Actually, it is even worse. You can patent an idea for a "solution" without actually knowing how to implement it. For instance (to quote a Friends episode), you can attempt to patent "selling icecream over the Internet". The problem, the fact that icecream melts, is not solved. However, as soon as someone solves that problem and starts a viable business of selling icecream over the Internet, the patent holder can jump in with his claims. So the patent holder did nothing more than identify a problem and lay claims to any solution to that problem.
Your real goal though is to write something, get it patented and then sell it for millions to the big boys.
Except what will happen is something like:
Microsoft: "You have a patent we need."
Lone Coder: "OK. It'll cost you a million bucks."
Microsoft: "Forget it, we're going to use your stuff without paying."
Lone Coder: "Then I'll sue you."
Microsoft: "But our defense will be that our code is a tiny but significant bit different from yours. So we're not infringing."
Lone Coder: "I'll get a really good lawyer who'll show that you actually ARE infringing."
Microsoft: "That will take years. Lawyers are expensive. In the meantime, we believe that you are infringing on several of our patents. So we're going to unleash our army of lawyers onto you. Can you really afford the legal costs?"
Lone Coder: "You bastards."
Microsoft: "Now now. There's no need for ugly language. We're your buddies. We're here to help you! If you just sign over the rights to your patent to us, we will allow you to continue using it yourself. Of course, when our new blockbuster software comes out, your market is destroyed, but until that time you will be able to make a living."
Lone Coder: "Arghhhhhh!"
Microsoft: "And you pay us a million dollars, just so that we won't have our lawyers start sueing you for infringing our patents, as soon as we have thought up a couple that you probably are infringing."
One player as Mr. X trying to run from I believe 5 Scotland Yard agents.
Against 5 detectives, Mr.X has no chance at all.
The game is best played with Mr.X against 4 detectives. The detectives will almost always win the game, but occassionaly Mr.X will make a brilliant escape.
Against 3 detectives, Mr.X easily wins the game.
Actually, that is the bad thing about Scotland Yard: the game is fun but never fair; either Mr.X has a huge chance of winning, or the detectives have, depending on the number of players.
One cannot help wondering whether with a different map design, the game could be made more fair. Perhaps with different maps for 4-player, 5-player, and 6-player games.
They have armies of lawyers sitting around the offices. Their job is making life hell for the competition -- any competition. Sending out a C&D letter is riskfree. Best possible result: competition decides to change the name, which costs them time and money. Worst possible result: competition laughs it off, and you can decide whether or not to take further steps later.
It wasn't goatse, but I felt a bit ill after reading about ten lines.
Is this really what is called "politics" in the US? Have W and his consorts any real opinion? It's an official campaign website and all they can do is slander and insinuate!
In your example you want to create an addition function. You provide a test function that tests whether 3 is the result when 1 and 2 are put into the function.
Now, I can create many different functions that give this result. One is a function that always returns 3, regardless your input. Another is that takes the first number, and adds 2 to it, regardless the second number. Yet another is a function that always returns 3 when 1 and 2 are the parameters, and zero otherwise.
So, to evolve a function that can add 1 and 2, your test function is sufficient. However, you already knew the answer to the addition of 1 and 2 (you needed the answer to write your test function), so to evolve a function for this is useless.
On the other hand, if your goal is to evolve a function that can do ANY addition, you need a far more complex test function. And it doesn't matter how many examples you put in your test function, there'll always be an infinite number of possible functions that can generate these test answers. The only viable test function is a function that completely imitates the required function: and if you can imitate it, you already programmed it.
If you wonder whether GP is of any use at all, the answer is yes, but only for problems where you find it very hard to come up with the right function to optimise certain results, and you are not interested in very specific results: any approximation should be enough.
For most software, approximations are not sufficient. I cannot have a Word Processor that only checks the spelling of 90% of the words, it must check them all. Or a Word Processor that saves my files correctly 99% of the time but corrupts them the remaining 1% of the time.
I use GP myself, but mainly for things like control of robots in unknown environments. The only real requirement is that the robot does not destroy itself, for the rest I find that GP delivers suitable solutions for specific environments wherein it is located - maybe not the best, but far better than I would be able to program when I would have to take into account all possible environments.
If a "robot" is defined as a machine that does a task in the same way as a human, an AIBO is not a robot, since it replaces a dog. Furthermore, neither a lawnmower nor a floorsweeper are robots, they would only be robots if they took the old lawnmower/broom out of the garage/closet and started using that to mow/sweep the lawn/floor.
But if you have to specify test code, you are either incomplete, so when used as a fitness function the GP Word Processor would be incomplete, or you would be complete (very difficult, but possible in theory), but then your test code would basically BE the Word Processor you want to evolve, and GP would be useless. So I'm afraid even in theory this would not work.
I am very interested to see how Jackson manages to connect up with the previous EEs. There are scenes added to the previous two EEs, which are not in the theatrical releases, which will have an impact on scenes already shown in the cinemas. I am thinking, for instance, of the scene where Merry and Pippin drink the Ent's stuff that makes them grow. At the crowning, the four hobbits are standing next to each other. In the theatrical release, the hobbits are of equal length. In the EE, Merry and Pippin need to be at least a head taller than Frodo and Sam! Will new special effects do the trick here?
Goff thinks AI will be able to take over the job of a programmer. Hah! What a ridiculous notion. Evidently he fell for some vague claims of specialists in genetic programming, of which he understands nothing himself. Well, evolutionary algorithms (of which genetic programming is a subclass) are one of my specialities too. And I can tell you that, while evolutionary algorithms are fun to toy with, and can be used for some interesting real-life applications, they will never be able to design complete programs for us. And why is that? Because their results are notoriously untrustworthy. You will only use them for problems for which you can't design a solution by hand, and for which you are satisfied with an approximate solution. Most applications are simply not like that. Furthermore, evolutionary algorithms can only be applied to problems for which a fitness function can be designed. I wonder what Goff thinks a fitness function for a Word Processor looks like.
Obviously, Goff knows nothing about the practice of computing science. But he still manages to hawk out three crappy articles about it. Which get on the/. frontpage... Man, that guy is good.
Since the two Southampton programs work together, I would say they both earn half of their combined winnings. From that point of view, Southampton lost big. Of course, if you are allowed to completely discount your losing entries, it is easy to defeat Southampton: simply create one master with TWO slaves.
Once I "won" a PD-tournament by fiddling with the organising engine. The friendly engine supplied the programs with the recent histories, and I simply inserted all cooperates in my opponent's table. The scores were only counted at the end. This was in a trial session, and I only wrote the program to expose a problem in the engine. I wouldn't dream of entering the program for real, because it would defeat the purpose of the tournament. I can't say I have much respect for the Southampton team that didn't have any qualms about cheating.
Axelrod never claimed that Tit-for-Tat was the best PD-playing program. He just stated that Tit-for-Tat would play well against any other combination of programs. Actually, IIRC, in the second tournament he organised Tit-for-Tat came in second. There was a different program that managed to exploit faults in other programs.
It is easy to score better than Tit-for-Tat in Axelrod's (original) tournament. He included a program that played random moves. It is not difficult to recognise this program after, say, ten moves have been played. You can always defect against random, because its moves are unrelated to its history. So, a program that plays Tit-for-Tat by default, but always defects against Random, scores better than Tit-for-Tat.
Does this dillute Tit-for-Tat's accomplishment? Of course not. Tit-for-Tat still plays well. And it is such a simple strategy that it can be programmed in two lines ("C on move 1, then copy opponent's previous move"), which none of the other programs achieve. Tit-for-Tat is simple, elegant, and strong. It's beautiful.
Southamptom entries, on the other hand, are complex, sneaky, and cheating against (perhaps unwritten, but nonetheless agreed-upon) rules. They're ugly. They only prove that backstabbing cheating bastards may defeat just-and-fair if the referee is looking the other way for a moment.
You'll get no argument from me, I completely agree.
I was just reacting to the parent's statement that humans are better chess players than chess computers. They are not.
Furthermore, I would say that not only the game of chess remains interesting when humans play each other, but also when humans play against a computer. The reason is that chess games against a computer feel completely different from chess games against humans. And computers sometimes come up with very surprising moves.
Of course, I may have a coloured view because chess programs are part of my studies.
If I'm allowed a person opinion, I WANT the article to be true. I WANT something completely evil and "not-normal" to blame the fall of the Original Horizons on.
But what the story describes is NOT "not-normal". It is all too common in the "business world". To a geek's mind, it is incredible that a manager would destroy a product, and take down a creative mind, just to step up in the ranks. However, you must remember that people that aspire management jobs instead of creative jobs are basically only in it for personal gain, "gain" consisting mainly of money and power. I have seen enough companies inside and out to realise that those at the top did not get there on their personal merits, but only on their ability to be sleazebags. The only exception are CEOs who formed their own company and made it grow.
If you see that, you will realise that Allen's story is entirely believable, and provides many opportunities for learning. Methinks his biggest mistake was trying to run the business-side of a company, without having the education for doing that. With the right education, he would have known the problems he could run into, and the way he could be exploited. He didn't know, which made him all the more vulnerable for exploitation. Such a man attracts vultures, and he got his share.
While it is a bygone conclusion that computers will someday surpass humans at chess by a large margin, this has yet to happen.
Hm. Computers play a draw against the top-three of all human chess players. They beat everyone else. So, basically, computers play better chess than (5,999,999,997 / 6,000,000,000) * 100 % = 99.99999998 % of all humans, and lose of no-one. To me, that is no different than saying that computers are better at playing chess than humans.
Also, I know lots of people who would like to play go.
I know a lot about Go playing programs. To make a long story short: they all suck. AI research is simply not up to the task of getting a computer to play Go. It will take a long time before we get a challenging Go program.
However, you can combine IT with anything. I was at a company once where an IT Manager got in early every morning to clean the toilets. The cleaning staff was hired because it was cheap, not because of the swell job they did...
Why is it news? I assume it is considered news because Catan has been one of the most popular family-oriented boardgames for many years now. The rights to put it online have probably been coveted by many publishers. The fact that MSN got it is reason enough to worry (do you need a "passport" to join? will they sue heavily the open source versions?).
They all have an unwritten agreement that they won't use their patents against each other
Not unwritten. It is called cross-licensing. Basically, they allow each other unlimited, royalty-free access to all their patents.
It backfires on them occasionally when a small startup patents an idea and it gets under the radar but usually all they have to do then is buy out the startup and they are back to square one.
No, it backfires when a litigation company, consisting only of lawyers, acquires a patent they violate. They cannot offer cross-licensing, because the litigation company does not need any patents since it doesn't produce anything. It is only interested in money. And it doesn't want just a tip, usually it goes for something like 10% of the profits.
Sorry friend but your University owns the patent rights. Read what you signed when you started your Ph.D. and then cry.
Actually, no. It is stated specifically that I myself own all the rights to my own ideas. Even if the university pays me to generate them. But I would have no problems with the university patenting them. If they would have the money. Which they don't.
Actually, I know what you mean. Recently my sister came to me, telling me there was a problem with spyware on her computer, and would I be willing to install a new OS on her computer? So I said I would do that, if she would hand me a CD-ROM with the OS, and a registration number. She asked me why I couldn't pluck those from the net somewhere. Yes, I could, but I'm not going to. I paid for my OS, and I am not going to become an accomplice to my sisters kleptomanic urges. Believe it or not, in my family I was the one who had to defend my position. My sister had a problem, so why wouldn't I help her? I could not expect her to pay up for a legel version of Windows, could I? Didn't I know how expensive that is?
Still, at work most of my co-workers only use legal software. They are computer scientists, and have a high sense of ethics. On the other hand, at my wife's work, an administrative department, you can easily buy the latest warez-DVD's if you just tap a random person on the shoulder.
There is something to say for stealing Windows. Basically, you don't need to steal an initial OS for your computer, because virtually every new computer is delivered with a legal one already installed. However, the problem comes when the computer starts to mature. I would say that as long as MS ensures that older versions of Windows remain safe to use and are compatible with newer versions, there is absolutely no reason to steal the latest version. However, MS does not. I would say that that makes them kind-of liable for things that go wrong on your computer, and if you steal a new version which makes the thing work right again, everybody wins. I mean, it's not legal, but it is a gray area (incidentally, I bought every version of Windows I have ever installed).
However, there is NO excuse to steal any other piece of software. You steal photoshop? Why? Because you need to edit graphics? There are loads of free alternatives. Either use those, or pay up for Photoshop if you really like it so much. I assume you don't do that because all your friends steal software and you would feel silly if you would not (granted, the chance your actions will bite you in the bum are negligible). But it IS stealing. As long as you realise that.
Slashdot Mirror
The US have spy satellites up. Europe wants spy satellites up. Don't complain if the Chinese want theirs up there too.
Actually, it is even worse. You can patent an idea for a "solution" without actually knowing how to implement it. For instance (to quote a Friends episode), you can attempt to patent "selling icecream over the Internet". The problem, the fact that icecream melts, is not solved. However, as soon as someone solves that problem and starts a viable business of selling icecream over the Internet, the patent holder can jump in with his claims. So the patent holder did nothing more than identify a problem and lay claims to any solution to that problem.
Except what will happen is something like:
Microsoft: "You have a patent we need."
Lone Coder: "OK. It'll cost you a million bucks."
Microsoft: "Forget it, we're going to use your stuff without paying."
Lone Coder: "Then I'll sue you."
Microsoft: "But our defense will be that our code is a tiny but significant bit different from yours. So we're not infringing."
Lone Coder: "I'll get a really good lawyer who'll show that you actually ARE infringing."
Microsoft: "That will take years. Lawyers are expensive. In the meantime, we believe that you are infringing on several of our patents. So we're going to unleash our army of lawyers onto you. Can you really afford the legal costs?"
Lone Coder: "You bastards."
Microsoft: "Now now. There's no need for ugly language. We're your buddies. We're here to help you! If you just sign over the rights to your patent to us, we will allow you to continue using it yourself. Of course, when our new blockbuster software comes out, your market is destroyed, but until that time you will be able to make a living."
Lone Coder: "Arghhhhhh!"
Microsoft: "And you pay us a million dollars, just so that we won't have our lawyers start sueing you for infringing our patents, as soon as we have thought up a couple that you probably are infringing."
Lone Coder: *whimper*
Against 5 detectives, Mr.X has no chance at all.
The game is best played with Mr.X against 4 detectives. The detectives will almost always win the game, but occassionaly Mr.X will make a brilliant escape.
Against 3 detectives, Mr.X easily wins the game.
Actually, that is the bad thing about Scotland Yard: the game is fun but never fair; either Mr.X has a huge chance of winning, or the detectives have, depending on the number of players.
One cannot help wondering whether with a different map design, the game could be made more fair. Perhaps with different maps for 4-player, 5-player, and 6-player games.
They have armies of lawyers sitting around the offices. Their job is making life hell for the competition -- any competition. Sending out a C&D letter is riskfree. Best possible result: competition decides to change the name, which costs them time and money. Worst possible result: competition laughs it off, and you can decide whether or not to take further steps later.
Usually followed by:
If it is in computers, pick one.
If it is from Microsoft, pick none.
It wasn't goatse, but I felt a bit ill after reading about ten lines.
Is this really what is called "politics" in the US? Have W and his consorts any real opinion? It's an official campaign website and all they can do is slander and insinuate!
Geez!
In your example you want to create an addition function. You provide a test function that tests whether 3 is the result when 1 and 2 are put into the function.
Now, I can create many different functions that give this result. One is a function that always returns 3, regardless your input. Another is that takes the first number, and adds 2 to it, regardless the second number. Yet another is a function that always returns 3 when 1 and 2 are the parameters, and zero otherwise.
So, to evolve a function that can add 1 and 2, your test function is sufficient. However, you already knew the answer to the addition of 1 and 2 (you needed the answer to write your test function), so to evolve a function for this is useless.
On the other hand, if your goal is to evolve a function that can do ANY addition, you need a far more complex test function. And it doesn't matter how many examples you put in your test function, there'll always be an infinite number of possible functions that can generate these test answers. The only viable test function is a function that completely imitates the required function: and if you can imitate it, you already programmed it.
If you wonder whether GP is of any use at all, the answer is yes, but only for problems where you find it very hard to come up with the right function to optimise certain results, and you are not interested in very specific results: any approximation should be enough.
For most software, approximations are not sufficient. I cannot have a Word Processor that only checks the spelling of 90% of the words, it must check them all. Or a Word Processor that saves my files correctly 99% of the time but corrupts them the remaining 1% of the time.
I use GP myself, but mainly for things like control of robots in unknown environments. The only real requirement is that the robot does not destroy itself, for the rest I find that GP delivers suitable solutions for specific environments wherein it is located - maybe not the best, but far better than I would be able to program when I would have to take into account all possible environments.
If a "robot" is defined as a machine that does a task in the same way as a human, an AIBO is not a robot, since it replaces a dog. Furthermore, neither a lawnmower nor a floorsweeper are robots, they would only be robots if they took the old lawnmower/broom out of the garage/closet and started using that to mow/sweep the lawn/floor.
But if you have to specify test code, you are either incomplete, so when used as a fitness function the GP Word Processor would be incomplete, or you would be complete (very difficult, but possible in theory), but then your test code would basically BE the Word Processor you want to evolve, and GP would be useless. So I'm afraid even in theory this would not work.
I am very interested to see how Jackson manages to connect up with the previous EEs. There are scenes added to the previous two EEs, which are not in the theatrical releases, which will have an impact on scenes already shown in the cinemas. I am thinking, for instance, of the scene where Merry and Pippin drink the Ent's stuff that makes them grow. At the crowning, the four hobbits are standing next to each other. In the theatrical release, the hobbits are of equal length. In the EE, Merry and Pippin need to be at least a head taller than Frodo and Sam! Will new special effects do the trick here?
Obviously, Goff knows nothing about the practice of computing science. But he still manages to hawk out three crappy articles about it. Which get on the /. frontpage... Man, that guy is good.
Once I "won" a PD-tournament by fiddling with the organising engine. The friendly engine supplied the programs with the recent histories, and I simply inserted all cooperates in my opponent's table. The scores were only counted at the end. This was in a trial session, and I only wrote the program to expose a problem in the engine. I wouldn't dream of entering the program for real, because it would defeat the purpose of the tournament. I can't say I have much respect for the Southampton team that didn't have any qualms about cheating.
It is easy to score better than Tit-for-Tat in Axelrod's (original) tournament. He included a program that played random moves. It is not difficult to recognise this program after, say, ten moves have been played. You can always defect against random, because its moves are unrelated to its history. So, a program that plays Tit-for-Tat by default, but always defects against Random, scores better than Tit-for-Tat.
Does this dillute Tit-for-Tat's accomplishment? Of course not. Tit-for-Tat still plays well. And it is such a simple strategy that it can be programmed in two lines ("C on move 1, then copy opponent's previous move"), which none of the other programs achieve. Tit-for-Tat is simple, elegant, and strong. It's beautiful.
Southamptom entries, on the other hand, are complex, sneaky, and cheating against (perhaps unwritten, but nonetheless agreed-upon) rules. They're ugly. They only prove that backstabbing cheating bastards may defeat just-and-fair if the referee is looking the other way for a moment.
I was just reacting to the parent's statement that humans are better chess players than chess computers. They are not.
Furthermore, I would say that not only the game of chess remains interesting when humans play each other, but also when humans play against a computer. The reason is that chess games against a computer feel completely different from chess games against humans. And computers sometimes come up with very surprising moves.
Of course, I may have a coloured view because chess programs are part of my studies.
But what the story describes is NOT "not-normal". It is all too common in the "business world". To a geek's mind, it is incredible that a manager would destroy a product, and take down a creative mind, just to step up in the ranks. However, you must remember that people that aspire management jobs instead of creative jobs are basically only in it for personal gain, "gain" consisting mainly of money and power. I have seen enough companies inside and out to realise that those at the top did not get there on their personal merits, but only on their ability to be sleazebags. The only exception are CEOs who formed their own company and made it grow.
If you see that, you will realise that Allen's story is entirely believable, and provides many opportunities for learning. Methinks his biggest mistake was trying to run the business-side of a company, without having the education for doing that. With the right education, he would have known the problems he could run into, and the way he could be exploited. He didn't know, which made him all the more vulnerable for exploitation. Such a man attracts vultures, and he got his share.
Hm. Computers play a draw against the top-three of all human chess players. They beat everyone else. So, basically, computers play better chess than (5,999,999,997 / 6,000,000,000) * 100 % = 99.99999998 % of all humans, and lose of no-one. To me, that is no different than saying that computers are better at playing chess than humans.
I know a lot about Go playing programs. To make a long story short: they all suck. AI research is simply not up to the task of getting a computer to play Go. It will take a long time before we get a challenging Go program.
However, you can combine IT with anything. I was at a company once where an IT Manager got in early every morning to clean the toilets. The cleaning staff was hired because it was cheap, not because of the swell job they did...
Why is it news? I assume it is considered news because Catan has been one of the most popular family-oriented boardgames for many years now. The rights to put it online have probably been coveted by many publishers. The fact that MSN got it is reason enough to worry (do you need a "passport" to join? will they sue heavily the open source versions?).
Not unwritten. It is called cross-licensing. Basically, they allow each other unlimited, royalty-free access to all their patents.
It backfires on them occasionally when a small startup patents an idea and it gets under the radar but usually all they have to do then is buy out the startup and they are back to square one.
No, it backfires when a litigation company, consisting only of lawyers, acquires a patent they violate. They cannot offer cross-licensing, because the litigation company does not need any patents since it doesn't produce anything. It is only interested in money. And it doesn't want just a tip, usually it goes for something like 10% of the profits.
The rest of your text is accurate.
Actually, no. It is stated specifically that I myself own all the rights to my own ideas. Even if the university pays me to generate them. But I would have no problems with the university patenting them. If they would have the money. Which they don't.
Let's think, what would I put in it...
Still, at work most of my co-workers only use legal software. They are computer scientists, and have a high sense of ethics. On the other hand, at my wife's work, an administrative department, you can easily buy the latest warez-DVD's if you just tap a random person on the shoulder.
However, there is NO excuse to steal any other piece of software. You steal photoshop? Why? Because you need to edit graphics? There are loads of free alternatives. Either use those, or pay up for Photoshop if you really like it so much. I assume you don't do that because all your friends steal software and you would feel silly if you would not (granted, the chance your actions will bite you in the bum are negligible). But it IS stealing. As long as you realise that.