Having taken several courses on AI, I never found a contributor to the field that promised it to be the silver bullet -- or even remotely comparable to the human mind.
There's also something wrong with spouting forth conclusions and condemning the opposing viewpoint as being idiotic without citing any evidence (which makes this somewhat ironic, I guess).
The Wikipedia article on the subject is convoluted and doesn't really offer any strong conclusions, but at least some studies reported in the article have suggested a small negative correlation between intelligence and fertility (i.e., number of offspring), and another study showed a strong negative correlation between education and fertility (and education is sometimes used as a proxy for measuring intelligence). There's also a well-known negative correlation between economic well-being and fertility which may be related.
Small Glurmo: But, your Highness, she's a commoner. Her Slurm will taste foul. Slurm Queen: Yes, which is why we'll market it as New Slurm. Then, when everyone hates it, we'll bring back Slurm Classic and make billions!
Sorry, but that doesn't even read on the independent claim, which requires that the OS disables some function first, displays an ad, and then enables the disabled function in response to the ad ending.
I haven't read the Nature article (no free access from home), but my guess is that the "adaptive" mutations are indicative of a brittle genome where most mutations result in a drastic decrease in fitness, such that the individual mutations aren't propagated very well and thus aren't detectable amid the large population of bacteria.
In the latter case, it sounds as though a particular portion of the genome was rendered inert or unimportant, such as by a modification of the metabolic pathways to eliminate a portion of it, or through some duplication of genes such that parts of the copies are unimportant. That is, there is no difference in fitness arising from mutations in these portions of the genome, and the result is random diffusion across a fairly flat region of fitness space.
An alternative is that there was some auto-repair mechanism that was disabled by mutation at some point, and the mechanism was somehow set up to prevent mutations except in a very few areas (such as areas which typically allow for adaptation to environmental change).
For the past 20 years or so, anytime you needed a robocopter chasing someone down a dark alley, computer graphics was the answer. But how cool would it be to do this as a practical effect with no wires involved? Of course, people are so used to CG effects that they'd probably think the real thing looked fake.
P.S. Michael Bay can shove this robocopter up his ass. And I seriously hope he does.
"Law like a free software project" would at least require a patch to the patent code to make it more efficient at rejecting obvious inventions.
The Supreme Court submitted the KSR patch to the case law branch back in 2007 which helped tremendously with this bug.
Most of the problem now seems to be that since patent claims resemble Perl scripts, most users end up reading the comments at the top of the file rather than the claim code because it's easier to understand. Then they start submitting bug reports based on the comments without even finding out whether the new code conflicts with other modules that are already loaded.
Actually, if you were to publish your paper now, and then within a year file a provisional patent application where the specification was just your paper again, and then within another year you file the nonprovisional application, you would still be in the clear - as long as the inventorship on your application and the authorship on your paper were the same, and as long as everything in your claims is mentioned in your provisional app.
The difference between Vger and TNG was that in Vger, the fantastic technological solution would be both discovered and successfully implemented in the last three minutes of the episode. Also, it would restore everything to pretty much the way it was when the episode started.
Not all of those have anything to do with FOSS. One of them, for example, is about Electronic Gaming Monthly going out of business and redeeming unused partial subscriptions with issues of Maxim. Another is about GoDaddy's advertisements.
But I don't have time to parse them all. Could you go through the list and pick out the incidents which (a) actually have something to do with FOSS, and (b) indicate a prevalence of sexism in the FOSS community that goes beyond there being occasional assholes in the community in proportions similar to those found in real life?
There's also the brute-force approach of sampling sub-images from the image at a ton of different sizes and locations and then feeding the sub-images to your classifier. And for controlled images (such as in face recognition) normalizing the image isn't too difficult (by scaling the image to normalize the distance between the eyes). Even then, dimension reduction by PCA and/or LDA is useful to reduce the needed complexity of the classifier.
FFNNs lost their popularity probably 15 or 20 years ago and haven't been touched very often in image analysis since then. The SIFT features you mentioned seem to hold a lot of promise, and there are already folks using them in web-based image search and video indexing.
By interviewing Frist, a former Senator who was the Senate Republican leader during part of his time in office, instead of some other well-known physician, Maher interposed wholly unrelated politics into the discussion about whether or not to get an H1N1 vaccination. The end result is to convince some people who disagree with Frist on other issues to accept what is essentially a "reductio ad Hitlerum" argument: that if Bill Frist believes you should get a flu shot, then clearly, that's reason enough not to.
I agree with you. In the interest of full disclosure, I'm a little bit right of center, and I voted for McCain (although I'm not a Republican). I'm willing to give Obama a fair shake for what he does while in office, and if he does wonderful things over the next few years, then give the man a truckload of awards. But really, let's wait and see what he does first.
Of course, now the Nobel Peace Prize folks have shot their wad, so what will they do in a few years if Obama really does bring us world peace?
All told, this was really their way of giving G.W. Bush negative-one Nobel Peace Prizes. They should have just *said* that (and maybe "fined" Bush a million dollars or whatever the prize is now, just for grins), instead of further misusing the actual prize.
That's part of why the Soviet Union developed nukes, because without the ability to retaliate, they felt themselves to be at the mercy of the US.
But fast forward to today. Considering (a) the huge number of nukes available, and (b) the successful treaty organizations which ensure that an attack against almost any developing or developed nation would bring into play a nuclear-armed nation, there's only one reason that any nation would want to obtain nukes themselves: to use them, consequences be damned.
This goes double for Iran, whose leaders are driven in part by religious ideology and irrational hatred of Israel, and for North Korea, whose leader is outright bat-shit insane.
Not to nitpick an interesting post, but patent terms in the US are 20 years from the filing date (plus any patent term adjustments resulting from the USPTO backlog, which can be 3 years or more in some cases). The law was changed in 1995 to address "submarine patents". Any patents on applications filed before the law was changed are still granted the 17 year term from the date of issue, but there are only a very few applications still pending from those days.
It's right there in 35 USC 101. In fact, I quoted the relevant part in my previous post.
A general purpose computer programmed with software is a machine, and 35 USC 101 says that machines are patent-eligible. A computer-readable storage medium embodying software which, when executed, performs the steps of some method is an article of manufacture, and 35 USC 101 says that articles of manufacture are patent-eligible.
Slashdot Mirror
How much for a really big styrofoam cooler?
Having taken several courses on AI, I never found a contributor to the field that promised it to be the silver bullet -- or even remotely comparable to the human mind.
Douglas Lenat, perhaps?
the paranoia of /. getting turned up to 11
Frankly, I'm always paranoid about Slashdot getting turned up to 11.....
There's also something wrong with spouting forth conclusions and condemning the opposing viewpoint as being idiotic without citing any evidence (which makes this somewhat ironic, I guess).
The Wikipedia article on the subject is convoluted and doesn't really offer any strong conclusions, but at least some studies reported in the article have suggested a small negative correlation between intelligence and fertility (i.e., number of offspring), and another study showed a strong negative correlation between education and fertility (and education is sometimes used as a proxy for measuring intelligence). There's also a well-known negative correlation between economic well-being and fertility which may be related.
Get off my lawn!
Small Glurmo: But, your Highness, she's a commoner. Her Slurm will taste foul.
Slurm Queen: Yes, which is why we'll market it as New Slurm. Then, when everyone hates it, we'll bring back Slurm Classic and make billions!
Sorry, but that doesn't even read on the independent claim, which requires that the OS disables some function first, displays an ad, and then enables the disabled function in response to the ad ending.
I dunno - it seems pretty apt to me. It does facilitate the whole "eating your own babies" concept, after all....
I haven't read the Nature article (no free access from home), but my guess is that the "adaptive" mutations are indicative of a brittle genome where most mutations result in a drastic decrease in fitness, such that the individual mutations aren't propagated very well and thus aren't detectable amid the large population of bacteria.
In the latter case, it sounds as though a particular portion of the genome was rendered inert or unimportant, such as by a modification of the metabolic pathways to eliminate a portion of it, or through some duplication of genes such that parts of the copies are unimportant. That is, there is no difference in fitness arising from mutations in these portions of the genome, and the result is random diffusion across a fairly flat region of fitness space.
An alternative is that there was some auto-repair mechanism that was disabled by mutation at some point, and the mechanism was somehow set up to prevent mutations except in a very few areas (such as areas which typically allow for adaptation to environmental change).
For the past 20 years or so, anytime you needed a robocopter chasing someone down a dark alley, computer graphics was the answer. But how cool would it be to do this as a practical effect with no wires involved? Of course, people are so used to CG effects that they'd probably think the real thing looked fake.
P.S. Michael Bay can shove this robocopter up his ass. And I seriously hope he does.
"Law like a free software project" would at least require a patch to the patent code to make it more efficient at rejecting obvious inventions.
The Supreme Court submitted the KSR patch to the case law branch back in 2007 which helped tremendously with this bug.
Most of the problem now seems to be that since patent claims resemble Perl scripts, most users end up reading the comments at the top of the file rather than the claim code because it's easier to understand. Then they start submitting bug reports based on the comments without even finding out whether the new code conflicts with other modules that are already loaded.
Actually, if you were to publish your paper now, and then within a year file a provisional patent application where the specification was just your paper again, and then within another year you file the nonprovisional application, you would still be in the clear - as long as the inventorship on your application and the authorship on your paper were the same, and as long as everything in your claims is mentioned in your provisional app.
Like vampires & ghouls fear the light of day, the RIAA lawyers fear the truth.
Also goes to show just how high the stakes are here.
ba dum bum
Thanks, I'll be here all week.
The difference between Vger and TNG was that in Vger, the fantastic technological solution would be both discovered and successfully implemented in the last three minutes of the episode. Also, it would restore everything to pretty much the way it was when the episode started.
Not all of those have anything to do with FOSS. One of them, for example, is about Electronic Gaming Monthly going out of business and redeeming unused partial subscriptions with issues of Maxim. Another is about GoDaddy's advertisements.
But I don't have time to parse them all. Could you go through the list and pick out the incidents which (a) actually have something to do with FOSS, and (b) indicate a prevalence of sexism in the FOSS community that goes beyond there being occasional assholes in the community in proportions similar to those found in real life?
There's also the brute-force approach of sampling sub-images from the image at a ton of different sizes and locations and then feeding the sub-images to your classifier. And for controlled images (such as in face recognition) normalizing the image isn't too difficult (by scaling the image to normalize the distance between the eyes). Even then, dimension reduction by PCA and/or LDA is useful to reduce the needed complexity of the classifier.
FFNNs lost their popularity probably 15 or 20 years ago and haven't been touched very often in image analysis since then. The SIFT features you mentioned seem to hold a lot of promise, and there are already folks using them in web-based image search and video indexing.
As far as I can tell, he's a 26 year old programmer/blogger who doesn't much like to program in his free time.
Also, he needs to get the hell off my lawn.
That was essentially Maher's ploy.
By interviewing Frist, a former Senator who was the Senate Republican leader during part of his time in office, instead of some other well-known physician, Maher interposed wholly unrelated politics into the discussion about whether or not to get an H1N1 vaccination. The end result is to convince some people who disagree with Frist on other issues to accept what is essentially a "reductio ad Hitlerum" argument: that if Bill Frist believes you should get a flu shot, then clearly, that's reason enough not to.
I agree with you. In the interest of full disclosure, I'm a little bit right of center, and I voted for McCain (although I'm not a Republican). I'm willing to give Obama a fair shake for what he does while in office, and if he does wonderful things over the next few years, then give the man a truckload of awards. But really, let's wait and see what he does first.
Of course, now the Nobel Peace Prize folks have shot their wad, so what will they do in a few years if Obama really does bring us world peace?
All told, this was really their way of giving G.W. Bush negative-one Nobel Peace Prizes. They should have just *said* that (and maybe "fined" Bush a million dollars or whatever the prize is now, just for grins), instead of further misusing the actual prize.
No, because Israel already has nukes - or at least, it's been widely suspected that Israel has had nukes for some time.
That's part of why the Soviet Union developed nukes, because without the ability to retaliate, they felt themselves to be at the mercy of the US.
But fast forward to today. Considering (a) the huge number of nukes available, and (b) the successful treaty organizations which ensure that an attack against almost any developing or developed nation would bring into play a nuclear-armed nation, there's only one reason that any nation would want to obtain nukes themselves: to use them, consequences be damned.
This goes double for Iran, whose leaders are driven in part by religious ideology and irrational hatred of Israel, and for North Korea, whose leader is outright bat-shit insane.
No, Bilski and Warmerdam were addressing two different issues altogether.
Actually, In re Warmerdam ruled out claiming software directly.
Not to nitpick an interesting post, but patent terms in the US are 20 years from the filing date (plus any patent term adjustments resulting from the USPTO backlog, which can be 3 years or more in some cases). The law was changed in 1995 to address "submarine patents". Any patents on applications filed before the law was changed are still granted the 17 year term from the date of issue, but there are only a very few applications still pending from those days.
It's right there in 35 USC 101. In fact, I quoted the relevant part in my previous post.
A general purpose computer programmed with software is a machine, and 35 USC 101 says that machines are patent-eligible. A computer-readable storage medium embodying software which, when executed, performs the steps of some method is an article of manufacture, and 35 USC 101 says that articles of manufacture are patent-eligible.