You've put your finger on a potential big problem. As companies naturally try to reduce costs, they'll try to make the artificial meat with cheaper & cheaper ingredients. And, just like we saw with certain Chinese imported food, what's going to stop them from using ingredients that might turn out to be poisonous, so long as it helps them pass certain "quality" tests?
We'd like to think that we can count on the FDA to protect us, but, like others have pointed out, they'll probably be in the pockets of big agriculture business (ie, the current "regime"), and mostly just work to keep artificial meat off the market until other countries have proven it safe for decades. Then, once it's finally been let onto the US market, they'll do a poor job with inspection and quality control until people start dieing from the excessively-cheaply-made stuff.
Test Method:... Tests the luminance of televisions in both its default/home mode and the retail or brightest (aka torch) selectable mode....
Also, both technology and the desire to reduce manufacturing costs have been driving down TV power consumption already. It's cheaper to make smaller power supplies and have smaller heatsinks and heat shields, and if adding a 3M film to the panel increases light output 30%, you need that much less backlight power.
Still, it would probably be a better idea if the law just required TV makers/retailers to put a sticker on the front of each set indicating the power consumption in dollars and cents (units people understand). Indeed, they should do this for everything with a plug.
The "3Dmouse" mentioned above is not a dial. It is a puck that's spring-loaded to stay centered. You cannot rotate it freely, so it is a relative control and not an absolute control.
I kind of miss the Big Red Switch. A friend would sometimes say "Oh, that's a BRS error," when describing bugs for which the only solution involved said switch.
I'm more inclined to believe that Carlin was an optimist. My impression is that: - there's a few exceptionally bright people out there, - there's a fair number of reasonably understanding people, - and there's this mind-bogglingly huge number of incredibly stupid people.
At least, that's the impression I get from reading the news each day.
You are reading my comment from a US-centric point of view ("foreign" = "not US"). My comment is mostly correct from a global point of view ("foreign" = "not local"). If you are in Japan and buy a Japanese phone, for instance, you can most likely enter Japanese (local) characters, but not likely Arabic (foreign) ones. You can always enter Latin letters, though.
What you describe has many of the elements that I suggested, such as the paper record and multiple parties overseeing the count.
Many parts of the US use a similar system. Unfortunately, the method of doing voting is not something that is set at the national level in the US. Every state has its own laws & regulations, and every county makes its own choices within what the state allows.
The result is that there are thousands of bodies making the same mistakes over and over again, being taken advantage of by voting machine companies, and generally making a big mess of things.
I think the US federal government should make certain regulations regarding voting, such as requiring a verifiable, recountable paper record, and requiring multiple party oversight of the counting process. Hopefully, within those requirements, opportunity for screwups will be reduced significantly.
This would result in an immediate discrepancy. But I take your point, in that there would need to be very careful handling of each & every scan sheet.
The fact that something "easy" like voting is hard (when there is motivation to hack the system) should be a lesson to every lawmaker and programmer. Laws & programs are easy to make when you can trust people to do the right thing all the time. But, in the real world, you need to design them both as if people will try to punch holes in them any way they can, because they will.
Yes, and what prevents the machine from then throwing that away and using the secondary code from the hidden hardware?
You cannot trust any single piece of hardware. That's why I suggested the only way to gain trust is through consensus (multiple parties doing the counting and checking each other).
My thought exactly. In fact, there's no way to trust vendor-supplied hardware on this account, or any hardware of reasonable complexity at all.
I still think there's only one sensible way to do voting:
1. Let the voter fill in an optical scan form. 2. Let lots of different interested parties scan the form. 3. Verify that all parties have the same count after every form. 4. Lock the forms away in case a recount is needed.
If there's only one party doing the counting, they can never be trusted. Only by having every competing interest do the counting (with constant cross-checking) can a system be potentially trusted. Even then, you have to have enough parties involved to avoid the possibility of collusion.
Combine this with a system like Punchscan.org to add privacy, and maybe you've got something.
Ah, but I beg to differ again. It's true that we're arguing about the definition of "image", but you aren't quite correct either. Let me explain by way of example.
Let's start off with an array of 10x10 light bulbs laid out in a surface. This forms a "real image" in that it's a light-emitting surface, and it just happens to be that you can see it from anywhere (as long as you're in front of it) since the bulbs emit light diffusely in all (forward) directions.
Now, let's add a couple of layers of black construction paper in front of the bulbs, with pinholes placed in them such that the only rays of light that escape focus down to a point somewhere in front of the bulbs.
The light bulb grid still forms a real image, but you can only see it if you happen to place your pupil right at the point where the rays are "aimed" at (by way of the pinholes).
So now take away the light bulbs and replace them with very low power lasers (ones that are safe to look into). Aim all the lasers at the same point as before. Again, this will give you a 10x10 real image of dots, so long as you place your pupil at the right point.
So now take away the 100 lasers, and replace them with one laser, plus a big converging lens where the original surface was. Add a scanner such that the single beam can scan over the where the original 100 lasers originated from, and adjust the laser/scanner position such that the lens focuses the beam on the same point as before.
As far as you can tell, from the original surface, it's still the same "image" as with the light bulbs or the 100 lasers.
This is a "real" image in the optical sense, in that if you were to place a screen there, you'd see the same 10x10 grid of dots that you'd see if you placed your eye at the focus point. (That's what a "real" image is in the optical sense. A virtual image is one that cannot be formed by placing a screen at the focal plane.)
Like I suggested with the CRT example, it's not the same kind of "image" as a painting, which happens to continually reflect light in all directions over all its surface at the same time. It's scanned, so that there is only one point emitting at a time (like the CRT), and it's non-diffuse, such that there's only 1 focused ray coming off from each part of the lens surface. So it is an extremely cut down image that you can only see if your eye is in the right place (and you consider persistence of vision), but it's an image nonetheless.
If you take a picture of a CRT with a fast shutter setting, you'll see that there's never more than a few scan lines lit up at any one time. See: http://www.youtube.com/watch?v=lRidfW_l4vs
The "persistence of vision" effect allows ANY light source, whether a laser beam or a flashlight, to generate an image when it is scanned out appropriately. For one neat example, check out: http://www.ladyada.net/make/spokepov/
A CRT also has "no real equivalent to the image." It's just a scanned electron beam that lights up an area of phosphor for a small fraction of a second. Any image you see is just the result of light from multiple directions (the area of the image) entering your eyes within a small window of time.
Also, no practical HMDs use just a point light source scanned over the retina, since this results in a "eye box" that's much too small. The eye box is the volume of space that your pupil must be in in order to see the image. In order to be practical, a scanned point light source needs to pass through a beam spreader of some sort in order to enlarge the eye box. It's difficult to make an HMD with both a large eye box and a large FOV. (It's actually difficult to make an HMD with a big FOV period.)
As far as safety goes, there are typically several levels of fail-safes built into laser-scanned displays, including a "dead man's switch" which cuts power to the lasers if the scanner stops scanning. Hopefully, no manufacturer is stupid enough to depend upon software alone to control the laser power.
The problems you mention appear to have been solved already. There was a/. article a while back about Punchscan, a two-part ballot system, where the order of the choices for each selection is randomized, and a special keyed algorithm is necessary to know the choice order for any given ballot. The 2-part ballot has two layers, separating the choice description from the choice label; one part alone would not indicate how a choice was made. When the voter marked through a choice, it marked both layers. The voter could choose which layer to shred, while the other is counted and retained as a receipt. It has a system where the voter can go online to verify his ballot.
The real problem with voting is that you cannot see what happens once your ballot has been cast. Remember the quote attributed to Stalin: "Those who cast the votes decide nothing. Those who count the votes decide everything."
The only solution to this problem that I can think of is to provide a system that allows any interested party to count the votes. If only one party can count the votes, they are automatically suspect. If a limited number of parties can count the votes, then they can create problems that are not easily resolved. If anyone can count the votes, then most of those problems go away. How to achieve this, while still retaining anonymity, might not be easy (I can't say for certain, not having studied the problem deeply).
Slashdot Mirror
> One basic concern is that there is probably a reason why axon growth is suppressed...
Right - we need to be careful where this is applied, otherwise people might start thinking with other areas of their bodies. Oh wait...
You've put your finger on a potential big problem. As companies naturally try to reduce costs, they'll try to make the artificial meat with cheaper & cheaper ingredients. And, just like we saw with certain Chinese imported food, what's going to stop them from using ingredients that might turn out to be poisonous, so long as it helps them pass certain "quality" tests?
We'd like to think that we can count on the FDA to protect us, but, like others have pointed out, they'll probably be in the pockets of big agriculture business (ie, the current "regime"), and mostly just work to keep artificial meat off the market until other countries have proven it safe for decades. Then, once it's finally been let onto the US market, they'll do a poor job with inspection and quality control until people start dieing from the excessively-cheaply-made stuff.
I hope I'm wrong.
Instead, we'll get the disobedience vaccine and the willful thinking vaccine.
I just hope it doesn't result in the quick extinction of man-kind.
Indeed. Presentation on the law: http://www.energy.ca.gov/appliances/2009_tvregs/documents/2009-10-13_hearing/2009-10-13_STAFF_PRESENTATION.PDF
Test Method: ... ...
Tests the luminance of televisions in both its default/home mode and the retail or brightest (aka torch) selectable mode.
Also, both technology and the desire to reduce manufacturing costs have been driving down TV power consumption already. It's cheaper to make smaller power supplies and have smaller heatsinks and heat shields, and if adding a 3M film to the panel increases light output 30%, you need that much less backlight power.
Still, it would probably be a better idea if the law just required TV makers/retailers to put a sticker on the front of each set indicating the power consumption in dollars and cents (units people understand). Indeed, they should do this for everything with a plug.
You can get a bunch of these: http://www.griffintechnology.com/products/powermate/
These used to be common, long ago: http://en.wikipedia.org/wiki/Dial_box
The "3Dmouse" mentioned above is not a dial. It is a puck that's spring-loaded to stay centered.
You cannot rotate it freely, so it is a relative control and not an absolute control.
Linux has been ported to various non-x86 architectures. See: http://www.ibm.com/developerworks/linux/library/l-nonx86.html
And some folks have even ported a Linux subset to 8086. See: http://elks.sourceforge.net/FAQ-English.html
Still, it would be quite a small subset to fit on the 4004. Being so small, you could consider it a subset of virtually any OS.
> Unless human intelligence changes significantly...
Ah, so now we get to the meat of the matter!
I guess I'm wrong. They crammed 45 instructions into the architecture using instruction words of varying width.
I don't think you can get more RISC than the 4004's instruction set. Remember, it's a 4-bit CPU!
When we get the Core i7 details, will it seem as quaint as the 4004 does now?
I kind of miss the Big Red Switch. A friend would sometimes say "Oh, that's a BRS error," when describing bugs for which the only solution involved said switch.
I'm more inclined to believe that Carlin was an optimist.
My impression is that:
- there's a few exceptionally bright people out there,
- there's a fair number of reasonably understanding people,
- and there's this mind-bogglingly huge number of incredibly stupid people.
At least, that's the impression I get from reading the news each day.
You are reading my comment from a US-centric point of view ("foreign" = "not US"). My comment is mostly correct from a global point of view ("foreign" = "not local"). If you are in Japan and buy a Japanese phone, for instance, you can most likely enter Japanese (local) characters, but not likely Arabic (foreign) ones. You can always enter Latin letters, though.
Indeed. And most cell phones and video game consoles and other specialized devices have no way at all to enter foreign alphabets.
What you describe has many of the elements that I suggested, such as the paper record and multiple parties overseeing the count.
Many parts of the US use a similar system. Unfortunately, the method of doing voting is not something that is set at the national level in the US. Every state has its own laws & regulations, and every county makes its own choices within what the state allows.
The result is that there are thousands of bodies making the same mistakes over and over again, being taken advantage of by voting machine companies, and generally making a big mess of things.
I think the US federal government should make certain regulations regarding voting, such as requiring a verifiable, recountable paper record, and requiring multiple party oversight of the counting process. Hopefully, within those requirements, opportunity for screwups will be reduced significantly.
This would result in an immediate discrepancy. But I take your point, in that there would need to be very careful handling of each & every scan sheet.
The fact that something "easy" like voting is hard (when there is motivation to hack the system) should be a lesson to every lawmaker and programmer. Laws & programs are easy to make when you can trust people to do the right thing all the time. But, in the real world, you need to design them both as if people will try to punch holes in them any way they can, because they will.
Um, that's why I mentioned http://punchscan.org/ . Check it out.
Yes, and what prevents the machine from then throwing that away and using the secondary code from the hidden hardware?
You cannot trust any single piece of hardware. That's why I suggested the only way to gain trust is through consensus (multiple parties doing the counting and checking each other).
My thought exactly. In fact, there's no way to trust vendor-supplied hardware on this account, or any hardware of reasonable complexity at all.
I still think there's only one sensible way to do voting:
1. Let the voter fill in an optical scan form.
2. Let lots of different interested parties scan the form.
3. Verify that all parties have the same count after every form.
4. Lock the forms away in case a recount is needed.
If there's only one party doing the counting, they can never be trusted.
Only by having every competing interest do the counting (with constant cross-checking) can a system be potentially trusted.
Even then, you have to have enough parties involved to avoid the possibility of collusion.
Combine this with a system like Punchscan.org to add privacy, and maybe you've got something.
Ah, but I beg to differ again. It's true that we're arguing about the definition of "image", but you aren't quite correct either. Let me explain by way of example.
Let's start off with an array of 10x10 light bulbs laid out in a surface. This forms a "real image" in that it's a light-emitting surface, and it just happens to be that you can see it from anywhere (as long as you're in front of it) since the bulbs emit light diffusely in all (forward) directions.
Now, let's add a couple of layers of black construction paper in front of the bulbs, with pinholes placed in them such that the only rays of light that escape focus down to a point somewhere in front of the bulbs.
The light bulb grid still forms a real image, but you can only see it if you happen to place your pupil right at the point where the rays are "aimed" at (by way of the pinholes).
So now take away the light bulbs and replace them with very low power lasers (ones that are safe to look into). Aim all the lasers at the same point as before. Again, this will give you a 10x10 real image of dots, so long as you place your pupil at the right point.
So now take away the 100 lasers, and replace them with one laser, plus a big converging lens where the original surface was. Add a scanner such that the single beam can scan over the where the original 100 lasers originated from, and adjust the laser/scanner position such that the lens focuses the beam on the same point as before.
As far as you can tell, from the original surface, it's still the same "image" as with the light bulbs or the 100 lasers.
This is a "real" image in the optical sense, in that if you were to place a screen there, you'd see the same 10x10 grid of dots that you'd see if you placed your eye at the focus point. (That's what a "real" image is in the optical sense. A virtual image is one that cannot be formed by placing a screen at the focal plane.)
Like I suggested with the CRT example, it's not the same kind of "image" as a painting, which happens to continually reflect light in all directions over all its surface at the same time. It's scanned, so that there is only one point emitting at a time (like the CRT), and it's non-diffuse, such that there's only 1 focused ray coming off from each part of the lens surface. So it is an extremely cut down image that you can only see if your eye is in the right place (and you consider persistence of vision), but it's an image nonetheless.
No, I took your point perfectly well.
If you take a picture of a CRT with a fast shutter setting, you'll see that there's never more than a few scan lines lit up at any one time. See: http://www.youtube.com/watch?v=lRidfW_l4vs
The "persistence of vision" effect allows ANY light source, whether a laser beam or a flashlight, to generate an image when it is scanned out appropriately. For one neat example, check out: http://www.ladyada.net/make/spokepov/
A CRT also has "no real equivalent to the image." It's just a scanned electron beam that lights up an area of phosphor for a small fraction of a second. Any image you see is just the result of light from multiple directions (the area of the image) entering your eyes within a small window of time.
Also, no practical HMDs use just a point light source scanned over the retina, since this results in a "eye box" that's much too small. The eye box is the volume of space that your pupil must be in in order to see the image. In order to be practical, a scanned point light source needs to pass through a beam spreader of some sort in order to enlarge the eye box. It's difficult to make an HMD with both a large eye box and a large FOV. (It's actually difficult to make an HMD with a big FOV period.)
As far as safety goes, there are typically several levels of fail-safes built into laser-scanned displays, including a "dead man's switch" which cuts power to the lasers if the scanner stops scanning. Hopefully, no manufacturer is stupid enough to depend upon software alone to control the laser power.
The problems you mention appear to have been solved already. There was a /. article a while back about Punchscan, a two-part ballot system, where the order of the choices for each selection is randomized, and a special keyed algorithm is necessary to know the choice order for any given ballot. The 2-part ballot has two layers, separating the choice description from the choice label; one part alone would not indicate how a choice was made. When the voter marked through a choice, it marked both layers. The voter could choose which layer to shred, while the other is counted and retained as a receipt. It has a system where the voter can go online to verify his ballot.
The real problem with voting is that you cannot see what happens once your ballot has been cast. Remember the quote attributed to Stalin: "Those who cast the votes decide nothing. Those who count the votes decide everything."
The only solution to this problem that I can think of is to provide a system that allows any interested party to count the votes. If only one party can count the votes, they are automatically suspect. If a limited number of parties can count the votes, then they can create problems that are not easily resolved. If anyone can count the votes, then most of those problems go away. How to achieve this, while still retaining anonymity, might not be easy (I can't say for certain, not having studied the problem deeply).
If you try to leave, you'll find yourself on an island, subject to the whims of Number Two.