There are many cases where the symbolic formula is much simpler than the equivalent program. Take any matrix factorization for example (QR, LU, SVD, etc), the programs to obtain them are rather complex (take a look at a good implementation of the divide and conquer algorithm for eigenvalue decomposition), while the equivalent formula is nothing more than 3 or 4 characters.
I'm not really sure I want to reduce all my trust to the emotional mindset of the programer. Emotions are often proven to be very instable, and in case of critical systems the consequences will be desastrous. I'd prefer to use a well defined and robust methodology, ideally involving proven programs.
Seriously? You are judging the work of someone solely on the shape of his citation curve on scholar? Compressed sensing is all but new, right, but selling these guys as bad because they have only 2k citations (!!!) on scholar is a bit exagerated, to say the least.
I've always thought of Verisign&co as a very reliable evidence that I'm paying something over the real paypal and not to some Russian based thugs. I would never trust them for my privacy.
Wasn't it the same situation 2 years ago in Libya with the gov certificate being trusted by default by some OS?
Why do so many developers waste time on obfuscation and other ways of hiding the source in scripting languages?
Using utilities like IonCube to 'protect' PHP-code will never stop the dedicated people from reverse engineering the application or re-engineering it. I've seen that countless times. It is security-through-obscurity at best and it will prevent people from both fixing bugs and re-submitting the fixed code to the developers, and finding security issues from simple code reviewing.
If developers of competing applications needs to steal code they're really crappy developers and whatever that makes their application unique will be equally crappy and thus not a threat.
Which brings us to the next point: If obfuscation is worthless and someone will steal you code whatever you do, just release it with an open source license in the first place.
My guess is that the short amount of time between the release and the cracking is where the management expects to make profit, and even more profit than if it was FLOSS in the first place. This highlights greatly the short-term objectives of today's business.
Isn't this like unreported employment, where workers have no rights and the state gets nothing (for maintaining the infrastructures used). I know/. is US-centric and my little European country seems communist to most of you (I'm from France). But seriously unreported employment is a bad idea, although it might look better than unenployment. Firstly, it's a downhill to slavery, like the world was before the introduction of labour laws. And secondly, it's not sharing at all because there is no collectivity in such shemes. It's everyone is on its own without any place for a collective structure, which is obviously not the way humankind has eveloved for the last couple of thousands of years.
These deregulated systems are utopias that only work if people are equally smart and potent, which will definitely never be the case.
Not really in France. "Le Mans" is pronounced without the final 's' (this is a french classic), with the sound [], like in the end of croissant (again, no final 't'), which doesn't exist in many languages.
Still it is difficult not to feel a little shame from the fact we all belong to the same species. Of course you are by no way responsible for the actions of others, but you can still fell a little sad and ashamed by the incommensurable stupidity of our fellows.
I am exactly in the same case. It has been a very long time since I wiped out any windows off my computers at home (never had one at work). I am so used to linux it would take me ages to regain my productivity if I were to switch to another OS.
Moreover, as far as I know, Windows is still no free software. I cannot count the number of things I learned thanks to the openness of linux by just browsing the code (things on filesystems, memory management, processes and scheduling, etc). Plus colsed things do not correspond to my ethics, which is a personal choice.
What is the point in having a TV or not? I do not get it.
As far as I can tell, "AI" has succeeded only in keeping the same name after endless redefinitions resulting from it's numerous failures.
Your blind faith in AI seems to indicate that you're either hopelessly misguided or one of those singularity nuts.
Absolutely not. The improvement in fields that were said to be impossible for AI are just astonishing, and I am among the first surprised by such successes. Let's state it clear: computing power is increasing, theoretical models are improving, practical implementations are getting more efficient. So yeah, basically Turing was right, we are just impressively capable computers and nothing more.
Around 5 or 6 years ago, there were some image classification benchmarks that were incredibly tough and said to be almost impossible to solve with a machine, with very low accuracies. Were are we now? Well the improvements have been far better than expected. Far better than I expected, to be honest. In some sense I would have loved if it didn't, since the pressure of this ever growing progress is stressful to my students and complicates the publication of novel ideas. But basically, yeah, it is improving a lot. It is science, it just works.
You could argue computer vision is not AI, but it is. Everything that allows a computer to make a statement that you thought was only possible by a human is AI.
Again, I am not interested in the the colorful stories about consciousness or whatever. What I'm saying is that there is basically no task that a computer will not be able to perform in the long run. Get over it, we are all replaceable by machines. Engineers, researchers, artists, name what you want, it is only a question of time and not of possibility.
Well, the truck can deliver the goods to a local market. Then, you can go to that market using your feet or even a bike. I guess it is even more green. It is the way our grandparents did. Why do we different? Because we have plenty of cheap energy and it is more comfortable the other way.
It might change when the energy will not be that cheap, though. I am pretty pessimistic at the idea some environmental enlightenment will win against laziness...
These exact same fears were written about in 1980. There was a famous BBC TV programme about how robots and microprocessors would replace everyone.
We already know the outcome.
I have the greatest difficulties with the assumption our society is a stationary process. Especially when most of the things around us are expected to grow exponentially. These 2 are mutualy exclusive, that is, you should take one but not both. Either you take the stationary society, and the "it was so before, thus it will be so in the future" is a very valid argument. Or you take the exponential growth everyone is looking for, and things will surely not be in the future as they were in the past.
it takes a powerfully broken worldview to even begin to think that people only do create stuff so that they'll get paid.
With that kind of thinking, I'm surprised you aren't advocating the abolition of payment for all jobs. Doctors, teachers, taxi drivers - they should all work for free according to this argument, right?
You didn't understand his point. He says that whether you pay artists or not, they will continue to create new things, because the primary reason they do it is that they like it much more than everything else. Which may obviously not be the case of you taxi driver, who takes his job as a necessity to survive instead of a pleasant activity.
So yeah, you can basically cut some art revenues with little (if any) effect on art creation.
Clearly, the exponential model has the best fit (which is not very surprising), and says 2015, take or give 1 year for 95% confidence. Of course, there is no theoretical model behind, but most of the time, the theoretical explanation comes after the empirical fit.
Same for speed. Unless you have a brain dead "repeat 1000 times" benchmark, Java is as fast as any other language.
After all: it gets compiled down to the same machine code...
Not exactly. Java is unable to vectorize floating point operations now (might change with java8 though), which is very common in any multimedia application. So it remains a lot slower than what you get by default in C/C++ with a decent compiler.
There are a ton of off-the-shelf machine learning toolkits that are sufficient for 90% of possible use cases. The problem is getting annotated data to feed into these tools so they can learn the appropriate patterns. But all that requires is a host of annotators (i.e. undergrads and interns), not machine learning experts.
Exactly this!
Almost everything you ever dreamed of as a non machine learning expert is available at https://mloss.org/software/ Please now annotate more data so that we can tune the algorithms;-)
Is sustained focus even possible in mass audiences anymore? If not, what have we lost?
As an associate professor at the university, I can tell you many students have lost sustained focus, even in very small groups. If an explanation takes longer than 5 minutes, you lose them. If a problem takes longer than 5 minutes to solve, you lose them too. Starting 2 years ago, I modified all my lectures to have like "breakpoints" very often, so that no-one gets lost.
However, I think we already lost the Cartesian approach to breaking problems into smaller tasks. If you give them a rather simple but big problem, very few students are able the break it down and solve each part. Most will just try a global solution for a few minutes, then try the internet for a global solution, and finally get bored and say it's too complicated. One of my hypotheses is that the internet permits to solve most of the problems instantaneously, so you don't need sustained attention anymore. For the few cases where it is needed, well, that's the difference between the elite and the others...
Well, many trades will never be outsourced, or replaced by a robot.. namely landscaping, plumbing and electrical work. Much of which isn't safe for robots, and likely won't be for a very long time.
You might be surprised, but a lot of people in France think to go back to 40 hours a week, because they want higher wages whatever the unemployment rate. They don't get that going from 35 to 40 roughly fires 1 in 8 workers (of course it's not exactly true, because people get better wages, and this tends to create jobs - or the company doesn't raise the wages and instead invests in more machines to fire more workers).
I'm pretty sure it would be better for every one to lower this to 32 or 30 hours per week, and it won't have a significant impact on productivity. But the impact on the quality of living would be great.
I would also have chosen the BY-NC-ND package even if I don't care about the NC aspect just because it is the only one to have the ND claim. This one is fundamental for a research paper.
If you take into account the time spent to write a good paper, every single word has been carefully crafted for hours. The idea to allow paraphrase or remixes is at best non-sense, most of the time it's just a very bad idea.
I'm pretty sure the authors in the study choose ND, and what ever the remaining condition, because as a researcher, there is just no way I could allow you to make me say something I was not meaning to say in the first place.
Slashdot Mirror
There are many cases where the symbolic formula is much simpler than the equivalent program. Take any matrix factorization for example (QR, LU, SVD, etc), the programs to obtain them are rather complex (take a look at a good implementation of the divide and conquer algorithm for eigenvalue decomposition), while the equivalent formula is nothing more than 3 or 4 characters.
I'm not really sure I want to reduce all my trust to the emotional mindset of the programer. Emotions are often proven to be very instable, and in case of critical systems the consequences will be desastrous. I'd prefer to use a well defined and robust methodology, ideally involving proven programs.
Seriously? You are judging the work of someone solely on the shape of his citation curve on scholar? Compressed sensing is all but new, right, but selling these guys as bad because they have only 2k citations (!!!) on scholar is a bit exagerated, to say the least.
I've always thought of Verisign&co as a very reliable evidence that I'm paying something over the real paypal and not to some Russian based thugs. I would never trust them for my privacy.
Wasn't it the same situation 2 years ago in Libya with the gov certificate being trusted by default by some OS?
Why do so many developers waste time on obfuscation and other ways of hiding the source in scripting languages?
Using utilities like IonCube to 'protect' PHP-code will never stop the dedicated people from reverse engineering the application or re-engineering it. I've seen that countless times. It is security-through-obscurity at best and it will prevent people from both fixing bugs and re-submitting the fixed code to the developers, and finding security issues from simple code reviewing.
If developers of competing applications needs to steal code they're really crappy developers and whatever that makes their application unique will be equally crappy and thus not a threat.
Which brings us to the next point: If obfuscation is worthless and someone will steal you code whatever you do, just release it with an open source license in the first place.
My guess is that the short amount of time between the release and the cracking is where the management expects to make profit, and even more profit than if it was FLOSS in the first place. This highlights greatly the short-term objectives of today's business.
Isn't this like unreported employment, where workers have no rights and the state gets nothing (for maintaining the infrastructures used). I know /. is US-centric and my little European country seems communist to most of you (I'm from France). But seriously unreported employment is a bad idea, although it might look better than unenployment. Firstly, it's a downhill to slavery, like the world was before the introduction of labour laws. And secondly, it's not sharing at all because there is no collectivity in such shemes. It's everyone is on its own without any place for a collective structure, which is obviously not the way humankind has eveloved for the last couple of thousands of years.
These deregulated systems are utopias that only work if people are equally smart and potent, which will definitely never be the case.
le-mans (which is prounounced le mons).
Not really in France. "Le Mans" is pronounced without the final 's' (this is a french classic), with the sound [], like in the end of croissant (again, no final 't'), which doesn't exist in many languages.
When working on Unix systems, it is easy to get lost at C. Let's hope she'll hack her way to the coast.
Still it is difficult not to feel a little shame from the fact we all belong to the same species. Of course you are by no way responsible for the actions of others, but you can still fell a little sad and ashamed by the incommensurable stupidity of our fellows.
I am exactly in the same case. It has been a very long time since I wiped out any windows off my computers at home (never had one at work). I am so used to linux it would take me ages to regain my productivity if I were to switch to another OS.
Moreover, as far as I know, Windows is still no free software. I cannot count the number of things I learned thanks to the openness of linux by just browsing the code (things on filesystems, memory management, processes and scheduling, etc). Plus colsed things do not correspond to my ethics, which is a personal choice.
What is the point in having a TV or not? I do not get it.
As far as I can tell, "AI" has succeeded only in keeping the same name after endless redefinitions resulting from it's numerous failures.
Your blind faith in AI seems to indicate that you're either hopelessly misguided or one of those singularity nuts.
Absolutely not. The improvement in fields that were said to be impossible for AI are just astonishing, and I am among the first surprised by such successes. Let's state it clear: computing power is increasing, theoretical models are improving, practical implementations are getting more efficient. So yeah, basically Turing was right, we are just impressively capable computers and nothing more.
Around 5 or 6 years ago, there were some image classification benchmarks that were incredibly tough and said to be almost impossible to solve with a machine, with very low accuracies. Were are we now? Well the improvements have been far better than expected. Far better than I expected, to be honest. In some sense I would have loved if it didn't, since the pressure of this ever growing progress is stressful to my students and complicates the publication of novel ideas. But basically, yeah, it is improving a lot. It is science, it just works.
You could argue computer vision is not AI, but it is. Everything that allows a computer to make a statement that you thought was only possible by a human is AI.
Again, I am not interested in the the colorful stories about consciousness or whatever. What I'm saying is that there is basically no task that a computer will not be able to perform in the long run. Get over it, we are all replaceable by machines. Engineers, researchers, artists, name what you want, it is only a question of time and not of possibility.
It is bound to fail trying to do it with AI.
And still you know it will work very well in the long run, like almost every other task that was set to fail with AI.
Well, the truck can deliver the goods to a local market. Then, you can go to that market using your feet or even a bike. I guess it is even more green. It is the way our grandparents did. Why do we different? Because we have plenty of cheap energy and it is more comfortable the other way.
It might change when the energy will not be that cheap, though. I am pretty pessimistic at the idea some environmental enlightenment will win against laziness...
These exact same fears were written about in 1980. There was a famous BBC TV programme about how robots and microprocessors would replace everyone.
We already know the outcome.
I have the greatest difficulties with the assumption our society is a stationary process. Especially when most of the things around us are expected to grow exponentially. These 2 are mutualy exclusive, that is, you should take one but not both. Either you take the stationary society, and the "it was so before, thus it will be so in the future" is a very valid argument. Or you take the exponential growth everyone is looking for, and things will surely not be in the future as they were in the past.
With that kind of thinking, I'm surprised you aren't advocating the abolition of payment for all jobs. Doctors, teachers, taxi drivers - they should all work for free according to this argument, right?
You didn't understand his point. He says that whether you pay artists or not, they will continue to create new things, because the primary reason they do it is that they like it much more than everything else. Which may obviously not be the case of you taxi driver, who takes his job as a necessity to survive instead of a pleasant activity.
So yeah, you can basically cut some art revenues with little (if any) effect on art creation.
Here are more curves that were posted in the comments of the blog you're linking:
https://sites.google.com/site/arctischepinguin/home/piomas
Clearly, the exponential model has the best fit (which is not very surprising), and says 2015, take or give 1 year for 95% confidence. Of course, there is no theoretical model behind, but most of the time, the theoretical explanation comes after the empirical fit.
Same for speed. Unless you have a brain dead "repeat 1000 times" benchmark, Java is as fast as any other language.
After all: it gets compiled down to the same machine code ...
Not exactly. Java is unable to vectorize floating point operations now (might change with java8 though), which is very common in any multimedia application. So it remains a lot slower than what you get by default in C/C++ with a decent compiler.
There are a ton of off-the-shelf machine learning toolkits that are sufficient for 90% of possible use cases. The problem is getting annotated data to feed into these tools so they can learn the appropriate patterns. But all that requires is a host of annotators (i.e. undergrads and interns), not machine learning experts.
Exactly this!
Almost everything you ever dreamed of as a non machine learning expert is available at https://mloss.org/software/ ;-)
Please now annotate more data so that we can tune the algorithms
Is sustained focus even possible in mass audiences anymore? If not, what have we lost?
As an associate professor at the university, I can tell you many students have lost sustained focus, even in very small groups. If an explanation takes longer than 5 minutes, you lose them. If a problem takes longer than 5 minutes to solve, you lose them too. Starting 2 years ago, I modified all my lectures to have like "breakpoints" very often, so that no-one gets lost.
However, I think we already lost the Cartesian approach to breaking problems into smaller tasks. If you give them a rather simple but big problem, very few students are able the break it down and solve each part. Most will just try a global solution for a few minutes, then try the internet for a global solution, and finally get bored and say it's too complicated. One of my hypotheses is that the internet permits to solve most of the problems instantaneously, so you don't need sustained attention anymore. For the few cases where it is needed, well, that's the difference between the elite and the others...
Well, many trades will never be outsourced, or replaced by a robot.. namely landscaping, plumbing and electrical work. Much of which isn't safe for robots, and likely won't be for a very long time.
I would not bet a penny on that...
You might be surprised, but a lot of people in France think to go back to 40 hours a week, because they want higher wages whatever the unemployment rate. They don't get that going from 35 to 40 roughly fires 1 in 8 workers (of course it's not exactly true, because people get better wages, and this tends to create jobs - or the company doesn't raise the wages and instead invests in more machines to fire more workers).
I'm pretty sure it would be better for every one to lower this to 32 or 30 hours per week, and it won't have a significant impact on productivity. But the impact on the quality of living would be great.
Forget about miles and think in kilometers. In Europe, we say that 640k should be enough for anybody.
I would also have chosen the BY-NC-ND package even if I don't care about the NC aspect just because it is the only one to have the ND claim. This one is fundamental for a research paper.
If you take into account the time spent to write a good paper, every single word has been carefully crafted for hours. The idea to allow paraphrase or remixes is at best non-sense, most of the time it's just a very bad idea.
I'm pretty sure the authors in the study choose ND, and what ever the remaining condition, because as a researcher, there is just no way I could allow you to make me say something I was not meaning to say in the first place.
How far all thess jokes will go until we decide collectively for a stop, and just throw all those IP crap out the window?
I mean, seriously?