That was what XUL and XPCOM were for. XULRunner is still a very respectable development environment. (The Firefox and Thunderbird UIs code is written almost exclusively in XUL/JavaScript.) Unfortunately, while these technologies have been around for quite a while, they haven't really taken off beyond Mozilla's apps. Up until recently, it was still possible to load XUL from a remote site and get an interface with native widgets, but no longer.
Graphics card doesn't make a difference because it presumably doesn't use WebGL. Browser makes a huge difference because Firefox 4+ has typed arrays, whereas Firefox 3.5 doesn't.
Wrong. Invertebrate model organisms are how most discoveries about the mammalian brain started off and continue to be how we discover the basics. On the most obvious level, THIS IS A NEURON. Same type of cell your brain is made up of.
Most of the important early discoveries in the invertebrate nervous system that were shown to hold in the mammalian brain had to do with the dynamics of individual neurons, not systems. The important discovery here is at the system level, not the single neuron level. While quite a few people would disagree, I couldn't care less about the dynamics of the insect olfactory bulb if they do not translate to the primate brain.
As far as one individual meganeuron in your head, maybe not. I think the histologists of the past would have realized if there were giant neurons similar to this. In the 1800's, they were using advanced staining techniques to show the shape of cells, I think if one neuron were synapsing with that many neurons, it would have shown up with golgi staining back then, or with the brainbow mouse [scienceblogs.com] more recently.
The average neuron in cerebral cortex makes on the order of 10,000 synapses. I would not be surprised if you can find neurons that make more than 50,000, particularly in small regions with a large number of long-range projections (e.g., VTA and striatum). I would be surprised if anything goes wrong if you kill a single one, which was my point. I'd also be very surprised if they don't fire action potentials.
FWIW, tracing the full extent of the synaptic connections of a single neuron in a mammalian brain is hard with any of these techniques, because you are typically labeling a large number of cells and trying to trace them through a large number of slices. Ed Callaway has a solution for labeling all presynaptic neurons targeting a single neuron, but I'm not aware of any good solutions for targeting postsynaptic cells.
The concept of bottlenecking information when sparsity is necessary: that probably IS a valuable lesson for human brains. It probably isn't a single cell, but the concept is still possible with a smaller number of cells.
This is possible, but I would say "maybe" and not "probably." There are computational principles that seem to apply in some brains but not others. (For example, intracellular recordings suggest that sequential firing during singing in songbird HVC is probably generated by synfire chains whereas sequential firing in the hippocampus during navigation likely has a more complex basis.) It's possible this principle holds in the mammalian olfactory bulb, although I would think a relatively large population of neurons would be involved. But, I would be surprised if it holds in neocortex. Most neocortical neurons are not all that sparse in comparison to neurons in archicortex, and connectivity patterns in neocortex are vastly different. The point is, we really don't know how sparseness is achieved in our own brains, and this article doesn't really add much on its own, although it suggests a path for further investigation. Since the article summary conveniently ignores the fact that this work was performed in an insect model, it makes it appear as if this strategy is used throughout the human brain, when this is very far from established.
For many reasons, there is almost certainly nothing like this in the mammalian brain.
Re:This is not the logic you are looking for
on
Is Sugar Toxic?
·
· Score: 2
If the article is correct, then fructose and sucrose are responsible for a large proportion of deaths in the United States, and not merely because of the calories they contain. (As far as I can tell, glucose and complex carbohydrates are fine.) If this is indeed the case, I think "toxic" is an accurate term. The headline is a little sensational in that it says "sugar" and not "fructose and sucrose," but no one has "fallen for a more subtle form of the Dihydrogen Monoxide troll."
Fortunately, using your GPU to render webpages means less load on your CPU, which means your CPU can run at a lower power mode. Also, if you render your webpage in half the time, both your GPU and CPU can go to idle faster. It's not so simple. In fact, if you look at this blog post, the fully GPU-accelerated browsers (Fx 4 and IE 9) are actually using significantly less power...
Even if you watch a movie without 3D, you are "tricking the brain into making you think you are physically moving in relation to your surroundings." There is a large overlap in the neural circuitry that processes motion parallax (the 3D effect that you get when you have a moving camera) and stereopsis (the 3D effect that you get when you have two different images projected onto your two retinas). This is the mechanism behind 3D animated GIFs, and one of the major depth cues in a 2D movie. Motion parallax is even more intricately linked to the vestibular system, since you need to know whether the image on your retina is changing because your head is moving or because the object you are looking at is in motion. (This is probably part of the reason that an ordinary movie is not an immersive 3D experience.) In contrast, stereopsis does not require motion to work as a depth cue, although all of these depth cues are ultimately integrated.
The potential for motion parallax without vestibular signals to alter the development of visual areas dedicated to depth perception seems at least as great as the potential for moving stereoscopic images without vestibular signals to alter the development of these areas. No one knew about this when the motion picture was invented, and kids who grew up with a TV are still perfectly capable of making use of vestibular signals.
Overall, that 3D is somehow "bad for the brain" is highly speculative. You don't get a headache or nausea when viewing 3D movies from very close up because you are damaging your brain. The malaise doesn't even necessarily have to do with the lack of a vestibular signal, and quite possibly doesn't, since you don't get nausea from simulated camera movement without associated head movement even though you have conflicting cues there as well. It can come from the visual system alone. If you are close enough to the screen, you are viewing 3D images with such high disparity that you can't fuse them. The brain interprets this as a sign that there is a problem with your visual system. You might even feel sick to your stomach, since in the environment in which we evolved, this kind of problem with your visual system would most likely have been caused by ingesting some kind of harmful psychotropic substance. There is absolutely no evidence that there is any permanent damage to or alteration of the brain itself.
If someone can show that there is any change in cortical thickness in the visual areas of children exposed to 3D movies from a very young age, or that these children exhibit significantly different performance in some set of psychophysical paradigms, I might reconsider, but the "evidence" presented in this article is complete bullshit.
You would need to stimulate the vestibular system, which means you'd need electrodes, not earphones. Even if you were to put electrodes into your central nervous system for the sake of a better movie watching experience, you would be moving your eyes to compensate, so you'd also need to inject modified rabies virus into the nucleus abducens and use arch or halorhodopsin and implant an optrode, so that you can reversibly inactivate the vestibulo-ocular reflex while you watch movies. This would be interesting, to say the least.
A Bluetooth interface is almost free. I don't think it's reasonable to expect a part that costs less than $5 to be optional. That it is in PC laptops is, IMHO, a scam.
The material cost of the lighted keyboard is probably also pretty low, although there is undoubtedly a design cost.
What mathematical constant? You can encode many things as numbers, including stories, songs, movies, and the CSS and AACS encryption keys, but these are not mathematical constants.
So if they sold the domain but kept the rights to the Sun name as a trademark, then how could anyone open up a new Sun.com without being in danger of violating Oracle's trademark? People have been sued over their domains named after themselves when it has the same name as a trademark, even when their domain has nothing to do with whatever area the trademark is in.
If the company operates in an entirely different area from Sun Microsystems, such that no confusion could arise, or if Oracle abandons the Sun trademark, which they probably would be if they sold Sun.com, then there's no problem.
WRT lawsuits, the CIA has also been sued for mind control. You can sue for anything; winning is the hard part.
No. Google's JIT is just as insecure. The problem is not in the implementation, it's that you need to disable the NX bit on an area of memory to run a JIT at all. There is no workaround to this, unless the JIT isn't actually a just in time compiler.
Wow. That doesn't sound safe. First thought was guess the 5 sec Apple pown to ownage will be even quicker with this?
Seriously, anybody have any comments on the safety of this / what prevents a small bug in Safari from easily becoming arbitrary code execution?
Or even the possibility that some special javascript doesn't even need a Safar bug, just this feature to do evil stuff on the device?
JavaScript gets compiled to ARM instructions, and those ARM instructions get executed. You can't just put ARM instructions in a webpage and expect it to run. Barring a bug in the Nitro JIT, it's perfectly secure. Any JIT (including V8, JaegerMonkey/TraceMonkey, and the JVM), on any platform, necessarily has the same potential vulnerabilities. If the JIT is written properly and doesn't attempt to execute code in the parts of RAM that contain data from untrusted sources, it's not a problem. Most desktop platforms are much worse because they allow execution of code anywhere in RAM, in any application, regardless of its source. (Ever heard of a buffer overflow?)
The problem is not using ARM instructions. The problem is where those ARM instructions are. The iPhone presumably uses something like the NX bit to segregate data from code. Because of the way a JIT works, it needs to be able to execute code in the data area of memory. Allowing every app to do this would effectively eliminate the additional security that the NX bit provides.
The technology is 100+ years old and has been used for 80 on human brain waves.
Almost 20 years ago, work at Radford was able to guess with 70 to 80 percent accuracy which of three possibilities within three parameters (size, shape and color) was being looked at, or being imagined with and without there being an attempt to verbalize it. They used a standard 16 channel external EEG. And a dozen different subjects.
I think the point here is that they are NOT using an EEG. They are using what are called "Utah arrays," which are a relatively new technological advance even for monkey electrophysiologists, and this is one of the first times they have ever been implanted in a human brain. This is a new technology, if not a new application.
Also, being able to distinguish size, shape, and color is not really a substitute for being able to communicate. People have had electrodes implanted in motor cortex to substitute for missing arms, etc. and that worked pretty well, but this is the first time anyone has (intentionally) implanted an electrode array in human speech areas to my knowledge.
You've got to have a damn good reason to carve open a skull. Surgical correction for epilepsy is a good reason, but the brain being tested before and after the surgery is hardly one to draw generalizations from. Given that previous work bested this without cutting into anyone, this is a dead end stunt.
The figures reported in the article are accuracy for selecting one of 10 words, where chance would be 10%. The figures you report are accuracy for 3 tasks, where chance would be 33%.
TFA is some scary shit. With all the alternatives available, safer, better AND cheaper, there's no reason to do stuff like this, and none at all to suggest that it should be used as a basis to develop a technology.
If you worked with a surface EEG, and then did single neuron electrophysiology, you would realize pretty quickly what the reason is. There is a pretty low ceiling for accuracy of any single-trial EEG approach because of the noise in the technique. It turns out that if you put an electrode directly into the brain, for various reasons, your measurements are much less noisy. Even if the paper in question has not succeeded in doing the decoding in a way that actually makes this useful, it is theoretically possible to make this work, whereas with EEG, you would still be limited to choosing from a very small number of alternatives.
And if you save them, and dump the sound track with mplayer, you usually end up with a nice 128kbps MP3.
Well, you end up with a 128 kbps MP3, but it's not usually too nice because YouTube compresses the hell out of audio. Depending on the source material, parameters used, and your ears, you may or may not care, but it's hard not to notice. I am pretty sure that YouTube does this for a reason.
There are some pretty benchmarks on the Mozilla site that show what TraceMonkey is faster at and what it's not. What you have here is slow because TraceMonkey doesn't optimize recursion. This feature is scheduled to be implemented in Firefox 3.1b2, so the final version of 3.1 may indeed perform this benchmark faster than V8.
But it is not free (as in libre) software. The APSL is a non-free license, albeit it does meet the OSI definition and hence is Open Source, but not Free Software.
Err, you can do this in Mac OS X with Spotlight relatively easily. I don't know where you got the idea that this would be impossible to do with a GUI, but the fact that I _can_ do it, and (for me) it is much easier than using the command line, serves as a counterexample.
Iriki, Tanaka, and Iwamura published a paper in 1996 showing that, when macaque monkeys were given a rake with which to retrieve distant objects, the receptive fields of neurons coding for the body schema of the hand changed to encompass the rake. I haven't read the Rizzolatti paper yet, and I'm sure that it adds something new to the research, but this basic idea was around more than 10 years ago.
Slashdot Mirror
That was what XUL and XPCOM were for. XULRunner is still a very respectable development environment. (The Firefox and Thunderbird UIs code is written almost exclusively in XUL/JavaScript.) Unfortunately, while these technologies have been around for quite a while, they haven't really taken off beyond Mozilla's apps. Up until recently, it was still possible to load XUL from a remote site and get an interface with native widgets, but no longer.
Graphics card doesn't make a difference because it presumably doesn't use WebGL. Browser makes a huge difference because Firefox 4+ has typed arrays, whereas Firefox 3.5 doesn't.
Maybe the link has the same text as the summary. Maybe we have no idea how this board is supposed to "speed up Linux development."
Wrong. Invertebrate model organisms are how most discoveries about the mammalian brain started off and continue to be how we discover the basics. On the most obvious level, THIS IS A NEURON. Same type of cell your brain is made up of.
Most of the important early discoveries in the invertebrate nervous system that were shown to hold in the mammalian brain had to do with the dynamics of individual neurons, not systems. The important discovery here is at the system level, not the single neuron level. While quite a few people would disagree, I couldn't care less about the dynamics of the insect olfactory bulb if they do not translate to the primate brain.
As far as one individual meganeuron in your head, maybe not. I think the histologists of the past would have realized if there were giant neurons similar to this. In the 1800's, they were using advanced staining techniques to show the shape of cells, I think if one neuron were synapsing with that many neurons, it would have shown up with golgi staining back then, or with the brainbow mouse [scienceblogs.com] more recently.
The average neuron in cerebral cortex makes on the order of 10,000 synapses. I would not be surprised if you can find neurons that make more than 50,000, particularly in small regions with a large number of long-range projections (e.g., VTA and striatum). I would be surprised if anything goes wrong if you kill a single one, which was my point. I'd also be very surprised if they don't fire action potentials.
FWIW, tracing the full extent of the synaptic connections of a single neuron in a mammalian brain is hard with any of these techniques, because you are typically labeling a large number of cells and trying to trace them through a large number of slices. Ed Callaway has a solution for labeling all presynaptic neurons targeting a single neuron, but I'm not aware of any good solutions for targeting postsynaptic cells.
The concept of bottlenecking information when sparsity is necessary: that probably IS a valuable lesson for human brains. It probably isn't a single cell, but the concept is still possible with a smaller number of cells.
This is possible, but I would say "maybe" and not "probably." There are computational principles that seem to apply in some brains but not others. (For example, intracellular recordings suggest that sequential firing during singing in songbird HVC is probably generated by synfire chains whereas sequential firing in the hippocampus during navigation likely has a more complex basis.) It's possible this principle holds in the mammalian olfactory bulb, although I would think a relatively large population of neurons would be involved. But, I would be surprised if it holds in neocortex. Most neocortical neurons are not all that sparse in comparison to neurons in archicortex, and connectivity patterns in neocortex are vastly different. The point is, we really don't know how sparseness is achieved in our own brains, and this article doesn't really add much on its own, although it suggests a path for further investigation. Since the article summary conveniently ignores the fact that this work was performed in an insect model, it makes it appear as if this strategy is used throughout the human brain, when this is very far from established.
But the comment still stands.
For many reasons, there is almost certainly nothing like this in the mammalian brain.
If the article is correct, then fructose and sucrose are responsible for a large proportion of deaths in the United States, and not merely because of the calories they contain. (As far as I can tell, glucose and complex carbohydrates are fine.) If this is indeed the case, I think "toxic" is an accurate term. The headline is a little sensational in that it says "sugar" and not "fructose and sucrose," but no one has "fallen for a more subtle form of the Dihydrogen Monoxide troll."
Fortunately, using your GPU to render webpages means less load on your CPU, which means your CPU can run at a lower power mode. Also, if you render your webpage in half the time, both your GPU and CPU can go to idle faster. It's not so simple. In fact, if you look at this blog post, the fully GPU-accelerated browsers (Fx 4 and IE 9) are actually using significantly less power...
Even if you watch a movie without 3D, you are "tricking the brain into making you think you are physically moving in relation to your surroundings." There is a large overlap in the neural circuitry that processes motion parallax (the 3D effect that you get when you have a moving camera) and stereopsis (the 3D effect that you get when you have two different images projected onto your two retinas). This is the mechanism behind 3D animated GIFs, and one of the major depth cues in a 2D movie. Motion parallax is even more intricately linked to the vestibular system, since you need to know whether the image on your retina is changing because your head is moving or because the object you are looking at is in motion. (This is probably part of the reason that an ordinary movie is not an immersive 3D experience.) In contrast, stereopsis does not require motion to work as a depth cue, although all of these depth cues are ultimately integrated.
The potential for motion parallax without vestibular signals to alter the development of visual areas dedicated to depth perception seems at least as great as the potential for moving stereoscopic images without vestibular signals to alter the development of these areas. No one knew about this when the motion picture was invented, and kids who grew up with a TV are still perfectly capable of making use of vestibular signals.
Overall, that 3D is somehow "bad for the brain" is highly speculative. You don't get a headache or nausea when viewing 3D movies from very close up because you are damaging your brain. The malaise doesn't even necessarily have to do with the lack of a vestibular signal, and quite possibly doesn't, since you don't get nausea from simulated camera movement without associated head movement even though you have conflicting cues there as well. It can come from the visual system alone. If you are close enough to the screen, you are viewing 3D images with such high disparity that you can't fuse them. The brain interprets this as a sign that there is a problem with your visual system. You might even feel sick to your stomach, since in the environment in which we evolved, this kind of problem with your visual system would most likely have been caused by ingesting some kind of harmful psychotropic substance. There is absolutely no evidence that there is any permanent damage to or alteration of the brain itself.
If someone can show that there is any change in cortical thickness in the visual areas of children exposed to 3D movies from a very young age, or that these children exhibit significantly different performance in some set of psychophysical paradigms, I might reconsider, but the "evidence" presented in this article is complete bullshit.
You would need to stimulate the vestibular system, which means you'd need electrodes, not earphones. Even if you were to put electrodes into your central nervous system for the sake of a better movie watching experience, you would be moving your eyes to compensate, so you'd also need to inject modified rabies virus into the nucleus abducens and use arch or halorhodopsin and implant an optrode, so that you can reversibly inactivate the vestibulo-ocular reflex while you watch movies. This would be interesting, to say the least.
A Bluetooth interface is almost free. I don't think it's reasonable to expect a part that costs less than $5 to be optional. That it is in PC laptops is, IMHO, a scam.
The material cost of the lighted keyboard is probably also pretty low, although there is undoubtedly a design cost.
What mathematical constant? You can encode many things as numbers, including stories, songs, movies, and the CSS and AACS encryption keys, but these are not mathematical constants.
ThumbEE has replaced Jazelle, so the latest ARM CPUs can in theory accelerate Dalvik as well as they could accelerate Java.
So if they sold the domain but kept the rights to the Sun name as a trademark, then how could anyone open up a new Sun.com without being in danger of violating Oracle's trademark? People have been sued over their domains named after themselves when it has the same name as a trademark, even when their domain has nothing to do with whatever area the trademark is in.
If the company operates in an entirely different area from Sun Microsystems, such that no confusion could arise, or if Oracle abandons the Sun trademark, which they probably would be if they sold Sun.com, then there's no problem.
WRT lawsuits, the CIA has also been sued for mind control. You can sue for anything; winning is the hard part.
No. Google's JIT is just as insecure. The problem is not in the implementation, it's that you need to disable the NX bit on an area of memory to run a JIT at all. There is no workaround to this, unless the JIT isn't actually a just in time compiler.
Wow. That doesn't sound safe. First thought was guess the 5 sec Apple pown to ownage will be even quicker with this?
Seriously, anybody have any comments on the safety of this / what prevents a small bug in Safari from easily becoming arbitrary code execution?
Or even the possibility that some special javascript doesn't even need a Safar bug, just this feature to do evil stuff on the device?
JavaScript gets compiled to ARM instructions, and those ARM instructions get executed. You can't just put ARM instructions in a webpage and expect it to run. Barring a bug in the Nitro JIT, it's perfectly secure. Any JIT (including V8, JaegerMonkey/TraceMonkey, and the JVM), on any platform, necessarily has the same potential vulnerabilities. If the JIT is written properly and doesn't attempt to execute code in the parts of RAM that contain data from untrusted sources, it's not a problem. Most desktop platforms are much worse because they allow execution of code anywhere in RAM, in any application, regardless of its source. (Ever heard of a buffer overflow?)
The problem is not using ARM instructions. The problem is where those ARM instructions are. The iPhone presumably uses something like the NX bit to segregate data from code. Because of the way a JIT works, it needs to be able to execute code in the data area of memory. Allowing every app to do this would effectively eliminate the additional security that the NX bit provides.
Phrases like "ends of the Earth," if taken literally, imply that the Earth is not only flat, but also one-dimensional.
The technology is 100+ years old and has been used for 80 on human brain waves.
Almost 20 years ago, work at Radford was able to guess with 70 to 80 percent accuracy which of three possibilities within three parameters (size, shape and color) was being looked at, or being imagined with and without there being an attempt to verbalize it. They used a standard 16 channel external EEG. And a dozen different subjects.
I think the point here is that they are NOT using an EEG. They are using what are called "Utah arrays," which are a relatively new technological advance even for monkey electrophysiologists, and this is one of the first times they have ever been implanted in a human brain. This is a new technology, if not a new application.
Also, being able to distinguish size, shape, and color is not really a substitute for being able to communicate. People have had electrodes implanted in motor cortex to substitute for missing arms, etc. and that worked pretty well, but this is the first time anyone has (intentionally) implanted an electrode array in human speech areas to my knowledge.
You've got to have a damn good reason to carve open a skull. Surgical correction for epilepsy is a good reason, but the brain being tested before and after the surgery is hardly one to draw generalizations from. Given that previous work bested this without cutting into anyone, this is a dead end stunt.
The figures reported in the article are accuracy for selecting one of 10 words, where chance would be 10%. The figures you report are accuracy for 3 tasks, where chance would be 33%.
TFA is some scary shit. With all the alternatives available, safer, better AND cheaper, there's no reason to do stuff like this, and none at all to suggest that it should be used as a basis to develop a technology.
If you worked with a surface EEG, and then did single neuron electrophysiology, you would realize pretty quickly what the reason is. There is a pretty low ceiling for accuracy of any single-trial EEG approach because of the noise in the technique. It turns out that if you put an electrode directly into the brain, for various reasons, your measurements are much less noisy. Even if the paper in question has not succeeded in doing the decoding in a way that actually makes this useful, it is theoretically possible to make this work, whereas with EEG, you would still be limited to choosing from a very small number of alternatives.
And if you save them, and dump the sound track with mplayer, you usually end up with a nice 128kbps MP3.
Well, you end up with a 128 kbps MP3, but it's not usually too nice because YouTube compresses the hell out of audio. Depending on the source material, parameters used, and your ears, you may or may not care, but it's hard not to notice. I am pretty sure that YouTube does this for a reason.
You mean like Nature, as the second link states?
There are some pretty benchmarks on the Mozilla site that show what TraceMonkey is faster at and what it's not. What you have here is slow because TraceMonkey doesn't optimize recursion. This feature is scheduled to be implemented in Firefox 3.1b2, so the final version of 3.1 may indeed perform this benchmark faster than V8.
But it is not free (as in libre) software. The APSL is a non-free license, albeit it does meet the OSI definition and hence is Open Source, but not Free Software.
Really? Even RMS seems to disagree.
"The Apple Public Source License (APSL) version 2.0 qualifies as a free software license."
Err, you can do this in Mac OS X with Spotlight relatively easily. I don't know where you got the idea that this would be impossible to do with a GUI, but the fact that I _can_ do it, and (for me) it is much easier than using the command line, serves as a counterexample.
Iriki, Tanaka, and Iwamura published a paper in 1996 showing that, when macaque monkeys were given a rake with which to retrieve distant objects, the receptive fields of neurons coding for the body schema of the hand changed to encompass the rake. I haven't read the Rizzolatti paper yet, and I'm sure that it adds something new to the research, but this basic idea was around more than 10 years ago.