands, all we're gonna get is clever engineering which we already know we're good at, but not good enough.
Driving across 150 miles of roadless, obstacle-ridden desert is not something most humans do, or even attempt. Don't be so sure that "even severely IQ handicapped humans" could handle it routinely.
I think that driving around a big city (New York, London, Paris, etc...) is much harder than driving around the desert, orders of magnitude harder, IMO. Especially during rush hour.
Yes, because being able to take two dimensional sensory input and use it to construct an acccurate three-dimensional representation of the local surroundings, and then plan a viable route through those surroundings, is not a trivial task.
That's the CMU team's approach. Rodney Brooks (MIT AI Lab Director) has shown with his subsumption architecture that this is precisely how not to do it. The coupling between sensors and effectors should be as short as possible, especially when your processors (neurons) are very slow. In fact, as Brooks pointed out in an interview with Edge.org, the connectivity diameter of the brain is no more than 5 or 6 neurons, that is, fron sensor layer to motor layer. Not nearly enough time to do the sort of serial processing needed to construct a 3-D model of the environment in real time. The brain is a reactive system, more than anything else. It learns to react appropriately from experience.
Did AI research implode for lack of funding, or is it really that hard?
None of the competitors are doing true AI. They are not using learning systems as far as I know. This is just good old fashioned programming where the designers/programmers try to think of all possibilities in advance. I don't see how this contest is advancing our understanding of intelligence. I think that the qualifying rules should have been more stringent and should have prohibited non-learning systems. Otherwise it's the same old traditional stuff.
No, the White House policy effectively prohibits using embryonic stem cells.
This is very much in doubt since it is legal to conduct embryonic stem cell research in the US. Some states (e.g., California) provide their own funding for embryonic stem cell research, in effect circumventing the federal ban on funding.
However, giving them the benefit of the doubt, it is too bad the field of stem cell research in the US has been badly damaged by policies the current Whitehouse administration have put into place.
The Korean researchers used umbilical stem cells, not embryonic cells from a fetus. It is a lie to insinuate that the white house forbids stem cell research. It only forbids federal grants to researchers using cells from aborted fetuses. Why be so disingenuous?
Software is in the same disconnected state. We build modules but we don't build their connections, on the relationships between these modules.
Yes indeed. But it goes deeper than the module level. The problem also exists at the roots of software, i.e., at the operations level. Modifying a variable should be immediately broadcasted to every part of the program that depends on the variable. This capability must be an inherent part of the software construction tools and should not be left to the programmer. Ineffective communication is 90% of the software reliability problem. It is especially serious in legacy systems after the old programmers are long gone.
It sounds like you have never heard of object-oriented programming - or even event-driven programming.
Object-oriented programming, while useful for software composition, does not solve the data dependency problem which is the biggest contributor to defective software. I have nothing against OOP. Event-driven programming is not strictly synchronous and does not go far enough. To solve the data-dependency problem at the program level, every effector operation (an operation that can modify data) must generate an event. I suggest you take a closer look at synchronous reactive programming (do a search on Google) so you can become more familiar with signal-based software in general.
Every piece of software starts with a clean, elegant structure - in the mind of whoever created it. Over time some of their assumptions prove false, and more importantly, many of the "true believers" who originally engineered the system move on.
...it became harder to strap new features onto the software since new code could affect everything else in unpredictable ways.
FTFA:
The problem is a communication problem, not between programmers (nothing can really be done about that since they come and go) but between different parts of a complex software system. It has to do with data dependencies, not only at the program level, but also at the system level. It's a matter of the left hand not knowing what the right hand is doing. The problem is proportional with complexity and it affects the entire software development industry, not just Microsoft. But is does not have to be that way. There is a solution, one which, unfortunately will require a fundamental rethinking of software construction. It's never too late to retrace one's steps. See my site for more.
VCR's, and DVD players have deep market penetration already - there is continuing growth, but its not what it was, and its much lower margin than when the technology is new and booming. Same with PS and PS2 - most everyone who wanted one has one by now.
In my opinion, Sony has fallen on hard times for the same reason that Zerox did several years ago. Their old patents expired and the competition has begun to compete on a level playing field. Serves them right, though. Rather than putting more money into research and development, they decided to get into the movie business instead. They only have themselves to blame.
There hasn't been a single article from either side saying 'cut off funding for the other'. All the scientists agree that 'more research, more funding, more computer models, etc. . .' are needed.
Are you saying that scientists are human just like lawyers? I'm also reminded of doctors and mechanics. Who would benefit if nobody got sick and cars fixed themselves?
In reality, languages have rational syntax in the sense that people can understand them, but none have the sort of rational syntax that a computer is good at understanding.
True. This means that computers should be more like people. That is to say, they should be intelligent. So, short of having a truly intelligent machine that can learn a language the way children do, a really good grammar checker is still in the future. But who knows, true AI may be just around the corner.
The whole IP stuff needs to be reviewed and changed, but as long as it's in place, the only option is to pick up the sword and defend yourself.
You're right, of course. Asking a company not to use IP laws would be like a prisonner not to eat prison food. One is forced to use the system currently in place. My beef is that none of the big players are in a hurrey to see the system changed. Why? Because the system allows them to make a shitload of money for comparatively little work. Patent a simple little algorithm, wait for some other guy to start making money with a product and sue him for a pice of the action. Some companies don't even make any products to sell. All they do is accumulate patents. Unmitigated greed will eventually destroy society as we know it.
I don't think they can fight Google...
Yes they can. They just need the right law firm. Google has a lot of money and everybody wants a piece of it. The IP hoarders deserve to be constantly at each other's throats, IMO. If you live by the sword, you will perish by the sword. Well, you will at least get hurt. Didn't Oracle have to pay 100 million US dollars or so to settle an IP-related suit just recently? Companies are like kids fighting over who the toys belong to. When will they wake up and grow up? It is obvious that the IP laws are stupid. So isn't something done about it? Answer: Greed.
I say let them rip out one another's throats. Problem is, the lawyers benefit immensely from this crap. And as long as lawyers are making both the laws and a shitload of money, we (the public) will continue to pay the price. In the meantime, the rich gets richer and the poor gets shafted. Sorry for being so cynical.
I look forward to the day when we relinquish all control of our cars once we enter the freeway.
Yeah. Problem is, manufacturers cannot guarantee that their software is bug-free. This is the reason that automated cars are legally banned in most of Europe and other parts of the world.
Software unreliability is a huge problem, one which is preventing us from achieving the full promise of the computer age. The more complex our systems the more unreliable they are. The reason is that there is something fundamentally wrong with the way we create software. Contrary to conventional wisdom, unreliability is not an essential characteristic of complex software programs.
The solution will require a radical change in the way we program our computers. In my opinion, the main reason that software is so unreliable has to do with a custom that is as old as the computer: the practice of using the algorithm as the basis of software construction Switching to a signal-based, synchronous software model will not only result in an improvement of several orders of magnitude in productivity, but also in programs that are guaranteed free of defects, regardless of their complexity.
Why? Because he hates Microsoft and loves Apple. Both MS and Google have a shitload of money to invest and both have tons of stuff they can do with their money. But if Google really wants to take on both MS and Apple, and even Intel, I think I got an excellent suggestion for them.
In my opinion, Intel and the rest of the big processor vendors can only come up with so many incremental improvements before they bore the market to death. Microsoft is mired in buggy code that they'll never be able to fix. Apple is playing second fiddle in the market. So what comes next?
I suggest that Google starts working on the biggest problem facing the computer industry today: unreliable software. It's costing us billions of dollars and even human lives. Consider that the basic architecture of the processor has not changed in more than 150 years, when a guy named Babbage and his girlfriend Ada built their mechanical computer around the "table of instructions". All processor architectures have been based on and optimized for the algorithm ever since.
A truly innovative architecture would abandon the algorithmic model altogether and embrace a non-algorithmic, signal-based synchronous software model. It would not only revolutionize the computer industry, it would solve its nastiest problem: software unreliability.
But can we really expect the big guys (Intel, AMD, IBM, etc...) to be truly innovative at this stage of the game? Their approach is evolutionary, not revolutionary; and they are doing just fine as it is. They have no great incentive to change. Hopefully, a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it's too late. The message is simple: There is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. The advantage is too good to ignore: 100% software reliability! Guaranteed!
This is the stuff that revolutions and great companies are made of. After a century and a half, I think it's time for a change. He who has an ear (and the venture capital) let him hear!
The reason chips have so few errors is because chip problems are *incredibly* expensive to fix (have to rip them off the circuit boards and solder on new ones), so chip companies take a *lot* of care to eliminate errors before shipping, or they go out of business. Software companies know they can let users find errors and send out a patch later for next to nothing, so they don't spend as much on QA. Plus software consumers have low expectations for quality, and high expectations for release schedules - they want the program now, even if it's not quite done.
The reason that it is so expensive to design and build harware has little to do with hardware logic. It has to do with the physical aspects of circuit layout and fabrication. Hardware failures are almost always physical failures and almost never logic failures. Logic failures are nipped in the bud because they are easy to find. In fact, they are found during simulation. This is because of the signal-based synchronous nature of logic. A computer system consists of many different chips interacting in very complex ways and yet they almost never fail due to logic defects. This is true regardless of complexity.
Sorry to be the one to tell you this, but after checking out your website, it looks like you've wasted a lot of your life on a crackpot theory. Sucks to be you!
One man's opinion, of course. A lot of people disagree with your point of view. I know, they tell me so. Peace and love to you.
In my opinion, Intel and the rest of the big processor vendors are running out of ideas. They can only come up with so many incremental improvements before they bore the market to death. So what comes next?
I suggest that they start working on the biggest problem facing the computer industry today: unreliable software. It's costing us billions of dollars and even human lives. Consider that the basic architecture of the processor has not change in more than 150 years, ever since a guy named Babbage and his girlfriend Ada built their mechanical computer around the "table of instructions". All processor architectures have benn based on and optimized for the algorithm ever since.
A truly innovative architecture would abandon the algorithm and embrace a non-algorithmic, signal-based synchronous software model. It would not only revolutionize the computer industry, it would solve its nastiest problem: software unreliability.
But can we really expect the big guys (Intel, AMD, IBM, etc...) to be truly innovative at this stage of the game? Their approach is evolutionary, not revolutionary and they are doing just fine as it is. They have no great incentive to change. Hopefully, a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it's too late. The message is simple: There is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. The advantage is too good to ignore: 100% software reliability! Guaranteed!
This is the stuff that revolutions and great companies are made of. After a century and a half, I think it's time for a change. He who has an ear (and the venture capital) let him hear!
Hopefully this will give nex-gen AMD chips a fresh design and hopefully push them to a significant majority over Intel.
This will not happen. Intel's marketing prowess is much better than its competition. What would scare Intel (and the others) is a revolutionary new chip that solves a major problem in the industry. Consider that all processor architectures are based on and optimized for the algorithm, a custom started by a guy named Babbage more than 150 years ago. Progress has only been incremental since.
A really new architecture should abandon the algorithmic model and adopt a non-algorithmic, signal-based synchronous software model. It would revolutionize computing and solve the nastiest problem in the computer industry: software unreliability.
But we cannot expect big companies like Intel, AMD and IBM to be truly innovative. Their approach is evolutionary, not revolutionary. Hopefully a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. They won't know what hit them until it is too late.
The message is that there is a solution to the software reliability crisis. The disadvantage is that it will require a radical change in both processor architecture and software construction methodology. But the advantage is too good to ignore: 100% software reliability! Guaranteed!
Thats what the Cell Archtecture does,
I don't think so. The Cell Processor is really a multicore processor. It encapsulates a number small bundles of algorithmic code and assigns them to multiple cores for concurrent processing. It's an interesting processor but I would not call it non-algorithmic. A truly synchronous, signal-based processor would be optimized for discrete signal communication between elementary cells/operations right out of the gate.
and it comes from some mighty big players.
Yep. I was surprised. But it does not go far enough. The revolution is still out there.
All processor architectures are based on the algorithm, a custom started by a guy named Babbage some 150 years ago.
A really new architecture should abandon the algorithmic model and adopt a non-algorithmic, signal-based synchronous software model. It would revolutionize computing and solve the nastiest problem in the computer industry: software unreliability.
But we cannot expect a big company like Intel to be truly innovative. Hopefully a bright upstart will get the message and make a killing while the behemoths are busy fighting each other for market share. The won't know what hit them until it's too late.
Compare that to CCS or CSP which actually provides an algebra via which you can logially reason about the system and hence formally mathematically prove things about the system... I don't see how COSA offers anything comparable - unless I've missed something.
Math has nothing to do with it. COSA solves the nastiest problem of complex software programs: blind code or unresolved data dependencies. That is, something is modified in one part of the program unbeknownst to another. This makes it almost impossible to modify complex code without introducing dangerous and unforeseen side effects. The second greatest cause of unreliability is non-deterministic timing. COSA solves this problem through synchronicity.
All other programming problems are, as Fred Brooks put it, are accidental and can be easily dealt with using traditional methods.
Not exactly "revolutionary".
COSA is revolutionary because it guarantees 100% reliability. No other model comes close.
As an aside: I suggest that you spend some time looking into concurrency theories such as CSP, CCS, and the pi-calculus.
Academia has turned software construction into a complete babelized mess over the last fifty years. Software engineering will not come of age until a single software construction model is universally adopted. Academics love to complicate the hell out of everything because they have to obtain grants and write their theses.
COSA is trying to be as simple as possible. The COSA motto is: If it is not simple, it's wrong.