Is it possible that the solution is that Netflix basically are forced to have multiple ISPs and connect directly to many networks?
Yes. The article indirectly confirms it:
Netflix has historically routed its streaming content to broadband providers through a number of Internet middlemen
Netflix probably has servers all over the world, so they would have many many ISPs. Those ISPs may have peering agreements, but Netflix still pays either way.
This does not sound like a case of Comcast "double dipping" where Comcast gets paid by the customer and by Netflix for bandwidth. And the article makes no implication that Comcast was going to do something non-neutral like slow down Netflix traffic. This looks more like Netflix is co-locating servers inside Comcast's data centers to avoid paying big bandwidth bills to tier 1 network providers.
Upon first reading this article I jumped to the conclusion that Comcast was slowing Netflix traffic and Netflix was paying them for some kind of "priority" service to get their bandwidth back up. But I don't think that is the case.
This is still a dangerous idea. What happens if Netflix decides to just stop supporting Internet streaming, and instead just streams to ISPs who they partner with? Then you will only get Netflix if you subscribe to an ISP that supports them. Sound familiar? It's basicall the Cable TV networks all over again.
My employer uses a MITM HTTPS proxy. The IT department pushed down a trusted corporate certificate, and most people don't even know their HTTPS connections aren't secure any more. The real problem is when some application, other than a browser, needs internet access and it fails. This includ sethings like web installers that download the app during installation, automatic update systems, secure file transfer software, or things that call home to confirm a license key. On occassion a developer curses some installer for not working, then we inspect the install.log file and find something about a certificate failure.
IT departments forget that HTTPS is used for more than just browsing the web.
Some of the competitors products do. I forget which one(s). There was a neat youtube video of some company's robot and it would decide when it was done a room and not go back into it, just by mapping out where the doors were as it went along.
Most artificial muscles work by applying electric current along the muscle. When the current is removed they snap back to their original shape. Using heat sounds very limiting. Presumably you cool it to get it back to the original shape, but the ABC article is light on details.
I've been wanting something like this since I was a kid. Even if for no reason other than to feel totally cool. (Yes, everyone else will think I'm a dork for wearing it, but that's the story of my life. I wear by 8-bit tie proudly!)
In 1960, a phone call could be placed from any point in the United States that had a 10 lb telephone hard-wired to it to any other point in the United States that had a 10 lb telephone hard-wired to it and the sound quality would be consistently good.
Of course the whole thing is just mental masturbation nonsense
If you wish to boil philosophy down to nonsense, so be it. This is somewhere between philosophy, physics, and mathematics. Maybe it is only theoretical right now. Perhaps, if we start thinking of the universe this way, it will yield testable predictions.
The implicit limit in binary representation isn't that you get something random if you try to slice between two values, you get either one or the other.
Not really. Check out the articles on numerical stability that I linked to and look at zwei2stein's post and my reply. Suppose for a moment they are using floating-point. As you approach the limits of numerical stability on floating-point, you get crazy results and aliasing. I proposed that randomness was added intentionally to avoid aliasing problems.
When that particular comment was made, the ubiquity of the home router dolling out DHCP addresses probably wasn't considered.
The report was in Feb 2011. Home routers have been doing this for 10 years. The reality is that home routers doing this is actually part of the problem. It's a real hassle for game developers, file sharers, Tor users, media servers,...
I didn't mean "we" as in "intelligent life" I meant "we" as in "any output of that wave function." I clarified this in another reply earlier on in the discussion. Sorry.
I agree. That is one of the holes in my pseudoscience.
I thought about it more and remembered my reasoning on this. I propose that the creator of the simulation intentionally randomized these small quantities to eliminate the aliasing. It is effectively an unpredictable version of symmetric rounding.
If they didn't do this, the simulation might gain or lose energy over time because of these small errors. Or you might wind-up with something like values tending to be positive slighty more often than they are negative (or vice-versa) which could imbalance some things. (Maybe they didn't get it totally right, and that explains CP-violation and/or matter antimatter asymmetry?)
I agree. That is one of the holes in my pseudoscience.
...not because of some weird universe limit, but because you effect particles by measuring them...
I would be fine with that if it weren't for the EPR paradox. Yes, we found a mathematical way to describe the result. It looks... suspicious to me. It seems like Einstein and friends found a way around the system, to read the *real* value. Then we tried it and noticed that value isn't actually there! It's like a splinter in my mind.
And actually computing whether something will be observed sounds like too much work to be considered a optimalization.
I can interpret that sentence two ways. By "too much work" do you mean "it would take more calculations to determine if something is observed than to simply calculate the result" or do you mean "the designer of the simulation would not bother to implement such an optimization." If you mean the second, I can't argue their intent so I don't know. If you mean the first, then I disagree. There are lots of cases of optimizations in simulators that work just like that. Simulations often calculate if something is off-screen or irrelevant before rendering it.
No optimization for intelligence is required. Wave don't merely collapse when something *intelligent* is observing it. They collapse when *anything* is observing it. Any time the output of the function is hooked to the input of another. If a snail opened Schrödinger's box, or if a stray rock struck the latch, the inside state would still collapse. It doesn't wait until a person walks over to it and looks.
I saw another post equating physics and mathematics. That is a new concept to me and not one I intended to imply. I postulate that that physics is *based on* mathematics. Curiously, we notice that the limitations of physics are similar to the limitations of numerical stability. That does not imply that physics is mathematics. It does not imply that mathematics has limitations. It implies that physics has limitations, which implies that it is not real.
Try it. I bet it won't work past a certain point.:-)
Just because you speculate that you could think of a big number or a small number isn't enough. It's fine, until we try to actually measure something that small or that big, and suddenly we find it doesn't work. I imagine you as a video game character who thought he could get infinite lives if he played really well, then one day notices his score is -2147483648 or number of lives is -127. But you have to play perfectly to ever even figure that out.
Yay someone who gets it! But you don't have to have the municipality own the infrastructure. You could still have a monopoly do it, so long as they only ran the lines and didn't act as an ISP. Permit me to expand on what you said, for the benefit of anyone stumbling upon this.
Those contracts are a result of physical limitations that create an almost natural monopoly. (You can't run dozens of cable lines along the poles, it would be a mess)...
This was true in the 1960s and 1970s when Comcast was first starting. But it is no longer true today since you can run multiple services over a single wire. This is how power is today, and how internet used to be in the 1990s. Everyone forgets that this limitation no longer applies, and so our regulatory framework is stuck in the past.
...Now, if towns owned the cable infrastructure and other network devices involved, and just leased to Comcast / Whoever to act as an ISP...
With power, one company builds and maintains the wires. Many companies provide the actual power, and customers can choose their power provider. This is really easy with power since power is fungible. The power companies just put the power on the wires and off it goes.
With telephone in the 1990s, one company built and maintained the wires. They also provided local dial-up or PPTP access. Many companies provided internet service, and customers could choose their ISP. The telephone wire company merely provided the means to connect the customer to the service.
With broadband internet, it could be the same way and it would work exactly like how telephone used to work. Telephone company provides PPTP, and you get a box that does PPTP to your ISP who could be anyone anywhere.
What happened that messed this up was that the cable and telephone monopolies were permitted to become ISPs. They had special accses tot he system, so the 3rd-party ISPs couldn't compete. For a while, in some states, there was legislation that actually required the telephone companies to allow these other ISPs to lease access to the lines at certain fixed prices. This was a poor attempt to allow competition. But ultimately, if you have 2 delivery companies and one owns the roads, there is no way to compete against that. So the telephone and cable monopolies bought out all the dial-up ISPs, then all the PPTP broadband ISPs, and now *poof* we have a "natural" monopoly.
Quantum physics seems to be the ultimate proof that the universe is a simulation.
The universe, intuitively, seems to be analog and continuous. That "feels" right to us. But quantum physics shows that it is actually discrete. But that is exactly how computer simulations work! They use very small time scales to make things appear continuous. We know that below certain time scales, things are essentially random. This is consistent with a computer simulation. You can't accurately simulate something that happens in less time than one "frame" of time. There is a whole area of mathematics that deals with how to make simulations work accurately given the limitation of discrete time scales.
The same happens with physical sizes. Below the Planck scale the universe starts to break-down and become random. This is exactly how things would work if the universe was using binary arithmetic. Suppose that every particle in the universe has a coordinate. You can represent it's position over a vast scale, but only with limited accuracy. The plank scale is that limit, and it indirectly tells us how many bits are in the coordinate field of each particle. When we try to measure the position of something accurately, we find that the position becomes random. And if you try to measure it's speed to more resolution than one "frame" of time, it becomes less accurate. Worse-yet: the only way we can measure the position or speed of a simulated particle is by comparing it to another simulated particule, which introduces yet more error. We are ultimately limited by the accuracy of the simulation.
One side-benefit of this is that we have an awesome source of stastically predictable randomness. Quantum computers are actually using the randomness of the simulator to take advantage of cpu-cycles that are "outside" of our universe. Within the simulator, we can only build a computer that is so fast. But if we find a way to tap into the computing power of the simulator, like by using the side-effects of one of it's built-in functions, then we can compute a result faster than anything we can do ourselves. It is like calling into "native code" while we are running in the interpreted bytecode.
Another indication that we are in a simulation is that quantum physics shows us that wave functions collapse when we observe them. That makes sense: why should the universal simulator waste time calculating quantities that are not currently being measured? Imagine a vast number of inputs, a vast number of calculations that produce outputs, and a smaller number of observers of those outputs. You can easily optimize away things that are not being observed. But we found a way to notice the side-effect of not calculating certain values. It's like a side-channel attack on an encryption algorithm. You can tell how many bits of a password are correct even without the output by seeing how long it took to calculate, or how much power the computer consumed. I wonder if the designers of the simulator didn't know that we could see these kinds of side-effects, or if they are too difficult to fix. Either way, we are seeing side-effects of some of the shortcuts and optimizations.
Perhaps one day one of the programmers will look over at their printer and find a little note from someone way down here inside the simulation. If you could hack a few words outside of the system, what would they be?
You are right, but with a caveat. But the reason they don't have overlapping markets is because the local governments give exclusive cable contracts. So it isn't that the companies were forming a cartel, it is that the governments were enforcing a cartel. The companies might have actually wanted to compete, and the government was forbidding it.
That's great to know! Why is it that commercial photo / music management tools include options like "upload to facebook" and "upload to dropbox" but not "upload to SFTP/SCP server" which should work with anything? Is there some recent abhorrence to using standard protocols?
True, but it would cost an obscene amount of money and/or be obnoxiously large.
A watch with 16 buttons and an LCD display that is 1.6 x x0.3 x 1.3 inches costs $25 at Target. This would need only 10 buttons, and no display. Completely feasible.
Slashdot Mirror
Is it possible that the solution is that Netflix basically are forced to have multiple ISPs and connect directly to many networks?
Yes. The article indirectly confirms it:
Netflix has historically routed its streaming content to broadband providers through a number of Internet middlemen
Netflix probably has servers all over the world, so they would have many many ISPs. Those ISPs may have peering agreements, but Netflix still pays either way.
This does not sound like a case of Comcast "double dipping" where Comcast gets paid by the customer and by Netflix for bandwidth. And the article makes no implication that Comcast was going to do something non-neutral like slow down Netflix traffic. This looks more like Netflix is co-locating servers inside Comcast's data centers to avoid paying big bandwidth bills to tier 1 network providers.
Upon first reading this article I jumped to the conclusion that Comcast was slowing Netflix traffic and Netflix was paying them for some kind of "priority" service to get their bandwidth back up. But I don't think that is the case.
This is still a dangerous idea. What happens if Netflix decides to just stop supporting Internet streaming, and instead just streams to ISPs who they partner with? Then you will only get Netflix if you subscribe to an ISP that supports them. Sound familiar? It's basicall the Cable TV networks all over again.
My employer uses a MITM HTTPS proxy. The IT department pushed down a trusted corporate certificate, and most people don't even know their HTTPS connections aren't secure any more. The real problem is when some application, other than a browser, needs internet access and it fails. This includ sethings like web installers that download the app during installation, automatic update systems, secure file transfer software, or things that call home to confirm a license key. On occassion a developer curses some installer for not working, then we inspect the install.log file and find something about a certificate failure.
IT departments forget that HTTPS is used for more than just browsing the web.
I wish my Roomba had this ability
Some of the competitors products do. I forget which one(s). There was a neat youtube video of some company's robot and it would decide when it was done a room and not go back into it, just by mapping out where the doors were as it went along.
Most artificial muscles work by applying electric current along the muscle. When the current is removed they snap back to their original shape. Using heat sounds very limiting. Presumably you cool it to get it back to the original shape, but the ABC article is light on details.
I've been wanting something like this since I was a kid. Even if for no reason other than to feel totally cool. (Yes, everyone else will think I'm a dork for wearing it, but that's the story of my life. I wear by 8-bit tie proudly!)
In 1960, a phone call could be placed from any point in the United States that had a 10 lb telephone hard-wired to it to any other point in the United States that had a 10 lb telephone hard-wired to it and the sound quality would be consistently good.
FTFY.
And the plaintiff was named Miranda, which in U.S. law has special meaning regarding detaining people! This is almost like a cartoon.
The US outlets are covering the meta-story that Snowden did this. The Guardian is releasing the actual data. Big difference.
Those leaks were reported through The Guardian, a UK newspaper.
Of course the whole thing is just mental masturbation nonsense
If you wish to boil philosophy down to nonsense, so be it. This is somewhere between philosophy, physics, and mathematics. Maybe it is only theoretical right now. Perhaps, if we start thinking of the universe this way, it will yield testable predictions.
The implicit limit in binary representation isn't that you get something random if you try to slice between two values, you get either one or the other.
Not really. Check out the articles on numerical stability that I linked to and look at zwei2stein's post and my reply. Suppose for a moment they are using floating-point. As you approach the limits of numerical stability on floating-point, you get crazy results and aliasing. I proposed that randomness was added intentionally to avoid aliasing problems.
When that particular comment was made, the ubiquity of the home router dolling out DHCP addresses probably wasn't considered.
The report was in Feb 2011. Home routers have been doing this for 10 years. The reality is that home routers doing this is actually part of the problem. It's a real hassle for game developers, file sharers, Tor users, media servers, ...
I didn't mean "we" as in "intelligent life" I meant "we" as in "any output of that wave function." I clarified this in another reply earlier on in the discussion. Sorry.
Replying to my own post.
we would see very clear patterns and aliases.
I agree. That is one of the holes in my pseudoscience.
I thought about it more and remembered my reasoning on this. I propose that the creator of the simulation intentionally randomized these small quantities to eliminate the aliasing. It is effectively an unpredictable version of symmetric rounding.
If they didn't do this, the simulation might gain or lose energy over time because of these small errors. Or you might wind-up with something like values tending to be positive slighty more often than they are negative (or vice-versa) which could imbalance some things. (Maybe they didn't get it totally right, and that explains CP-violation and/or matter antimatter asymmetry?)
we would see very clear patterns and aliases.
I agree. That is one of the holes in my pseudoscience.
...not because of some weird universe limit, but because you effect particles by measuring them...
I would be fine with that if it weren't for the EPR paradox. Yes, we found a mathematical way to describe the result. It looks... suspicious to me. It seems like Einstein and friends found a way around the system, to read the *real* value. Then we tried it and noticed that value isn't actually there! It's like a splinter in my mind.
And actually computing whether something will be observed sounds like too much work to be considered a optimalization.
I can interpret that sentence two ways. By "too much work" do you mean "it would take more calculations to determine if something is observed than to simply calculate the result" or do you mean "the designer of the simulation would not bother to implement such an optimization." If you mean the second, I can't argue their intent so I don't know. If you mean the first, then I disagree. There are lots of cases of optimizations in simulators that work just like that. Simulations often calculate if something is off-screen or irrelevant before rendering it.
No optimization for intelligence is required. Wave don't merely collapse when something *intelligent* is observing it. They collapse when *anything* is observing it. Any time the output of the function is hooked to the input of another. If a snail opened Schrödinger's box, or if a stray rock struck the latch, the inside state would still collapse. It doesn't wait until a person walks over to it and looks.
I saw another post equating physics and mathematics. That is a new concept to me and not one I intended to imply. I postulate that that physics is *based on* mathematics. Curiously, we notice that the limitations of physics are similar to the limitations of numerical stability. That does not imply that physics is mathematics. It does not imply that mathematics has limitations. It implies that physics has limitations, which implies that it is not real.
Try it. I bet it won't work past a certain point. :-)
Just because you speculate that you could think of a big number or a small number isn't enough. It's fine, until we try to actually measure something that small or that big, and suddenly we find it doesn't work. I imagine you as a video game character who thought he could get infinite lives if he played really well, then one day notices his score is -2147483648 or number of lives is -127. But you have to play perfectly to ever even figure that out.
Yay someone who gets it! But you don't have to have the municipality own the infrastructure. You could still have a monopoly do it, so long as they only ran the lines and didn't act as an ISP. Permit me to expand on what you said, for the benefit of anyone stumbling upon this.
Those contracts are a result of physical limitations that create an almost natural monopoly. (You can't run dozens of cable lines along the poles, it would be a mess)...
This was true in the 1960s and 1970s when Comcast was first starting. But it is no longer true today since you can run multiple services over a single wire. This is how power is today, and how internet used to be in the 1990s. Everyone forgets that this limitation no longer applies, and so our regulatory framework is stuck in the past.
...Now, if towns owned the cable infrastructure and other network devices involved, and just leased to Comcast / Whoever to act as an ISP...
With power, one company builds and maintains the wires. Many companies provide the actual power, and customers can choose their power provider. This is really easy with power since power is fungible. The power companies just put the power on the wires and off it goes.
With telephone in the 1990s, one company built and maintained the wires. They also provided local dial-up or PPTP access. Many companies provided internet service, and customers could choose their ISP. The telephone wire company merely provided the means to connect the customer to the service.
With broadband internet, it could be the same way and it would work exactly like how telephone used to work. Telephone company provides PPTP, and you get a box that does PPTP to your ISP who could be anyone anywhere.
What happened that messed this up was that the cable and telephone monopolies were permitted to become ISPs. They had special accses tot he system, so the 3rd-party ISPs couldn't compete. For a while, in some states, there was legislation that actually required the telephone companies to allow these other ISPs to lease access to the lines at certain fixed prices. This was a poor attempt to allow competition. But ultimately, if you have 2 delivery companies and one owns the roads, there is no way to compete against that. So the telephone and cable monopolies bought out all the dial-up ISPs, then all the PPTP broadband ISPs, and now *poof* we have a "natural" monopoly.
Quantum physics seems to be the ultimate proof that the universe is a simulation.
The universe, intuitively, seems to be analog and continuous. That "feels" right to us. But quantum physics shows that it is actually discrete. But that is exactly how computer simulations work! They use very small time scales to make things appear continuous. We know that below certain time scales, things are essentially random. This is consistent with a computer simulation. You can't accurately simulate something that happens in less time than one "frame" of time. There is a whole area of mathematics that deals with how to make simulations work accurately given the limitation of discrete time scales.
The same happens with physical sizes. Below the Planck scale the universe starts to break-down and become random. This is exactly how things would work if the universe was using binary arithmetic. Suppose that every particle in the universe has a coordinate. You can represent it's position over a vast scale, but only with limited accuracy. The plank scale is that limit, and it indirectly tells us how many bits are in the coordinate field of each particle. When we try to measure the position of something accurately, we find that the position becomes random. And if you try to measure it's speed to more resolution than one "frame" of time, it becomes less accurate. Worse-yet: the only way we can measure the position or speed of a simulated particle is by comparing it to another simulated particule, which introduces yet more error. We are ultimately limited by the accuracy of the simulation.
One side-benefit of this is that we have an awesome source of stastically predictable randomness. Quantum computers are actually using the randomness of the simulator to take advantage of cpu-cycles that are "outside" of our universe. Within the simulator, we can only build a computer that is so fast. But if we find a way to tap into the computing power of the simulator, like by using the side-effects of one of it's built-in functions, then we can compute a result faster than anything we can do ourselves. It is like calling into "native code" while we are running in the interpreted bytecode.
Another indication that we are in a simulation is that quantum physics shows us that wave functions collapse when we observe them. That makes sense: why should the universal simulator waste time calculating quantities that are not currently being measured? Imagine a vast number of inputs, a vast number of calculations that produce outputs, and a smaller number of observers of those outputs. You can easily optimize away things that are not being observed. But we found a way to notice the side-effect of not calculating certain values. It's like a side-channel attack on an encryption algorithm. You can tell how many bits of a password are correct even without the output by seeing how long it took to calculate, or how much power the computer consumed. I wonder if the designers of the simulator didn't know that we could see these kinds of side-effects, or if they are too difficult to fix. Either way, we are seeing side-effects of some of the shortcuts and optimizations.
Perhaps one day one of the programmers will look over at their printer and find a little note from someone way down here inside the simulation. If you could hack a few words outside of the system, what would they be?
You are right, but with a caveat. But the reason they don't have overlapping markets is because the local governments give exclusive cable contracts. So it isn't that the companies were forming a cartel, it is that the governments were enforcing a cartel. The companies might have actually wanted to compete, and the government was forbidding it.
That's great to know! Why is it that commercial photo / music management tools include options like "upload to facebook" and "upload to dropbox" but not "upload to SFTP/SCP server" which should work with anything? Is there some recent abhorrence to using standard protocols?
So do aircraft not take off or land in your part of the world?
In my part of the world, aircraft take off from airports.
True, but it would cost an obscene amount of money and/or be obnoxiously large.
A watch with 16 buttons and an LCD display that is 1.6 x x0.3 x 1.3 inches costs $25 at Target. This would need only 10 buttons, and no display. Completely feasible.
In theory yes. I simply don't believe it.