They can, but that increases the cost. The real danger is using cheap human power to train a statistical engine that can then solve CAPTCHAs without human assistance. For example, prebuilding a database of a million questions that only humans can answer would easily fall victim to this. Even if your CAPTCHA solver is only right 10% of the time, that's plenty for spamming.
You have to consider the source of the questions. If the questions are human-generated, it's not economically feasible. Remember that they can train their CAPTCHA-defeating software by paying large numbers of people to supply the answers to CAPTCHAs. Even a very large database could fall to that approach.
If the questions are machine-generated, then you're pitting a machine generating questions and answers against a machine designed to answer questions.
This is a very good analogy. The only problem is that after a while, people care less about your degree (though in some fields, it might still matter to an extent). Publishing in highly visible, peer-reviewed is critical to getting funding and keeping an academic career.
Air resistance on reasonably aerodynamic objects is actually more like |v|^1.4, not v^2. Air resistance for objects like bricks (or trucks) is roughly v^2.
Nonlocal hidden variables aren't really "extras" -- the hidden variables would have to be communicating information instantaneously across an arbitrary distance.
The "still haven't" is misleading. Bell's clearly shows that the only possibility for a deterministic mechanism behind quantum mechanics is a system containing nonlocal hidden variables.
Well, it won't be useful in a home environment. Why do you assume it will? The article doesn't suggest that. It even mentions particular situations where this technology is useful.
At a 3 mm wavelength, they're not going to be able to diverge much from LOS.
How much the LOS diverges depends on the antennae and the wavelength. It certainly won't be like a laser pointer, even though they use laser microwave production, because of the larger wavelength. It won't diverge nearly as much as LOS radio communication, though.
Infrared, which is above the frequency they're using, will cook your dinner too.
In fact, many RF frequencies are converted to heat by solid matter. Just because it's less efficient than the microwave oven frequency doesn't mean it doesn't work.
I wouldn't consider university deployment to be consumer-grade.
There are some situations, like what you mentioned, where a consumer-level device might be useful. (However, currently it's line-of-sight on both ends using low-frequency lasers. That would require one transceiver for each connected house on top of that church tower!)
Depends on the physical size of the signal and the redundancy. Obviously if you're relying on a single link that's physically roughly the size of the bird or smaller, the signal will be disrupted as long as the bird is in the beam (plus, probably, a short recovery time). It's as sure an outage as a backhoe taking out your fiber, although shorter-lived.
The researchers seem to have a pretty good idea of exactly where it's useful -- they don't really mention replacing backbones or end-user links, but using it places like college campuses, where setting up redundant links between buildings would be quite reasonable, but uptime isn't as critical as it is for a backbone.
Unfortunately, as point-to-point or line-of-sight sort of implies, they're also completely useless in most consumer applications. Both transmitter and receiver have to point directly at one another. So it doesn't work like cellular, 802.11, BlueTooth, etc.
In fact, it'll probably never make it to consumer devices. We have a tough enough time with satellite television, which is sort of a sloppy line-of-sight transmission.
While there's certainly a engineering difference between a prototype and a consumer-ready model, most of the problems you mention probably won't apply to this. It's point-to-point (so, line-of-sight) communications, so it will require an unobstructed path. However, proper transmitters and receivers for line-of-sight communications won't have nearly the sort of interference problems you experience with something like a Bluetooth. Line-of-sight communications also means you can use quite efficient antennae, so should get quite good range.
It's much less interesting if you don't use hyperbole.
The FDA won't sell you the tests to test 100% of your cattle when they know the test is not effective on the cattle you're testing (as you are killing them before they are old enough for the test to work), as the only function of such a test is to gain a competitive advantage through false advertisement.
The article is terrible. They're actually looking at non-equilibrium Brownian ratchets, which is very different from a Brownian ratchet. Much like how they're not reexamining the second law of thermodynamics, they're reexamining its implications.
As I read it, the general idea seems to be that instead of simply burning electricity and disposing of the waste heat, they're considering reclaiming some of the waste heat to help power the device (which could help reduce its heat output). Of course, since they're consuming energy to perform calculations (which are entropy-reducing), they're required to emit a certain amount of uncapturable heat.
It's much more effective to use a well-respected numerical analysis package rather than rolling your own everything from scratch. NumPy, for example, lets you do mathematical analysis far better than Excel very quickly.
I agree that they're orthogonal opinions. I just happen to disagree about the "moral degeneracy" bit. While the environmental damage is nonzero, it's sort of a curveball, as your original comment was only about killing animals.
While it may not apply to you, I find it interesting that many "moral vegetarians" are in the naturalistic-fallacy camp, but animal husbandry and the human consumption of meat are hardly unnatural.
Please make any responses more clear. The phrase "massive environmental damage moral degeneracy" is tough to parse without additional words and/or punctuation.
Slashdot Mirror
They can, but that increases the cost. The real danger is using cheap human power to train a statistical engine that can then solve CAPTCHAs without human assistance. For example, prebuilding a database of a million questions that only humans can answer would easily fall victim to this. Even if your CAPTCHA solver is only right 10% of the time, that's plenty for spamming.
Your system is only really successful if all humans agree on the "correct" answer -- which basically rules out things like "most attractive human".
Computers actually can do incredibly well with statistics and extensive training.
If I read the article and summary correctly, it's exactly the sort of CAPTCHA you're suggesting that people have found a reasonably-good solution to.
Unfortunately, often these solutions aren't actually useful AI solutions.
You have to consider the source of the questions. If the questions are human-generated, it's not economically feasible. Remember that they can train their CAPTCHA-defeating software by paying large numbers of people to supply the answers to CAPTCHAs. Even a very large database could fall to that approach.
If the questions are machine-generated, then you're pitting a machine generating questions and answers against a machine designed to answer questions.
This is a very good analogy. The only problem is that after a while, people care less about your degree (though in some fields, it might still matter to an extent). Publishing in highly visible, peer-reviewed is critical to getting funding and keeping an academic career.
Air resistance on reasonably aerodynamic objects is actually more like |v|^1.4, not v^2. Air resistance for objects like bricks (or trucks) is roughly v^2.
Many people, especially younger people, in U.S. urban areas have no cars -- as they, too, can actually use the public transportation.
Nonlocal hidden variables aren't really "extras" -- the hidden variables would have to be communicating information instantaneously across an arbitrary distance.
It's not imprecise, it's just nondeterministic.
The "still haven't" is misleading. Bell's clearly shows that the only possibility for a deterministic mechanism behind quantum mechanics is a system containing nonlocal hidden variables.
Well, it won't be useful in a home environment. Why do you assume it will? The article doesn't suggest that. It even mentions particular situations where this technology is useful.
It is a laser -- just in the microwave spectrum instead of visible-light. Both are not trivial to snoop on. (Not like 802.11 wireless.)
At a 3 mm wavelength, they're not going to be able to diverge much from LOS.
How much the LOS diverges depends on the antennae and the wavelength. It certainly won't be like a laser pointer, even though they use laser microwave production, because of the larger wavelength. It won't diverge nearly as much as LOS radio communication, though.
Infrared, which is above the frequency they're using, will cook your dinner too.
In fact, many RF frequencies are converted to heat by solid matter. Just because it's less efficient than the microwave oven frequency doesn't mean it doesn't work.
I wouldn't consider university deployment to be consumer-grade.
There are some situations, like what you mentioned, where a consumer-level device might be useful. (However, currently it's line-of-sight on both ends using low-frequency lasers. That would require one transceiver for each connected house on top of that church tower!)
Also, depending on the power in the signal, it could be an interesting experience for the bird.
Depends on the physical size of the signal and the redundancy. Obviously if you're relying on a single link that's physically roughly the size of the bird or smaller, the signal will be disrupted as long as the bird is in the beam (plus, probably, a short recovery time). It's as sure an outage as a backhoe taking out your fiber, although shorter-lived.
The researchers seem to have a pretty good idea of exactly where it's useful -- they don't really mention replacing backbones or end-user links, but using it places like college campuses, where setting up redundant links between buildings would be quite reasonable, but uptime isn't as critical as it is for a backbone.
They're very different beasts.
Unfortunately, as point-to-point or line-of-sight sort of implies, they're also completely useless in most consumer applications. Both transmitter and receiver have to point directly at one another. So it doesn't work like cellular, 802.11, BlueTooth, etc.
In fact, it'll probably never make it to consumer devices. We have a tough enough time with satellite television, which is sort of a sloppy line-of-sight transmission.
While there's certainly a engineering difference between a prototype and a consumer-ready model, most of the problems you mention probably won't apply to this. It's point-to-point (so, line-of-sight) communications, so it will require an unobstructed path. However, proper transmitters and receivers for line-of-sight communications won't have nearly the sort of interference problems you experience with something like a Bluetooth. Line-of-sight communications also means you can use quite efficient antennae, so should get quite good range.
It's much less interesting if you don't use hyperbole.
The FDA won't sell you the tests to test 100% of your cattle when they know the test is not effective on the cattle you're testing (as you are killing them before they are old enough for the test to work), as the only function of such a test is to gain a competitive advantage through false advertisement.
The article is terrible. They're actually looking at non-equilibrium Brownian ratchets, which is very different from a Brownian ratchet. Much like how they're not reexamining the second law of thermodynamics, they're reexamining its implications.
As I read it, the general idea seems to be that instead of simply burning electricity and disposing of the waste heat, they're considering reclaiming some of the waste heat to help power the device (which could help reduce its heat output). Of course, since they're consuming energy to perform calculations (which are entropy-reducing), they're required to emit a certain amount of uncapturable heat.
I really hope you're joking.
It's much more effective to use a well-respected numerical analysis package rather than rolling your own everything from scratch. NumPy, for example, lets you do mathematical analysis far better than Excel very quickly.
That doesn't really sound like you're doing significant scientific data analysis.
For what? The three items you discuss do different things.
Presumably nobody uses LaTeX for making charts or for performing calculations (the latter of which is really what's being addressed here).
I agree that they're orthogonal opinions. I just happen to disagree about the "moral degeneracy" bit. While the environmental damage is nonzero, it's sort of a curveball, as your original comment was only about killing animals.
While it may not apply to you, I find it interesting that many "moral vegetarians" are in the naturalistic-fallacy camp, but animal husbandry and the human consumption of meat are hardly unnatural.
Please make any responses more clear. The phrase "massive environmental damage moral degeneracy" is tough to parse without additional words and/or punctuation.