Making Tor dead simple to use is great, but this is such a nice device for three-letter agencies to target inserting a backdoor into.
While that is a possibility(albeit one that could theoretically be ameliorated, barring hardware-level backdoors, by 'here's how to build Tor from mainline and replace our firmware' documentation), I'd be more worried about the fact that Tor isn't dead simple.
The project itself has a list of handy warnings concerning What Not To Do on Tor and expect the anonymity to keep working, even assuming there are no unknown attacks and vulnerabilities at play. Tor has no magical ability to scrub dangerously identifying information from the assorted dumb, lazy, or just plain user-hostile chatter generated by various programs on your computer. It also, as a necessary side effect of its design, exposes some traffic to the exit node, which requires that you be careful about SSL/TLS for anything that the exit node shouldn't see.
That's what makes me nervous about the projects(hardware or software, boxes like this or Android VPN plugins, or whatever) that make it dead easy to route all traffic through Tor. Unless you know exactly what you are doing, that probably isn't what you want. Your day-to-day OS is very likely to be far too dangerously chatty(which means that you really shouldn't use it at all, unless booted to a liveCD; with the Tor browser bundle, that passes only traffic from the Tor browser as a distant second best); but you definitely shouldn't just plug it into the magic Tor box. Some applications you just don't want going through Tor at all. If the traffic is intrinsically personally identifying the best case is that you'll gain nothing and the worst case is that you'll be less secure than you were.
Things that keep people from running the browser bundle on their poxed XP machines and expecting anonymity are good; but Tor simply isn't easy to use, even if it is made easy to set up, and that can bite you in the ass.
They certainly are reluctant about it at times; but the military has the convenient 'reality is what shoots you even when you try to pretend it isn't there' incentive to remain members of the reality-based-community.
Has anybody actually examined the difference in bandwidth consumption(obviously Netflix has; but I was hoping for a 3rd party)?
'4k' is four times the pixels; but your target bitrate is a different question. I doubt they'd be gutsy enough to keep it the same as for 1080p; but they could have concluded that 3 times the bitrate actually looks just fine (or, less likely but possible, that anything they can get in '4k' is more likely to have been produced at high resolution all the way from camera to final output and more than four times the bandwidth is needed to keep from munging the result).
It would also be interesting to know(but a lot harder to test yourself) how much of the additional cost is storage and bandwidth and how much, if any, is the "More Pixels Means More Premium!" content markup that we saw between DVDs and BDs, where the same title would cost more if you wanted it at higher resolution.
Netflix can certainly attempt to charge what they wish, and it's certainly possible that the difference in price closely reflects the actual difference in cost; but I'd be interested to know more about how much more costly the additional resolution is, and where the additional costs come from.
This 'study' is invalid on its face because Free Will. It simply must not be possible that behavior is driven by causal, physical, mechanisms because that would be philosophically inconvenient. Therefore, it must be false.
I'm...not exactly sure... that the clandestine services of the world are worried about legal exposure incurred in the course of their activities. I certainly can't think of any being bitten in the ass for deploying spyware and it's a matter of public knowledge that it has been done reasonably frequently.
If you just take them down, you get nothing, not even traffic data. If you distribute malware, you get continued tracking of people who visited, possibly keylogger data, dumps of address books and contact lists, credentials for other accounts, and other fun stuff.
It'd be childish if the 'malware' were just serving pop-up ads or sending herbal viagra spam. The stuff designed for surveillance of infected targets, though, would be an entirely logical intelligence gathering strategy.
There seems to be one major difference between this situation and your analogy:
In the case of the neighbor and the TV, the neighbor is a moral person with an interest in his TV (and even if he is dead, his 'estate' has an interest in it, and even if the TV is going to be destroyed, it is likely to have insurance coverage that he won't get if you salvage it and it isn't destroyed after all). It quite likely would be ethical for you to help him out by saving some stuff from the fire; but you'd be a total dick to save some stuff from the fire, then tell him 'haha, no, mine now!' rather than "Here, I got these out for you."
In this case, your 'neightbor' only exists if you treat the possible-future-child (that might have been produced from the embryo, if somebody was found who wanted to use it, most IVF surplus just stays on ice until it eventually gets tossed, since people tend to want either their own children, or go with adoptions of already-born children) as a moral person. This is a position that some take, and arguably a cogent one; but it's far from immediately proven.
If you don't consider the embryo to be a moral person(at least not this early in development, if it was chosen for cellular plasticity it probably doesn't even have a neural network yet), then there is no 'neighbor' analog to speak of. Simply an item of value that can either be used or be allowed to be destroyed.
The design would be worth billions, if you add in a bunch of investment money. (which would flow to a contest winner) And the $20k easily pays back the materials cost of a garage/makerspace-built prototype.
That's why I included the bit about "absolutely no loss of patents or copyrights, or grants of license to amazon, for anything related to the robot." bit.
The value of something that is good enough to win the prize is far larger than the prize. If you can compete without compromising your control of your entry in any way, it only costs the time of whoever you send to unpack the robot and participate. If touching the contest involves granting any right to your entry, you should probably run away screaming.
Even if they did, a cure is only less profitable than daily maintenance treatment if you price the cure such that the profit you make from selling it is less than the net present value of the profit made on all sales of the maintenance treatment across the patient's expected lifetime(future profits are discounted because you don't get the money upfront and because of risk: the patient might die, switch suppliers, be cured by someone else, lose insurance and just suffer without, etc.)
Given the lifetime cost of maintenance treatment for Type I diabetes, an equivalent-profit cure sure as hell wouldn't be cheap; but there's nothing magically profitable about selling consumables that somehow doesn't work for one-time purchases.
Abort that fetus today, you could save a diabetic person's foot...
They may have other uses; but fetuses old enough to abort are not a good supply of embryonic stem cells (all those adorably baby-like features you see on the gory antiabortion posters? That's because the cells have already differentiated and lost most of the really cool capabilities of early stage embryonic cells...)
If you want to do your part for the stem cell industrial complex, get some IVF done.
Honestly, the hardest thing about subverting free will is actually chasing down a demonstrable instance of it to subvert.
The business of modifying people's behaviors and decisions, on the other hand, at least at the "all of the people some of the time or some of the people all of the time" population level is absolutely ubiquitous, rather effective, and at least as old as civilization.
It's hard to anticipate much of value from somebody who manages to cram so many terms that are both heavily loaded and almost totally vacuous into a single sentence.
In addition to your point (it's very, very, unlikely that the embryos were produced for this purpose at all, let alone solely, since most of the embryo supply is surplus from IVF work, which humans obviously do for other reasons), the rest of the sentence is little more than a fabric of implicit assertions padded with a few nearly meaningless bits.
"Destruction": This process only works because the embryonic cells are undifferentiated (any later in embryonic development and it'd be fun with cell reprogramming) and will only be medically relevant if the resulting beta cells form a reasonably long lived cell line(possibly not immortal; but the more frequently the patient needs new ones implanted, the lower the benefit over just injecting insulin). Does this change in developmental trajectory count as 'destruction'? Arguable; but hardly self-evident.
"Individual": As opposed to the other kind? Did I miss all those collective humans out there? Maybe a hive mind? What would a 'non-individual' human life even look like?
"Unique": Both irrelevant (would the procedure be somehow more or less ethical if it were non-unique? One of those creepy, soulless, clones?) and questionably accurate (very early stage embryos can, and sometimes do, split and form two cell masses that each continue to divide. We call them 'identical twins' and usually don't tell them that they are non-unique, or that a 'unique human life' was destroyed when the original zygote split into two). Given the age of the cells the researchers were working with, chosen specifically for their plasticity, it's actually somewhat tricky to argue that the embryo is 'individual' and 'unique'. In a terribly vacuous sense it is (this embryo is unique because no other embryo is also this embryo); but beyond that you really have to argue for it.
"Human life": This one is as old as the hills, and a classic of the abortion wars. Is it human? Yeah, sure, to the same degree that any other cells in my body are. Is it a 'human life' in the moral personhood sense that you are invoking? Arguable; but you certainly haven't argued it yet.
I sure hope that this bioethicist was either taken out of context or hasn't given up the day job.
John Henry? Isn't that the guy who voluntarily worked a bunch of unpaid overtime to finish his last project before being automated into obsolescence; and then died just in time to save his employer the risk of paying any sort of severance?
Unless you are Jeff Bezos, you might consider a story with a different moral.
It might be worth it if you can enter with absolutely no loss of patents or copyrights, or grants of license to Amazon, for anything related to the robot you enter; but only if you were already working on the problem for reasons related to an amount of money that ranks somewhere between 'insulting' and 'hilarious' in comparison to the value of the task. $20k would likely have some difficultly covering even your expenses, much less actually rewarding you, and that's the big prize.
I'm not sure that it's specifically an intelligence problem. It's going to be essentially impossible to write a convincing 'genius' or distinguish him from a 'savant' unless both the writer and the audience have at least an approximate idea of how the difficulty level in their discipline is distributed.
Computers are a horrific subject for that. People don't know what's easy, what's hard, what's suspected to be impossible but so far not proven to be, what would leave a *nix-using CS expert puzzled but be solved in moments by a geek squad kid who has fixed exactly the same broken update 30 times this week, and so on.
Oh, I was thinking of putting the TPM on the GPU and treating the rest of the system as untrusted, rather than trying to control everything. Controlling everything would work as well; but steps on a lot more toes, makes a great many more things Your Problem, and lengthens the chain of critical-things-you'll-probably-screw-up-somewhere. The GPU vendor only has an economic incentive to control a few performance-related parameters tightly coupled to the GPU silicon, so as long as the card can know its own model number(presumably lasered into the die when parts were being binned) and verify the signature on a trivial little manifest file that provides performance settings for each model number, it need trust nobody else.
As it happens, Nvidia is apparently trying something along these lines. They are apparently verifying the whole firmware, and denying specific items to unverified firmware, rather than ignoring most of the firmware and only signing for the specific items; but the effect is fairly similar. GPUs are tightly coupled to their hosts; but they have enough onboard capability to implement what is effectively a 'locked bootloader' behavior without exerting control over the remainder of the system(any more than only applying updates from signed repositories requires control of the internet).
(What won't be pleasant, though, is Team Copyright demanding that all video memory where precious 'premium content' might momentarily reside be protected from access by everyone and everything.)
I'm not holding my breath -- fusion power has been 20-30 years away since the 70s.
In fairness, fusion power works just fine if you scale it up. It's just the attempts to make it work in systems that don't weight ~2x10^29kg or more that haven't been so hot.
These hardware companies will never truly make their systems open, equally able to be used by any OS written by anyone under any license.
The reason? They want to be able to sell the same hardware in different "models" at different prices, with the difference enforced by binary blobs.
I'm not holding my breath for a blobless future; but architecturally that's a somewhat antique way of enforcing a restriction: a TPM, or TPM-like crypto chip costs maybe 2-3 dollars as a discrete chip in modest volumes, likely less in considerable quantity or if the necessary functions are baked into some other chip to save on packaging and board space. If you have one of those, you don't need some big binary blob driver, or even an OSS driver talking to a binary blob in userspace in order to cripple (did I say cripple? I meant 'differentiate'). A simple "model numbers and how fast they should go.txt" configuration file becomes essentially bulletproof once it's digitally signed.
The question isn't really whether last year's card can support this year's features; but whether they can do so remotely adequate speed. Recent-ish cards are turing complete on their own, and nothing forbids doing some amount of work on the CPU by building that into the driver. Frame rate probably won't be pretty, though.
It might help that 'appliance' pricing tends to be fairly close to software pricing; but with thousand to few thousand dollars worth of vaguely custom server thrown in. Unlike the price for hardware (especially the lower end commodity stuff), where the vendor is looking at actual losses (in shipping, refurbishing, and selling formerly new goods at refurb prices), much of the price of your appliance is probably copies of software and some sort of support agreement.
Not that they'll want to forgo profit; but the direct cost of letting you return the junk, cancel the remainder of the contract, and stop cursing their name will be lower.
Unless he got a fairly amazing contract, 'sucky and frustrating but more or less functional' will meet or exceed anything they actually made a legally binding promise to deliver.
It's not as though he had a very good shot at getting a better contract(better product, possibly; but even that vendor wouldn't actually promise any more), so I wouldn't be inclined to blame him; but it'd take some seriously impressive suck to actually give him any legal ground to work with.
Slashdot Mirror
Making Tor dead simple to use is great, but this is such a nice device for three-letter agencies to target inserting a backdoor into.
While that is a possibility(albeit one that could theoretically be ameliorated, barring hardware-level backdoors, by 'here's how to build Tor from mainline and replace our firmware' documentation), I'd be more worried about the fact that Tor isn't dead simple.
The project itself has a list of handy warnings concerning What Not To Do on Tor and expect the anonymity to keep working, even assuming there are no unknown attacks and vulnerabilities at play. Tor has no magical ability to scrub dangerously identifying information from the assorted dumb, lazy, or just plain user-hostile chatter generated by various programs on your computer. It also, as a necessary side effect of its design, exposes some traffic to the exit node, which requires that you be careful about SSL/TLS for anything that the exit node shouldn't see.
That's what makes me nervous about the projects(hardware or software, boxes like this or Android VPN plugins, or whatever) that make it dead easy to route all traffic through Tor. Unless you know exactly what you are doing, that probably isn't what you want. Your day-to-day OS is very likely to be far too dangerously chatty(which means that you really shouldn't use it at all, unless booted to a liveCD; with the Tor browser bundle, that passes only traffic from the Tor browser as a distant second best); but you definitely shouldn't just plug it into the magic Tor box. Some applications you just don't want going through Tor at all. If the traffic is intrinsically personally identifying the best case is that you'll gain nothing and the worst case is that you'll be less secure than you were.
Things that keep people from running the browser bundle on their poxed XP machines and expecting anonymity are good; but Tor simply isn't easy to use, even if it is made easy to set up, and that can bite you in the ass.
They certainly are reluctant about it at times; but the military has the convenient 'reality is what shoots you even when you try to pretend it isn't there' incentive to remain members of the reality-based-community.
Would you prefer advice on escapist fantasy or birth control? I can provide either.
Because it hurts those poor ISPs so much when they are forced to charge you an overage fee...
Has anybody actually examined the difference in bandwidth consumption(obviously Netflix has; but I was hoping for a 3rd party)?
'4k' is four times the pixels; but your target bitrate is a different question. I doubt they'd be gutsy enough to keep it the same as for 1080p; but they could have concluded that 3 times the bitrate actually looks just fine (or, less likely but possible, that anything they can get in '4k' is more likely to have been produced at high resolution all the way from camera to final output and more than four times the bandwidth is needed to keep from munging the result).
It would also be interesting to know(but a lot harder to test yourself) how much of the additional cost is storage and bandwidth and how much, if any, is the "More Pixels Means More Premium!" content markup that we saw between DVDs and BDs, where the same title would cost more if you wanted it at higher resolution.
Netflix can certainly attempt to charge what they wish, and it's certainly possible that the difference in price closely reflects the actual difference in cost; but I'd be interested to know more about how much more costly the additional resolution is, and where the additional costs come from.
This 'study' is invalid on its face because Free Will. It simply must not be possible that behavior is driven by causal, physical, mechanisms because that would be philosophically inconvenient. Therefore, it must be false.
I'm...not exactly sure... that the clandestine services of the world are worried about legal exposure incurred in the course of their activities. I certainly can't think of any being bitten in the ass for deploying spyware and it's a matter of public knowledge that it has been done reasonably frequently.
If you just take them down, you get nothing, not even traffic data. If you distribute malware, you get continued tracking of people who visited, possibly keylogger data, dumps of address books and contact lists, credentials for other accounts, and other fun stuff.
It'd be childish if the 'malware' were just serving pop-up ads or sending herbal viagra spam. The stuff designed for surveillance of infected targets, though, would be an entirely logical intelligence gathering strategy.
There seems to be one major difference between this situation and your analogy:
In the case of the neighbor and the TV, the neighbor is a moral person with an interest in his TV (and even if he is dead, his 'estate' has an interest in it, and even if the TV is going to be destroyed, it is likely to have insurance coverage that he won't get if you salvage it and it isn't destroyed after all). It quite likely would be ethical for you to help him out by saving some stuff from the fire; but you'd be a total dick to save some stuff from the fire, then tell him 'haha, no, mine now!' rather than "Here, I got these out for you."
In this case, your 'neightbor' only exists if you treat the possible-future-child (that might have been produced from the embryo, if somebody was found who wanted to use it, most IVF surplus just stays on ice until it eventually gets tossed, since people tend to want either their own children, or go with adoptions of already-born children) as a moral person. This is a position that some take, and arguably a cogent one; but it's far from immediately proven.
If you don't consider the embryo to be a moral person(at least not this early in development, if it was chosen for cellular plasticity it probably doesn't even have a neural network yet), then there is no 'neighbor' analog to speak of. Simply an item of value that can either be used or be allowed to be destroyed.
Am I out of consideration if I refer to the polygraph as 'truth dowsing' while it is being administered? How about asking if it can detect witches?
The guys with the space program get to use the acronym. Those are the rules.
The design would be worth billions, if you add in a bunch of investment money. (which would flow to a contest winner) And the $20k easily pays back the materials cost of a garage/makerspace-built prototype.
That's why I included the bit about "absolutely no loss of patents or copyrights, or grants of license to amazon, for anything related to the robot." bit.
The value of something that is good enough to win the prize is far larger than the prize. If you can compete without compromising your control of your entry in any way, it only costs the time of whoever you send to unpack the robot and participate. If touching the contest involves granting any right to your entry, you should probably run away screaming.
Even if they did, a cure is only less profitable than daily maintenance treatment if you price the cure such that the profit you make from selling it is less than the net present value of the profit made on all sales of the maintenance treatment across the patient's expected lifetime(future profits are discounted because you don't get the money upfront and because of risk: the patient might die, switch suppliers, be cured by someone else, lose insurance and just suffer without, etc.)
Given the lifetime cost of maintenance treatment for Type I diabetes, an equivalent-profit cure sure as hell wouldn't be cheap; but there's nothing magically profitable about selling consumables that somehow doesn't work for one-time purchases.
Abort that fetus today, you could save a diabetic person's foot...
They may have other uses; but fetuses old enough to abort are not a good supply of embryonic stem cells (all those adorably baby-like features you see on the gory antiabortion posters? That's because the cells have already differentiated and lost most of the really cool capabilities of early stage embryonic cells...)
If you want to do your part for the stem cell industrial complex, get some IVF done.
Honestly, the hardest thing about subverting free will is actually chasing down a demonstrable instance of it to subvert.
The business of modifying people's behaviors and decisions, on the other hand, at least at the "all of the people some of the time or some of the people all of the time" population level is absolutely ubiquitous, rather effective, and at least as old as civilization.
It's hard to anticipate much of value from somebody who manages to cram so many terms that are both heavily loaded and almost totally vacuous into a single sentence.
In addition to your point (it's very, very, unlikely that the embryos were produced for this purpose at all, let alone solely, since most of the embryo supply is surplus from IVF work, which humans obviously do for other reasons), the rest of the sentence is little more than a fabric of implicit assertions padded with a few nearly meaningless bits.
"Destruction": This process only works because the embryonic cells are undifferentiated (any later in embryonic development and it'd be fun with cell reprogramming) and will only be medically relevant if the resulting beta cells form a reasonably long lived cell line(possibly not immortal; but the more frequently the patient needs new ones implanted, the lower the benefit over just injecting insulin). Does this change in developmental trajectory count as 'destruction'? Arguable; but hardly self-evident.
"Individual": As opposed to the other kind? Did I miss all those collective humans out there? Maybe a hive mind? What would a 'non-individual' human life even look like?
"Unique": Both irrelevant (would the procedure be somehow more or less ethical if it were non-unique? One of those creepy, soulless, clones?) and questionably accurate (very early stage embryos can, and sometimes do, split and form two cell masses that each continue to divide. We call them 'identical twins' and usually don't tell them that they are non-unique, or that a 'unique human life' was destroyed when the original zygote split into two). Given the age of the cells the researchers were working with, chosen specifically for their plasticity, it's actually somewhat tricky to argue that the embryo is 'individual' and 'unique'. In a terribly vacuous sense it is (this embryo is unique because no other embryo is also this embryo); but beyond that you really have to argue for it.
"Human life": This one is as old as the hills, and a classic of the abortion wars. Is it human? Yeah, sure, to the same degree that any other cells in my body are. Is it a 'human life' in the moral personhood sense that you are invoking? Arguable; but you certainly haven't argued it yet.
I sure hope that this bioethicist was either taken out of context or hasn't given up the day job.
John Henry? Isn't that the guy who voluntarily worked a bunch of unpaid overtime to finish his last project before being automated into obsolescence; and then died just in time to save his employer the risk of paying any sort of severance?
Unless you are Jeff Bezos, you might consider a story with a different moral.
It might be worth it if you can enter with absolutely no loss of patents or copyrights, or grants of license to Amazon, for anything related to the robot you enter; but only if you were already working on the problem for reasons related to an amount of money that ranks somewhere between 'insulting' and 'hilarious' in comparison to the value of the task. $20k would likely have some difficultly covering even your expenses, much less actually rewarding you, and that's the big prize.
I'm not sure that it's specifically an intelligence problem. It's going to be essentially impossible to write a convincing 'genius' or distinguish him from a 'savant' unless both the writer and the audience have at least an approximate idea of how the difficulty level in their discipline is distributed.
Computers are a horrific subject for that. People don't know what's easy, what's hard, what's suspected to be impossible but so far not proven to be, what would leave a *nix-using CS expert puzzled but be solved in moments by a geek squad kid who has fixed exactly the same broken update 30 times this week, and so on.
Oh, I was thinking of putting the TPM on the GPU and treating the rest of the system as untrusted, rather than trying to control everything. Controlling everything would work as well; but steps on a lot more toes, makes a great many more things Your Problem, and lengthens the chain of critical-things-you'll-probably-screw-up-somewhere. The GPU vendor only has an economic incentive to control a few performance-related parameters tightly coupled to the GPU silicon, so as long as the card can know its own model number(presumably lasered into the die when parts were being binned) and verify the signature on a trivial little manifest file that provides performance settings for each model number, it need trust nobody else.
As it happens, Nvidia is apparently trying something along these lines. They are apparently verifying the whole firmware, and denying specific items to unverified firmware, rather than ignoring most of the firmware and only signing for the specific items; but the effect is fairly similar. GPUs are tightly coupled to their hosts; but they have enough onboard capability to implement what is effectively a 'locked bootloader' behavior without exerting control over the remainder of the system(any more than only applying updates from signed repositories requires control of the internet).
(What won't be pleasant, though, is Team Copyright demanding that all video memory where precious 'premium content' might momentarily reside be protected from access by everyone and everything.)
I'm not holding my breath -- fusion power has been 20-30 years away since the 70s.
In fairness, fusion power works just fine if you scale it up. It's just the attempts to make it work in systems that don't weight ~2x10^29kg or more that haven't been so hot.
These hardware companies will never truly make their systems open, equally able to be used by any OS written by anyone under any license. The reason? They want to be able to sell the same hardware in different "models" at different prices, with the difference enforced by binary blobs.
I'm not holding my breath for a blobless future; but architecturally that's a somewhat antique way of enforcing a restriction: a TPM, or TPM-like crypto chip costs maybe 2-3 dollars as a discrete chip in modest volumes, likely less in considerable quantity or if the necessary functions are baked into some other chip to save on packaging and board space. If you have one of those, you don't need some big binary blob driver, or even an OSS driver talking to a binary blob in userspace in order to cripple (did I say cripple? I meant 'differentiate'). A simple "model numbers and how fast they should go.txt" configuration file becomes essentially bulletproof once it's digitally signed.
The question isn't really whether last year's card can support this year's features; but whether they can do so remotely adequate speed. Recent-ish cards are turing complete on their own, and nothing forbids doing some amount of work on the CPU by building that into the driver. Frame rate probably won't be pretty, though.
It might help that 'appliance' pricing tends to be fairly close to software pricing; but with thousand to few thousand dollars worth of vaguely custom server thrown in. Unlike the price for hardware (especially the lower end commodity stuff), where the vendor is looking at actual losses (in shipping, refurbishing, and selling formerly new goods at refurb prices), much of the price of your appliance is probably copies of software and some sort of support agreement.
Not that they'll want to forgo profit; but the direct cost of letting you return the junk, cancel the remainder of the contract, and stop cursing their name will be lower.
Unless he got a fairly amazing contract, 'sucky and frustrating but more or less functional' will meet or exceed anything they actually made a legally binding promise to deliver.
It's not as though he had a very good shot at getting a better contract(better product, possibly; but even that vendor wouldn't actually promise any more), so I wouldn't be inclined to blame him; but it'd take some seriously impressive suck to actually give him any legal ground to work with.