There's Fisher's fundamental law of natural selection, later generalized to Price's covariance and selection theorem. Although hideously misinterpreted at times, these are testable statements about what one expects to observe when natural selection is true.
It would then of course be equally legal to take your binaries, examine the machine- or bytecode, decompile it, rip out the good stuff (your new features) and re-assemble it back in firefox. Cumbersome, but perfectly doable.
After some reflection on this issue (after submitting the post), I do concede that, theoretically, the author of Psyco has a point. However, in theory there's no difference between theory and practice. In practice, there is.
I'm apparently debating something else, that profit maximization can be counterproductive to the goals of a foundation such as Gates'. I do think we're debating orthogonal issues, and I agree completely with your point that the Gates foundation cannot be held responsible for the full actions of the companies they have a share in. They can however be held responsible proportionally.
So, the full responsibility of the Oil companies raid on the Nigerian population should not be ascribed to the Gates Foundation. They should however know better to support than to support such a raid by whatever small percentage they own. Buying shares is taking ownership is taking responsibility. It simply is a black spot on the foundation, the size is a matter of debate.
Although good, RDP is just too expensive. First you have to port all your applications to their proprietary framework to be able to run over RDP. Luckily there is some gcc support for that, but still. Then you have to buy the license. That one is not too steep, but is single user. If you want multiple users to use RDP on the same box, with session persistence, you will have to buy the much more expensive server version of RDP, for which you pay per user. And even then, you are stuck with one persistent RDP session per user.
Given that, the foundation will just bleed dry if they don't invest for maximum profits. And the more profitable their investments, the more impact the foundation can have.
Maximum profits doesn't follow. You can easily aim to invest for 'reasonable', or 'sustainable' profit to avoid the foundation to bleed dry. There's no need to maximize. In particular, when the investments go directly against the purpose of the foundation, you would have a net negative, in that the existence of the foundation would do more harm than good.
If that were the case, the world would be better off when Gates' fortune would be spent on gambling, hookers and blow.
That's easily solvable. The US is big and divided enough to actually send in redneck armies into the east or west, to 'prevent those darned librels to sell out god's own country to the UN', or vice-versa get the Metropolitan troops to 'quell a fundamentalist revolution in Jesus land'. The army of the underlying party can be shipped overseas, or sent out to guard Alaska.
I was specifically targetting the dynamic languages that the OP was aiming at. Those with a dynamic OO syntax, and for which function dispatch was based on a lookup instead of a pointer dereference. Although extremely dynamic, Lisp is a completely different beast, not based on hashtable lookup, and my point doesn't hold for it.
I am however suprised at the speed of Python+Psycho though. It must have come a long way as last time I looked it was roughly twice as slow. One word of caution though with this combination: you need to refrain from using much of python's dynamicity to be able to let psycho do its job. Effectively psycho figures out type information at runtime to eliminate the overhead of dynamic dispatch.
Overall the score is: Java 2.64, Python + psyco 9.22 (factor 3), Python 26.66 (factor 10), Ruby 76.56 (factor 25).
I don't really think the trade-off is in static typing and catching errors. Static typing gains you speed. Python, Ruby and Javascript are complete hogs, simply because each member function call is a hash-lookup in the objects hashtable for finding the method that is bound to the name. This is the cost of dynamicity. In Java (and C++ for that matter), a member function call translates to a function pointer call, i.e., fast.
The end result is that Java with Hotspot runs in the same order as C/C++, maybe a factor of 1.5/2 slower (if you're careful). Pure Python and Ruby will always run one order (factor 20 for Python, 40 for Ruby) slower because a function call is an expensive operation.
The only way to get speed out of the dynamic languages is to mix them with C. Python is pretty good at that, and I consider PythonWithEmbeddedC a reasonable language, yet a bit clumsy.
Interesting that you manage to put in a push for Python in the same post that complains that Java is too slow. On its own, python is a complete pig, you need to move significant portions of the code to C to even remotely reach the speed of Java. Java is decently fast, and I'm pretty sure Google uses it for various things that (a) are not time-critical, and (b) need to run fast.
Best way to do it is 'clean-room'. You hire/ get a couple of guys together who provably haven't seen Code Sample A, and ask them to write something to the requirements of the API. You meticulously let them document every step of the way, and in the end, all you did was implement a public API.
As long as there are no people involved that have seen sample A, sample B is clean.
Butting in a bit late, but nevertheless: why do you think it is important that a machine can simulate the universe in real-time? It's pretty much a given that this is impossible, as any computational device is part of that same universe, and would thus need to be able to simulate itself in real-time. Ad infinitum. What's your point with this?
Strictly speaking, if you construct a halting problem around a commercial static code analyzer to prove they can't solve all problems, you would likely be slapped with a lawsuit (a) for creating an unauthorized derivative work, and (b) for obtaining their source code through illegal means. Remember, a halting problem is constructed around a halting checker.
Probabilistically this works better. One definition:
Atheists believe god does not exist (P(god exist) = 0)
Theists believe god does exist (P(god exists) = 1)
Agnostics believe something in between.
With these definitions, there are probably hardly any atheists in existence, though many theists. My personal level of agnosticism is somewhere around P(god exists) = 1e-80. Effectively an atheist, but not technically so, as I deem it a philosophical error to assign irrevocable truth (0 or 1) to any proposition. Compared with my belief about the sun not coming up tomorrow, the god proposition is a few orders less likely.
An alternative definition (used in combination with the above, breeding much confusion) is:
Atheists believe that it's likely God doesn't exist (P(god exist) < 0.5)
Agnostics find it as likely that god exists than that she doesn't (P(god exists) = 0.5)
Theists have faith (P(god exist) = 1.0)
And we have a gap between 0.5 and 1.0 exclusive.
In this set of definitions, there are many atheists, staunch theists, a few agnostics of little brain, and a diffuse group that thinks that there 'must be something out there'.
Next to all this you have a large group that we might call 'abstract theists', who would assert that the probability of God existing in a strict physical sense is close to nothing, but that the God concept has merit as a basis of culture and morality and for reasoning about morality. Here existence is not an issue, as there are no implications for physics, nor for biology. I dislike this position for various reasons, but at least it's rational.
Depending on which definition you use, there are either lots of agnostics or lots of atheists. Theists usually misuse the two definitions, something like: "Ah, so according to your philosophy you cannot logically conclude that God doesn't exist (with probability 1), so you must be agnostic, and therefore find it likely (50%) that god exists, so you have to take me seriously.". No, I assign probability 1e-80 and both from a philosophical point of view, as well as out of my belief in the proposition and the reasons I came to this belief, I think your belief in god is completely wrong, and your assertion that god has a physical existence is irrational and stupid.
Well, given the historical track record of stock analysts, you could just as well believe a chimpansee, a drunken journalist, or for that matter a Slashdot Yahoo. Any appeal to the authority of Wall street has no basis in reality whatsoever.
But of course there's a shortage of SKILLED developers. That's the nice thing about being skilled: you usually have a job already. What you're complaining about is a shortage of skilled developers that are looking for another job. That this is rare simply says that your competitors are treating skilled people more-or-less decently. They're there, but just not dying to find another job.
Your company simply did a stupid acquisition. The acquired company doesn't have the skillset, and your company didn't plan well enough in advance to first set up the skillset internally, before they started buying. And now you're in a fix. Simple case of bad planning, says nothing about the shortage of skills.
If bittorrent will be heavily used to distribute mainstream media, the ISP's can cull cost without any caching arrangement. Just whitelist a number of torrent seeders (Mainstream, 'legit' seeders) at the ISP side, and cache/seed whatever your customers are downloading from those places. As you can provide best speed to the customers, they will mostly use your seeder and the traffic stays at home. This would probably be more flexible and powerful than current methods.
Best way to view this is to consider what would happen when there is no mutation to speak of in a certain species. Would the species thrive? Well, yes, as long as its environment stays the same. However, if the environment changes (for instance by an evolving predator, or an evolving parasite), the non-mutating species will be going the way of the dodo very fast, as in the absence of mutation, it has no means of adapting to the changes.
So, it might not be mentioned simply because the resolution is so obvious.
7 M euros is 2 eurocent per EU citizen of propaganda. How much has the US spent on the presidential election so far?
Fingerprints. Present in passports from June 2009 to be able to meet the US visa requirements
President of the EU? Statutory position. Blair would do just as well as any other politician that's survived national politics. They've all got blood on their hands anyway. Blair, Prodi, Aznar, Schroeder, Chirac, Berlusconi? Let's take Blair.
Yes, there's no knowledge, only belief. However, belief can be quantified in probability and can then be compared. I believe that electrons are real and quantized, roughly at the 0.99999 level. The basis for this is reputation of the scientists involved, length of this theory being unchallenged, possibility of doing the experiments myself (and therefore the amount of people that could falsify this easily and gain instant fame), the amount of work that is based on this work (the computer I write this one for instance). All that leads to a belief that, for all practical purposes, I consider to be truth.
I also believe that global warming is real, quantified as quite high (as compared with many), at the 0.9 level. The physics behind greenhouse gasses I believe to be true with a much higher likelihood, but the complexity of the atmosphere and all canondrum around it suppresses very strong belief in it. I merely consider it very likely.
Now you're saying it's all belief, and you're throwing out the quantification aspect, interrelatedness of fact, chains of reasoning, accuracy of results, and generally put a big gray blanket of mistrust over understanding (scientific or not).
Re:Shorting AMD stock: NASDAQ figures
on
Is AMD Dead Yet?
·
· Score: 1
You are in principle right, the GP was wrong. What he should have said was:
Remember we are no longer in a CPU speed race; we've nearly reached physical limits, and CPU's are not going to be significantly faster. Perhaps ever. Luckily they're plenty fast enough.
The CPU race is over because marketing Ghz's has been a losing proposition for a couple of years now. Your mother's machine in 92 has what clockspeed? And what was the clockspeed for a similarly priced machine in '95? What was the clockspeed in 2005? Also 2 or 3 Ghz like it was in 2003, and like it is now?
The Ghz race is over, we now have to look at energy consumption, IO speed, we'll witness the demise of harddrives in favour of solid state, and finally we'll go into parallel computing in earnest. Still lots to improve upon, but sheer single-core speed is not going to be it. I'd expect to be able to run, maybe, a 4 Ghz chip in my handheld in 15 years, but I'm not counting on it. For my workstation in 15 years, I do hope to have a few hundred CPUs, at least 1 GB of static ram cache, 1 TB of DRAM, all backup up (including my filesystem) by a 100-1000 TB solid state storage system. Good times ahead.
As someone interested in philosophy i agree hes way off, you need to understand something before you can build it.
Before creating AI there should be some consensus as to "What is AI", "What is the mind", "What is consciousness", (and as you hint at) "What does it mean to be human" and lots of other fundamental questions. I disagree. We're going to build increasingly complex computers for a wide array of tasks. We are going to connect them, build meta-computers to control the interconnectedness in order to define and solve sub-tasks. We're going to keep on building interfaces, and with tasks becoming more complex, the interfaces will become more complex. So we will try to create programs that create the interfaces.
At some point (way later than 2029), we will look at the whole intertangled mess and observe: "gosh, it's intelligent". At that point we might be able to define AI, not before.
Slashdot Mirror
There's Fisher's fundamental law of natural selection, later generalized to Price's covariance and selection theorem. Although hideously misinterpreted at times, these are testable statements about what one expects to observe when natural selection is true.
It would then of course be equally legal to take your binaries, examine the machine- or bytecode, decompile it, rip out the good stuff (your new features) and re-assemble it back in firefox. Cumbersome, but perfectly doable.
After some reflection on this issue (after submitting the post), I do concede that, theoretically, the author of Psyco has a point. However, in theory there's no difference between theory and practice. In practice, there is.
So, the full responsibility of the Oil companies raid on the Nigerian population should not be ascribed to the Gates Foundation. They should however know better to support than to support such a raid by whatever small percentage they own. Buying shares is taking ownership is taking responsibility. It simply is a black spot on the foundation, the size is a matter of debate.
No, I guess VNC is still the better option.
Given that, the foundation will just bleed dry if they don't invest for maximum profits. And the more profitable their investments, the more impact the foundation can have.
Maximum profits doesn't follow. You can easily aim to invest for 'reasonable', or 'sustainable' profit to avoid the foundation to bleed dry. There's no need to maximize. In particular, when the investments go directly against the purpose of the foundation, you would have a net negative, in that the existence of the foundation would do more harm than good.
If that were the case, the world would be better off when Gates' fortune would be spent on gambling, hookers and blow.
I think chances are higher that soldiers will shoot civilians when the civilians start shooting at the soldiers. Right to bear arms to the rescue!
That's easily solvable. The US is big and divided enough to actually send in redneck armies into the east or west, to 'prevent those darned librels to sell out god's own country to the UN', or vice-versa get the Metropolitan troops to 'quell a fundamentalist revolution in Jesus land'. The army of the underlying party can be shipped overseas, or sent out to guard Alaska.
I am however suprised at the speed of Python+Psycho though. It must have come a long way as last time I looked it was roughly twice as slow. One word of caution though with this combination: you need to refrain from using much of python's dynamicity to be able to let psycho do its job. Effectively psycho figures out type information at runtime to eliminate the overhead of dynamic dispatch.
Overall the score is: Java 2.64, Python + psyco 9.22 (factor 3), Python 26.66 (factor 10), Ruby 76.56 (factor 25).
The end result is that Java with Hotspot runs in the same order as C/C++, maybe a factor of 1.5/2 slower (if you're careful). Pure Python and Ruby will always run one order (factor 20 for Python, 40 for Ruby) slower because a function call is an expensive operation.
The only way to get speed out of the dynamic languages is to mix them with C. Python is pretty good at that, and I consider PythonWithEmbeddedC a reasonable language, yet a bit clumsy.
Interesting that you manage to put in a push for Python in the same post that complains that Java is too slow. On its own, python is a complete pig, you need to move significant portions of the code to C to even remotely reach the speed of Java. Java is decently fast, and I'm pretty sure Google uses it for various things that (a) are not time-critical, and (b) need to run fast.
As long as there are no people involved that have seen sample A, sample B is clean.
Butting in a bit late, but nevertheless: why do you think it is important that a machine can simulate the universe in real-time? It's pretty much a given that this is impossible, as any computational device is part of that same universe, and would thus need to be able to simulate itself in real-time. Ad infinitum. What's your point with this?
Strictly speaking, if you construct a halting problem around a commercial static code analyzer to prove they can't solve all problems, you would likely be slapped with a lawsuit (a) for creating an unauthorized derivative work, and (b) for obtaining their source code through illegal means. Remember, a halting problem is constructed around a halting checker.
Atheists believe god does not exist (P(god exist) = 0)
Theists believe god does exist (P(god exists) = 1)
Agnostics believe something in between.
With these definitions, there are probably hardly any atheists in existence, though many theists. My personal level of agnosticism is somewhere around P(god exists) = 1e-80. Effectively an atheist, but not technically so, as I deem it a philosophical error to assign irrevocable truth (0 or 1) to any proposition. Compared with my belief about the sun not coming up tomorrow, the god proposition is a few orders less likely.
An alternative definition (used in combination with the above, breeding much confusion) is:
Atheists believe that it's likely God doesn't exist (P(god exist) < 0.5)
Agnostics find it as likely that god exists than that she doesn't (P(god exists) = 0.5)
Theists have faith (P(god exist) = 1.0)
And we have a gap between 0.5 and 1.0 exclusive.
In this set of definitions, there are many atheists, staunch theists, a few agnostics of little brain, and a diffuse group that thinks that there 'must be something out there'.
Next to all this you have a large group that we might call 'abstract theists', who would assert that the probability of God existing in a strict physical sense is close to nothing, but that the God concept has merit as a basis of culture and morality and for reasoning about morality. Here existence is not an issue, as there are no implications for physics, nor for biology. I dislike this position for various reasons, but at least it's rational.
Depending on which definition you use, there are either lots of agnostics or lots of atheists. Theists usually misuse the two definitions, something like: "Ah, so according to your philosophy you cannot logically conclude that God doesn't exist (with probability 1), so you must be agnostic, and therefore find it likely (50%) that god exists, so you have to take me seriously.". No, I assign probability 1e-80 and both from a philosophical point of view, as well as out of my belief in the proposition and the reasons I came to this belief, I think your belief in god is completely wrong, and your assertion that god has a physical existence is irrational and stupid.
Well, given the historical track record of stock analysts, you could just as well believe a chimpansee, a drunken journalist, or for that matter a Slashdot Yahoo. Any appeal to the authority of Wall street has no basis in reality whatsoever.
So, quickly, how many bytes to a 1.44 MB floppy?
Want to get to the root account? try 'sudo su'.
Your company simply did a stupid acquisition. The acquired company doesn't have the skillset, and your company didn't plan well enough in advance to first set up the skillset internally, before they started buying. And now you're in a fix. Simple case of bad planning, says nothing about the shortage of skills.
If bittorrent will be heavily used to distribute mainstream media, the ISP's can cull cost without any caching arrangement. Just whitelist a number of torrent seeders (Mainstream, 'legit' seeders) at the ISP side, and cache/seed whatever your customers are downloading from those places. As you can provide best speed to the customers, they will mostly use your seeder and the traffic stays at home. This would probably be more flexible and powerful than current methods.
So, it might not be mentioned simply because the resolution is so obvious.
7 M euros is 2 eurocent per EU citizen of propaganda. How much has the US spent on the presidential election so far?
Fingerprints. Present in passports from June 2009 to be able to meet the US visa requirements
President of the EU? Statutory position. Blair would do just as well as any other politician that's survived national politics. They've all got blood on their hands anyway. Blair, Prodi, Aznar, Schroeder, Chirac, Berlusconi? Let's take Blair.
I also believe that global warming is real, quantified as quite high (as compared with many), at the 0.9 level. The physics behind greenhouse gasses I believe to be true with a much higher likelihood, but the complexity of the atmosphere and all canondrum around it suppresses very strong belief in it. I merely consider it very likely.
Now you're saying it's all belief, and you're throwing out the quantification aspect, interrelatedness of fact, chains of reasoning, accuracy of results, and generally put a big gray blanket of mistrust over understanding (scientific or not).
Remember we are no longer in a CPU speed race; we've nearly reached physical limits, and CPU's are not going to be significantly faster. Perhaps ever. Luckily they're plenty fast enough.
The CPU race is over because marketing Ghz's has been a losing proposition for a couple of years now. Your mother's machine in 92 has what clockspeed? And what was the clockspeed for a similarly priced machine in '95? What was the clockspeed in 2005? Also 2 or 3 Ghz like it was in 2003, and like it is now?
The Ghz race is over, we now have to look at energy consumption, IO speed, we'll witness the demise of harddrives in favour of solid state, and finally we'll go into parallel computing in earnest. Still lots to improve upon, but sheer single-core speed is not going to be it. I'd expect to be able to run, maybe, a 4 Ghz chip in my handheld in 15 years, but I'm not counting on it. For my workstation in 15 years, I do hope to have a few hundred CPUs, at least 1 GB of static ram cache, 1 TB of DRAM, all backup up (including my filesystem) by a 100-1000 TB solid state storage system. Good times ahead.
Before creating AI there should be some consensus as to "What is AI", "What is the mind", "What is consciousness", (and as you hint at) "What does it mean to be human" and lots of other fundamental questions.
I disagree. We're going to build increasingly complex computers for a wide array of tasks. We are going to connect them, build meta-computers to control the interconnectedness in order to define and solve sub-tasks. We're going to keep on building interfaces, and with tasks becoming more complex, the interfaces will become more complex. So we will try to create programs that create the interfaces.
At some point (way later than 2029), we will look at the whole intertangled mess and observe: "gosh, it's intelligent". At that point we might be able to define AI, not before.