Imagine you're running along in a one dimensional world. Any positioning errors will move any two points either directly towards each other or away from each other. The average bias should be zero.
If you allow two dimensions, errors can now occur to the right and left of your path, even if you're still running along in a straight line. Any lateral error, right or left, will increase the measured distance between points. That's where the bias comes from.
If I followed your post correctly, the effect you're referring to is the same as the interpolation error mentioned in the article, and actually tends to decrease the measured distance.
Middle eastern muslims thought pretty much the same thing about the murderous christian barbarians who invaded their heartland a thousand years ago. Both the bible and koran, and most other holy books, have some incredibly nasty bits. The vast majority of adherents to any major religion ignore those parts.
Terrorist organizations are generally very small. Attacks have two goals: terrorize the victims and get publicity to recruit more terrorists. If you shut down the publicity then you deny both those objectives.
Covering up terrorist attacks probably isn't a good idea. But publicizing them as a small number of criminals who committed a horrible crime is a lot better than millions of people wringing their hands and screaming about how the terrorists are winning.
It's both. It's a library for numerical computation using data flow graphs. Most people these days use such libraries to implement their machine learning... TensorFlow contains objects that do machine learning, on top of its lower level operations.
Tensors are used in differential geometry, but they're not limited to it. A tensor is just a multidimensional array. TensorFlow is built on a computation graph where tensors (arrays) flow between operations.
Yay! It's been a while since we had a nice armed standoff or riot. Of course, when the Mohawks are unhappy it can be difficult to get to the beach, so I hope they work it out by next summer.
Yup. I have a common name. Since 2001 I've gotten hassled when entering, travelling in, or flying near the US. Sometimes it's just extra "random" screening (five times while making a single connection in the US once). When crossing the border it's usually a forty minute interview in the back room.
The guy they're currently looking for scares the crap out of border guards. In airports, when they know you're coming, it's hand-on-gun-follow-me. But at land crossings, where they don't get an advance list of names it can be more exciting. I was crossing from Canada to New York once and, after being asked to pass my keys to the border guard, I looked around to see thirteen other guards rushing my car with drawn and aimed weapons. My passenger and I were handcuffed at gunpoint and dragged into separate isolation cells. I was left in handcuffs and further cuffed to the bench. An hour or so later an agent walked in and said "you're not black, are you?" After establishing that I didn't match the physical description in some very obvious ways, they fingerprinted me and let me go. The border guards were practically giddy, post stress.
"Prices quoted on the Elsevier website suggest that an academic library in the United States with a total student and faculty full-time equivalent number of around 10,000 would pay $2,211 for shared online access"
Yeah, not really sure why Elsevier wouldn't want to convert to open access. That's about what they charge per article at OA journals.
You'd need an awful lot of detectors because the plane would have to fly directly overhead. You could maybe do it the other way, with a satellite detecting a radar shadow from lots of emitters, but I would think it would be easier (nowhere near easy) to detect the actual shadow against background street lighting.
String theory, at least the core idea, makes predictions that are testable in principle. Even so, it's often criticized as being unscientific for not making predictions that are testable in the foreseeable future. Even so, there's the hope that, with further development, it might do so.
The general idea that consciousness causes wave function collapse cannot be tested. You simply cannot test whether a wave function collapsed or not without, at some point, being cognizant of the result. More restricted forms of the hypothesis ARE testable, have been tested, and have been determined not to be true.
I had a friend once who insisted that Reiki masters could bend light. I said "hey, I'd like to see that!" She replied that they can't bend visible light. Cool, let's set up IR, UV, whatever cameras then. No, they can only bend light when nobody is watching.... That's what the consciousness causes collapse theory is, and that's why it's rejected.
Many worlds is considered unscientific fiction by many physicists. I agree: it's a nice story, but until it makes some testable predictions, at least in principle, it's not science. This is a pragmatic viewpoint: a theory that does not make testable predictions has no predictive, and thus no practical value.
Rejecting many worlds and consciousness-causes-collapse isn't due to personal incredulity, it's due to both of those ideas being non-scientific. The scientists who proposed the consciousness-causes-collapse interpretation realized this and stopped supporting their own idea because of it. We're still waiting for the many worlds proponents to do the same.
I also think failing to dismiss unscientific theories gives the public the wrong idea, and harms science. As an example, the insistence of some physicists to take many worlds seriously seems to have given you the idea that the consciousness theory of wave function collapse was rejected based on personal belief.
It doesn't matter how many people are watching. You cannot make a measurement without interacting with the thing you're measuring. You need to bounce at least one test particle off your target to measure anything about it, and that interaction will change the thing you're trying to measure, as well as it's complementary property. You can optimize your measurement so that it has a minimal effect on position, but it will then have a big effect on momentum. If multiple people make different measurements at the same time you'll all just get crappy measurements.
The requirement for authorship on a paper is that you make a novel intellectual contribution. Collecting data in some standard way doesn't qualify, but coming up with a new way to do it does. In particle physics, many people get their names on papers because they've worked on building and designing equipment for that experiment. But that equipment is novel. In biology, the lab tech who runs a western blot doesn't (shouldn't) get her name on the paper. But the tech who invents an optimized western blot procedure for that experiment would.
We have? Penrose advanced that theory in the 80s but there's now a lot of evidence against it. It's quite difficult to imagine that something as big and messy as the brain could maintain superposition states.
It was an interesting thought experiment in the early days of quantum mechanics, to illustrate the idea of superposition. You can't take it literally though. Even if you could somehow get a vial of poison into superposition, the cat is far too big.
The general idea that consciousness causes wavefunction collapse is non-falsifiable and therefore not scientific. That's an excellent justification for dismissing it. If you propose a *specific*, falsifiable dependence on consciousness, such as "a human has to be watching in realtime" then those have already been disproven.
In quantum mechanics the likelihood of quantum effects diminishes as the particle interacts with larger systems of particles. One photon might leave a decent chance of tunneling while a million photons makes it extremely unlikely.
Recent research has tracked the popsci woo to it's origin. Apparently more often than not it's the university newspapers and press releases. Contrary to popular belief, university newspapers and press offices are generally staffed with non-scientists.
Whenever a pop science article talking about quantum mechanics says "observed" think "is in contact with a large number of other particles." You can't see anything without peppering it with particles, true, but quantum effects also become highly unlikely when a particle is in contact with a large particle system as well. That's why you don't come home to find that your desk chair has tunneled into the basement. Socks are an exception. They exist in a unique state of quantum grace.
Slashdot Mirror
You're missing something.
Imagine you're running along in a one dimensional world. Any positioning errors will move any two points either directly towards each other or away from each other. The average bias should be zero.
If you allow two dimensions, errors can now occur to the right and left of your path, even if you're still running along in a straight line. Any lateral error, right or left, will increase the measured distance between points. That's where the bias comes from.
If I followed your post correctly, the effect you're referring to is the same as the interpolation error mentioned in the article, and actually tends to decrease the measured distance.
Middle eastern muslims thought pretty much the same thing about the murderous christian barbarians who invaded their heartland a thousand years ago. Both the bible and koran, and most other holy books, have some incredibly nasty bits. The vast majority of adherents to any major religion ignore those parts.
Terrorist organizations are generally very small. Attacks have two goals: terrorize the victims and get publicity to recruit more terrorists. If you shut down the publicity then you deny both those objectives.
Covering up terrorist attacks probably isn't a good idea. But publicizing them as a small number of criminals who committed a horrible crime is a lot better than millions of people wringing their hands and screaming about how the terrorists are winning.
The quantum microtubule hypothesis has been pretty thoroughly disproven. Penrose doesn't think so anymore either.
It's both. It's a library for numerical computation using data flow graphs. Most people these days use such libraries to implement their machine learning... TensorFlow contains objects that do machine learning, on top of its lower level operations.
Tensors are used in differential geometry, but they're not limited to it. A tensor is just a multidimensional array. TensorFlow is built on a computation graph where tensors (arrays) flow between operations.
Yay! It's been a while since we had a nice armed standoff or riot. Of course, when the Mohawks are unhappy it can be difficult to get to the beach, so I hope they work it out by next summer.
Yup. I have a common name. Since 2001 I've gotten hassled when entering, travelling in, or flying near the US. Sometimes it's just extra "random" screening (five times while making a single connection in the US once). When crossing the border it's usually a forty minute interview in the back room.
The guy they're currently looking for scares the crap out of border guards. In airports, when they know you're coming, it's hand-on-gun-follow-me. But at land crossings, where they don't get an advance list of names it can be more exciting. I was crossing from Canada to New York once and, after being asked to pass my keys to the border guard, I looked around to see thirteen other guards rushing my car with drawn and aimed weapons. My passenger and I were handcuffed at gunpoint and dragged into separate isolation cells. I was left in handcuffs and further cuffed to the bench. An hour or so later an agent walked in and said "you're not black, are you?" After establishing that I didn't match the physical description in some very obvious ways, they fingerprinted me and let me go. The border guards were practically giddy, post stress.
"Prices quoted on the Elsevier website suggest that an academic library in the United States with a total student and faculty full-time equivalent number of around 10,000 would pay $2,211 for shared online access"
Yeah, not really sure why Elsevier wouldn't want to convert to open access. That's about what they charge per article at OA journals.
It's not for bombing Russia or China, it's for bombing some country with Russian or Chinese air defences as a proxy for bombing Russia or China.
You'd need an awful lot of detectors because the plane would have to fly directly overhead. You could maybe do it the other way, with a satellite detecting a radar shadow from lots of emitters, but I would think it would be easier (nowhere near easy) to detect the actual shadow against background street lighting.
String theory, at least the core idea, makes predictions that are testable in principle. Even so, it's often criticized as being unscientific for not making predictions that are testable in the foreseeable future. Even so, there's the hope that, with further development, it might do so.
The general idea that consciousness causes wave function collapse cannot be tested. You simply cannot test whether a wave function collapsed or not without, at some point, being cognizant of the result. More restricted forms of the hypothesis ARE testable, have been tested, and have been determined not to be true.
I had a friend once who insisted that Reiki masters could bend light. I said "hey, I'd like to see that!" She replied that they can't bend visible light. Cool, let's set up IR, UV, whatever cameras then. No, they can only bend light when nobody is watching.... That's what the consciousness causes collapse theory is, and that's why it's rejected.
There's no such thing as "proof." The Bullet Cluster is evidence in favour of dark matter, and, as I said, it's pretty strong evidence against MOND.
Many worlds is considered unscientific fiction by many physicists. I agree: it's a nice story, but until it makes some testable predictions, at least in principle, it's not science. This is a pragmatic viewpoint: a theory that does not make testable predictions has no predictive, and thus no practical value.
Rejecting many worlds and consciousness-causes-collapse isn't due to personal incredulity, it's due to both of those ideas being non-scientific. The scientists who proposed the consciousness-causes-collapse interpretation realized this and stopped supporting their own idea because of it. We're still waiting for the many worlds proponents to do the same.
I also think failing to dismiss unscientific theories gives the public the wrong idea, and harms science. As an example, the insistence of some physicists to take many worlds seriously seems to have given you the idea that the consciousness theory of wave function collapse was rejected based on personal belief.
It doesn't matter how many people are watching. You cannot make a measurement without interacting with the thing you're measuring. You need to bounce at least one test particle off your target to measure anything about it, and that interaction will change the thing you're trying to measure, as well as it's complementary property. You can optimize your measurement so that it has a minimal effect on position, but it will then have a big effect on momentum. If multiple people make different measurements at the same time you'll all just get crappy measurements.
The requirement for authorship on a paper is that you make a novel intellectual contribution. Collecting data in some standard way doesn't qualify, but coming up with a new way to do it does. In particle physics, many people get their names on papers because they've worked on building and designing equipment for that experiment. But that equipment is novel. In biology, the lab tech who runs a western blot doesn't (shouldn't) get her name on the paper. But the tech who invents an optimized western blot procedure for that experiment would.
Also, they don't stop moving. They stop tunneling.
We have? Penrose advanced that theory in the 80s but there's now a lot of evidence against it. It's quite difficult to imagine that something as big and messy as the brain could maintain superposition states.
It was an interesting thought experiment in the early days of quantum mechanics, to illustrate the idea of superposition. You can't take it literally though. Even if you could somehow get a vial of poison into superposition, the cat is far too big.
The general idea that consciousness causes wavefunction collapse is non-falsifiable and therefore not scientific. That's an excellent justification for dismissing it. If you propose a *specific*, falsifiable dependence on consciousness, such as "a human has to be watching in realtime" then those have already been disproven.
You do control the light that provides illumination at night.
In quantum mechanics the likelihood of quantum effects diminishes as the particle interacts with larger systems of particles. One photon might leave a decent chance of tunneling while a million photons makes it extremely unlikely.
Recent research has tracked the popsci woo to it's origin. Apparently more often than not it's the university newspapers and press releases. Contrary to popular belief, university newspapers and press offices are generally staffed with non-scientists.
Whenever a pop science article talking about quantum mechanics says "observed" think "is in contact with a large number of other particles." You can't see anything without peppering it with particles, true, but quantum effects also become highly unlikely when a particle is in contact with a large particle system as well. That's why you don't come home to find that your desk chair has tunneled into the basement. Socks are an exception. They exist in a unique state of quantum grace.
The title of their paper says it's a complex optimization algorithm.