Civilization might return by making use of different resources. Also, even if civilization is destroyed, the accumulated knowledge it has built up would likely be more difficult to destroy (and thus civilization could possibly return more quickly). For example, the Aztecs did pretty well prior to their conquest by the Europeans considering that their technology was technically stone-age.
It would quite simply herald the end of regular humans. We'd augment, replace, or just die off, but the age of biological humans would come to a close - we couldn't compete with that.
Anyway, that's still a far way off, and just building models yields no actual understanding (until we start all of the ethically questionable poking around with the model to see how it works).
I know that IEEE measures "Patent Power" in some way that is based on citations. However, how is such strength objectively measured? Number of citations? Number of products marketed on those patents? How many scientists think the patent is useful? Scientists are subject to trends as much as others, and such a question of use tends to be predicated upon how "hot" the field is. For example, patents on SVMs were probably very well cited a few years ago; perhaps now techniques involving manifold learning and and Gaussian processes have replaced them.
The truth is that the ramifications of any scientific discovery cannot be fully assessed. Some things definitely have more applications than others, but I don't trust anyone who claims to have an objective measure of research quality - especially a quantitative one like IEEE's. I've seen many articles I'd consider great get rejected and I've seen many articles with incorrect formulas or results that were later refuted win "best paper" awards. I've even seen the same papers get rejected at one conference then win an award at another with very few changes. Likewise, I've seen articles I'd call poor garner more citations than those I'd consider better.
To use some famous specific examples, neural network research was stalled for a decade because people misunderstood Minsky's book on perceptrons, people once thought it ridiculous that bacteria could cause stomach ulcers, and what is now referred to as one of the centerpieces of Objectivist philosophy, The Fountainhead, was rejected by 12 different publishers before finally being published. That's saying nothing about the even more astounding controversies over, say, heliocentricism and evolution, which tend to become charged with special interests.
All of this has thoroughly convinced me that people have no way of judging the quality of research... or ideas in general.
Radio is a bit different, however. With all of the other forms of advertising you mentioned, the consumer is making a deliberate choice to attend a specific concert or buy a specific CD. The consumer already has an idea of the music he or she wishes to buy. With radio, this is not necessarily the case - radio is nonspecific, and therefore is likely the medium in which the first exposure to new music will occur.
Now, different arguments can be made as to what a consumer buys when he purchases a CD. The music itself, certainly - but he can already listen to that for free by waiting for it to appear on the radio. In my opinion, what the consumer buys when he buys a CD is choice - the choice to listen to a particular song whenever he wishes rather than waiting for it to appear whenever a radio station plays it. The radio then becomes the advertisement for this purchase.
I doubt they're interested in this sort of thing right now. I submitted a paper on how to do multimedia similarity search to them when interviewing there, and was told that similarity-based image search isn't an area they're concerned with at the moment. Because it isn't, it's probably a better idea for him to go into it on his own (also, once they see it in action, they might want it).
Also, everyone in computer graphics has some sort of image similarity search method, and I don't see anything particularly novel about this one. Wake me up when he has some classification accuracies or something showing that his method works better than the others.
Ah, but can such an approximation truly result in intelligence or would it just be a very slow and complex classifier? Or both?:)
Are all of the functions of a neuron required to produce intelligent behavior? If not, which can we omit? How will we even know when a system is behaving intelligently? Even humans take years to learn how to communicate and rationalize. Could we provide even a perfect simulation of the human brain the proper environment to train in to ensure these results? Once we have such a model, can we extend it or are we limited to the level of artificial people? Is it ethical to perform questionable experiments on a model with an essentially human brain in attempts to extend it?
These are some of the questions we need to confront before taking this route (another should always be "do we really want to do this?"). It's not enough to just say "connect lots of wires and we have intelligence". We have to understand at least some of the theory linking the low and high level functionality before we can hope to construct an intelligent agent based on this model, and I suspect we'll need a much more complete understanding to (a) make it computationally tractable and (b) generalize the system to above human-level intelligence.
It was more a matter of range than correctness (although as the poster below pointed out, "volts" would still be correct because the human body can be treated like a resistor). 1kA doesn't seem a likely number to suggest, so I figured he probably meant 1kV.
If they were 500 light years away, it would take at least 500 years for our signal to reach them and presumably at least 500 years for them to reply (probably much longer), unless they have some method of faster-than-light travel. By that time, we'd either be on more than one world or extinct.
In fact, I've always taken the fact that we haven't found extraterrestrial life as empirical evidence that FTL travel doesn't exist (although it could also be a matter of sheer scale even if FTL is possible). It solves the Fermi paradox nicely.
That's any sort of elite, not just politicians. For example, it's a lot easier to get into a good grad. school if your professors know someone there. Likewise for corporate "networking" and jobs.
I just received a fake call from "George W" that was effectively meant as a smear campaign against "the Republicans". "The Republicans" are no better; they've been calling my house multiple times daily with a fake caller ID # to sling mud at "the Democrats". Political Joe jobs and other nasty things don't only happen on the Internet.
(Of course, online, any idiot could do this, whereas calling people requires a bit more coordination and resources).
But honestly, we should be asking ourselves if we want people who stoop to such measures to make the policy for our country in the first place. I don't think I'm voting for any of them.
From an individual perspective, yes, especially if you're rooted to one society (otherwise just move wherever the innovation is going). But in terms of total innovation by humanity, it isn't a major issue.
If societies tend to become complacent and pass anti-innovative policy such as this once they've achieved power, all the better; it lets other societies advance (relatively speaking), which encourages competition, which spurs innovation like nothing else.
Thus, I'm not happy about this as an individual, particularly because I'm a scientist who is philosophically opposed to filing for any patents of my own, but at the same time, I'm glad to see indications that individual societies are still making the sorts of dumb decisions that further all of us in the long run.
You act as if this is a bad thing. Evolution works on the scale of societies as well. If one society becomes unfriendly to innovation, it will be marginalized as it loses technical dominance. Ultimately the most powerful societies will be the ones that are friendly to innovation because the others couldn't keep up. So it works out nicely.
Doing a thorough analysis of several unique sort when you were trying to push it on Wikipedia revealed no asymptotic advantage over basic bubble sort. I'm not going to repeat the analysis with this algorithm, but I'm inclined to be skeptical based on that experience. Also, perhaps you have a non-asymptotic advantage over quicksort (or let's say a modified quicksort without the O(n^2) worst-case complexity), but the O(n log n) bound on comparison sorting has been proved optimal.
They've had people working on their algorithms for quite some time now. I doubt it's in the state where it's something you can just give away all at once... or precisely target, for that matter. It's probably hundreds of thousands of lines of code by now, if not more. They should have systems in place to notify them when that much data is copied at once.
Still waiting for them to allow weighting of search terms, though:)
Or if they have cameras in the room with the computers, which they probably would. Being a member of the cleaning staff wouldn't qualify you to install software on the system under any pretense.
Hopefully without breaking the NDA, I should mention that people at Google looked at me strangely when I suggested that they blur faces on street view. They couldn't understand why the privacy implications of such a service are a problem, as what they are doing is technically legal in the USA. However, when people are posting images of random people picking their noses or something on Digg for millions to gawk at (and such things have appeared even on the Digg front page from time to time), there's a problem - it can ruin someone's reputation for a rather stupid reason if the person is identified. To me, that's evil. To them, fixing it should be the cautious thing to do so they don't get sued (weren't they already involved in a lawsuit for this?), even if it happens to jive with their morals.
I don't know if the "don't be evil" thing is practiced as rigorously by the individual employees there as the company would like you to believe. Creating nifty things seems to win out over most moral considerations; at least, this was the impression I got while I was there. Nifty things are good, but people should think about how their technology is going to be used rather than just what they could make.
If this happens, Microsoft will probably come out with cheaper editions of Windows that limit your capabilities in additional ways. Then you'll end up with $250 PCs with a $30 Windows "Starter Edition" or something. A lower Microsoft tax, but also a far less useful system. Linux might still stand to gain due to the dramatically increased functionality relative to Windows that this will give it.
And wait until you see what it's like if you do!:)
Seriously, doing a Ph. D. is depressing. My experience is that it's essentially a source of cheap labor for your professors, usually involving tasks that are rather devoid of original thought, such as running the experiments and writing the papers. It's training in every aspect of science except actual science. The workload is heavy, the days are long, and you frequently don't get credit for your work. In a challenging school, the work would probably be difficult too; if you're not so fortunate, it will simply be mind-numbing. And forget it if you want to study something that isn't "hot" or well-funded! At best you'll be rejected in admissions since no one can match a faculty member to you, but if your credentials are too good for that, you'll be dragged into a well-funded field that doesn't interest you... but only after you've gone too far to turn back. It's not an enviable position under any circumstance, and only the hope of an end to it keeps most students going.
Slashdot Mirror
David Brin also explored this concept. IIRC, the book was called "The Transparent Society".
Civilization might return by making use of different resources. Also, even if civilization is destroyed, the accumulated knowledge it has built up would likely be more difficult to destroy (and thus civilization could possibly return more quickly). For example, the Aztecs did pretty well prior to their conquest by the Europeans considering that their technology was technically stone-age.
It would quite simply herald the end of regular humans. We'd augment, replace, or just die off, but the age of biological humans would come to a close - we couldn't compete with that.
Anyway, that's still a far way off, and just building models yields no actual understanding (until we start all of the ethically questionable poking around with the model to see how it works).
Well back in my day we shifted disks on an abacus and made the beeping noises ourselves! :)
I know that IEEE measures "Patent Power" in some way that is based on citations. However, how is such strength objectively measured? Number of citations? Number of products marketed on those patents? How many scientists think the patent is useful? Scientists are subject to trends as much as others, and such a question of use tends to be predicated upon how "hot" the field is. For example, patents on SVMs were probably very well cited a few years ago; perhaps now techniques involving manifold learning and and Gaussian processes have replaced them.
The truth is that the ramifications of any scientific discovery cannot be fully assessed. Some things definitely have more applications than others, but I don't trust anyone who claims to have an objective measure of research quality - especially a quantitative one like IEEE's. I've seen many articles I'd consider great get rejected and I've seen many articles with incorrect formulas or results that were later refuted win "best paper" awards. I've even seen the same papers get rejected at one conference then win an award at another with very few changes. Likewise, I've seen articles I'd call poor garner more citations than those I'd consider better.
To use some famous specific examples, neural network research was stalled for a decade because people misunderstood Minsky's book on perceptrons, people once thought it ridiculous that bacteria could cause stomach ulcers, and what is now referred to as one of the centerpieces of Objectivist philosophy, The Fountainhead, was rejected by 12 different publishers before finally being published. That's saying nothing about the even more astounding controversies over, say, heliocentricism and evolution, which tend to become charged with special interests.
All of this has thoroughly convinced me that people have no way of judging the quality of research... or ideas in general.
Radio is a bit different, however. With all of the other forms of advertising you mentioned, the consumer is making a deliberate choice to attend a specific concert or buy a specific CD. The consumer already has an idea of the music he or she wishes to buy. With radio, this is not necessarily the case - radio is nonspecific, and therefore is likely the medium in which the first exposure to new music will occur.
Now, different arguments can be made as to what a consumer buys when he purchases a CD. The music itself, certainly - but he can already listen to that for free by waiting for it to appear on the radio. In my opinion, what the consumer buys when he buys a CD is choice - the choice to listen to a particular song whenever he wishes rather than waiting for it to appear whenever a radio station plays it. The radio then becomes the advertisement for this purchase.
I doubt they're interested in this sort of thing right now. I submitted a paper on how to do multimedia similarity search to them when interviewing there, and was told that similarity-based image search isn't an area they're concerned with at the moment. Because it isn't, it's probably a better idea for him to go into it on his own (also, once they see it in action, they might want it).
Also, everyone in computer graphics has some sort of image similarity search method, and I don't see anything particularly novel about this one. Wake me up when he has some classification accuracies or something showing that his method works better than the others.
Ah, but can such an approximation truly result in intelligence or would it just be a very slow and complex classifier? Or both? :)
Are all of the functions of a neuron required to produce intelligent behavior? If not, which can we omit? How will we even know when a system is behaving intelligently? Even humans take years to learn how to communicate and rationalize. Could we provide even a perfect simulation of the human brain the proper environment to train in to ensure these results? Once we have such a model, can we extend it or are we limited to the level of artificial people? Is it ethical to perform questionable experiments on a model with an essentially human brain in attempts to extend it?
These are some of the questions we need to confront before taking this route (another should always be "do we really want to do this?"). It's not enough to just say "connect lots of wires and we have intelligence". We have to understand at least some of the theory linking the low and high level functionality before we can hope to construct an intelligent agent based on this model, and I suspect we'll need a much more complete understanding to (a) make it computationally tractable and (b) generalize the system to above human-level intelligence.
It was more a matter of range than correctness (although as the poster below pointed out, "volts" would still be correct because the human body can be treated like a resistor). 1kA doesn't seem a likely number to suggest, so I figured he probably meant 1kV.
Ohms are a measure of resistance. Did you mean volts?
If they were 500 light years away, it would take at least 500 years for our signal to reach them and presumably at least 500 years for them to reply (probably much longer), unless they have some method of faster-than-light travel. By that time, we'd either be on more than one world or extinct.
In fact, I've always taken the fact that we haven't found extraterrestrial life as empirical evidence that FTL travel doesn't exist (although it could also be a matter of sheer scale even if FTL is possible). It solves the Fermi paradox nicely.
Ah, you're right; the 0 is neither. I don't gamble much :)
(13 c 5) * .48^5 * .52^8 = .18, to be exact.
That's any sort of elite, not just politicians. For example, it's a lot easier to get into a good grad. school if your professors know someone there. Likewise for corporate "networking" and jobs.
The next step is to automatically analyze them. Like researchers (including myself) are doing already.
I just received a fake call from "George W" that was effectively meant as a smear campaign against "the Republicans". "The Republicans" are no better; they've been calling my house multiple times daily with a fake caller ID # to sling mud at "the Democrats". Political Joe jobs and other nasty things don't only happen on the Internet.
(Of course, online, any idiot could do this, whereas calling people requires a bit more coordination and resources).
But honestly, we should be asking ourselves if we want people who stoop to such measures to make the policy for our country in the first place. I don't think I'm voting for any of them.
From an individual perspective, yes, especially if you're rooted to one society (otherwise just move wherever the innovation is going). But in terms of total innovation by humanity, it isn't a major issue.
If societies tend to become complacent and pass anti-innovative policy such as this once they've achieved power, all the better; it lets other societies advance (relatively speaking), which encourages competition, which spurs innovation like nothing else.
Thus, I'm not happy about this as an individual, particularly because I'm a scientist who is philosophically opposed to filing for any patents of my own, but at the same time, I'm glad to see indications that individual societies are still making the sorts of dumb decisions that further all of us in the long run.
You act as if this is a bad thing. Evolution works on the scale of societies as well. If one society becomes unfriendly to innovation, it will be marginalized as it loses technical dominance. Ultimately the most powerful societies will be the ones that are friendly to innovation because the others couldn't keep up. So it works out nicely.
Doing a thorough analysis of several unique sort when you were trying to push it on Wikipedia revealed no asymptotic advantage over basic bubble sort. I'm not going to repeat the analysis with this algorithm, but I'm inclined to be skeptical based on that experience. Also, perhaps you have a non-asymptotic advantage over quicksort (or let's say a modified quicksort without the O(n^2) worst-case complexity), but the O(n log n) bound on comparison sorting has been proved optimal.
They've had people working on their algorithms for quite some time now. I doubt it's in the state where it's something you can just give away all at once... or precisely target, for that matter. It's probably hundreds of thousands of lines of code by now, if not more. They should have systems in place to notify them when that much data is copied at once.
:)
Still waiting for them to allow weighting of search terms, though
Or if they have cameras in the room with the computers, which they probably would. Being a member of the cleaning staff wouldn't qualify you to install software on the system under any pretense.
I was waiting for someone to say that.
Hopefully without breaking the NDA, I should mention that people at Google looked at me strangely when I suggested that they blur faces on street view. They couldn't understand why the privacy implications of such a service are a problem, as what they are doing is technically legal in the USA. However, when people are posting images of random people picking their noses or something on Digg for millions to gawk at (and such things have appeared even on the Digg front page from time to time), there's a problem - it can ruin someone's reputation for a rather stupid reason if the person is identified. To me, that's evil. To them, fixing it should be the cautious thing to do so they don't get sued (weren't they already involved in a lawsuit for this?), even if it happens to jive with their morals.
I don't know if the "don't be evil" thing is practiced as rigorously by the individual employees there as the company would like you to believe. Creating nifty things seems to win out over most moral considerations; at least, this was the impression I got while I was there. Nifty things are good, but people should think about how their technology is going to be used rather than just what they could make.
If this happens, Microsoft will probably come out with cheaper editions of Windows that limit your capabilities in additional ways. Then you'll end up with $250 PCs with a $30 Windows "Starter Edition" or something. A lower Microsoft tax, but also a far less useful system. Linux might still stand to gain due to the dramatically increased functionality relative to Windows that this will give it.
It isn't just money, though. I like doing research. It's something very few people would let me touch without a Ph. D.
(The nice thing about mathematically-derived fields is that no one can deny me the resources necessary to do it, however, Ph. D. or no).
And wait until you see what it's like if you do! :)
Seriously, doing a Ph. D. is depressing. My experience is that it's essentially a source of cheap labor for your professors, usually involving tasks that are rather devoid of original thought, such as running the experiments and writing the papers. It's training in every aspect of science except actual science. The workload is heavy, the days are long, and you frequently don't get credit for your work. In a challenging school, the work would probably be difficult too; if you're not so fortunate, it will simply be mind-numbing. And forget it if you want to study something that isn't "hot" or well-funded! At best you'll be rejected in admissions since no one can match a faculty member to you, but if your credentials are too good for that, you'll be dragged into a well-funded field that doesn't interest you... but only after you've gone too far to turn back. It's not an enviable position under any circumstance, and only the hope of an end to it keeps most students going.
Being smart can indeed suck.