See the 'or simply emotional majoritarianism' option. Anything the majority (or loud minority that pines for the good ol' days) approves of isn't 'offensive' in the sense used to argue against things; because it is simply right and proper that they exercise their power; but such things are quite frequently calculated to offend those who don't have the power to do anything about it.
Having an implant that can 'patch over' damaged brain areas should make lesion studies more subtle and precise. They can certainly tell us something already, at least in broad strokes, about what functions go where; but it's hard to shake the question of 'if you damage area X, does function Y suffer because area X handles it, or because it depends on connectivity through area X between areas W and Q?' If we have a technique for replacing a functional area with a mere transmission line, that gives us greater ability to differentiate between an area with a functional role in some function and an area with a merely connective role (presumably, there are also areas that are mostly connective; but apply some amount of signal processing between input and output. In the future, maybe we will be able to write arbitrary signal processing filters and patch them in, in software, between the input and the output of this 'bridge' device. That'd be extra neat).
If anything, it's the opposite: Experimentally, we observe a variety of effects as we prod the black box in various ways, examine it with various clever inferential techniques, and so on. In some cases, we are able to develop reliable "If you poke it there, it does that, reliably" type rules based on repeated observation. However, our prodding of the system has not yet provided enough information to posit the underlying structure, that would unify all our disparate observations of patterns. So, we don't have one. We might have some theorists, on a parallel track, tinkering with candidate unified theories; but operationally we only have our pattern inferences.
If it were faith based, it'd be way easier: we could just posit a Grand Unified Theory, and then bodge like crazy to force any observations into agreement with it.
Instead, we start with a whole bunch of tedious empiricism, observational inference of patterns, and hope that eventually somebody comes up with something that will elegantly explain the mess. It's Newton's old "Hypotheses non fingo": He had no idea how or why universal gravitation worked the way it did; but accorded with astronomical and physical observation, so he proposed it, without any metaphysical entanglements, as a model.
The evangelicals really should have remembered the Danbury Baptists...
Exercising the temporal power of fire and sword against your religious enemies is fun and all; but (even when you are on top) tends to be corrupting, and when you aren't, it opens the door to being at the mercy of every different group out there.
Plus, even among people who would ordinarily be inclined to treat your choices of faith as purely personal and let you believe as you will, nothing sours toleration quite like making it clear that you are ready and willing to impose what you believe on everyone else. Suddenly, and wholly because of your actions, your beliefs are now everybody's business; because everybody will suffer for them. That's when the gloves come off (most notably among atheists: 'god-not-existing' is something that isn't even worth mentioning, except that people who believe otherwise keep pushing the matter. In absence of pressure from theists, the nonexistence of god is about as interesting as the nonexistence of Russel's teapot.)
"Offensive" is an unhelpful criterion because it's a subjective assessment, and one that varies from person to person. Unless you couple it with a statistical cut-off of some kind, you probably couldn't do anything without offending somebody. Even deciding whether to go with grass or masonry paving for the public space wouldn't be uncontroversial...
"Offensive", in practice, is either meaningless (since everything is, to somebody) or simply emotional majoritarianism (if you only count as 'offensive' things that offend large and influential groups of people). Lousy criterion.
If you think atheists drive evangelical conservatives nuts, you ain't seen nothing yet.
Which is a trifle ironic, because 'satanists' (to the degree that they actually take the stuff seriously, and aren't just into heavy metal and upsetting their parents), are far closer, in terms of opinions on metaphysics, to Christians than atheists are. Especially to some of the protestant outfits that are practically Manichean in their emphasis on the power of satan in the world...
Though, given how much they like Muslims, who are closer still, I suppose that it may be a matter of hating your competitors even more than people in a different industry altogether.
No reason to worry, with our exciting advances in military robot technology, we can have robots solve the unemployment problems that other robots create!
Not to worry! With new FungusLease Managed Robotic Workforce technology, businesses mired in boring old 'making physical objects' industries can experience the joys of The Cloud by replacing an increasing number of aspects of their production process with robots whose hardware is leased from me, and whose software and configuration data are licensed in exchange for monthly fees in perpetuity! Don't worry, the part of the license granting me a perpetual nonexclusive license to all the configuration data and job parameters is, um, just legal boilerplate, not something I'd ever use to sell services to your competitors!
I suspect that one skilled in the art could perform a physical attack. Decapping chips and analyzing both their logical structure and their non-volatile memory contents isn't trivial; but it's a commercially available service with today's technology. TPMs are usually FIPS 140-2 compliant; but tend not to mention whether they are level 1, 2, 3, or 4 devices (which makes one suspect that the answer is unflattering).
The big question would be whether the DRM system has some sort of revocation support (blacklisting on XBL, 'burn' lists on game disks authored after the compromised unit is discovered, etc. The design of AACS is an example of how this sort of thing can be done). If not, it might cost a couple hundred thousand, more or less depending on the difficulty of the chip; but all you'd need is one private key. If there is a revocation mechanism, that makes physical attacks substantially less helpful, since MS(or any other party in the position of DRMed-platform vendor) can force you to pay the price, again and again, every time they catch you.
In my (admittedly layman's) understanding, the difficulty with the XB360/PS3 isn't so much that their cores are all that cryptic (the xbox, and the PS3's main core, are all basic PPCs, and the PS3's 'SPE' elements are weird; but IBM talked a lot about them in the course of trying to build interest in using them as accelerator cards for compute applications); but because they aren't x86; but are clocked as high as contemporary x86s, which makes it difficult to get emulation at anything remotely resembling usable speed, much less support anything that does some fancy timing-dependent tricks.
That's fair, Back when Apple sold "Apple"-line computers, they had a number of neat tricks up their sleeve. Once Woz got the shove, though, they still produced some very refined products(and some messes); but all in the direction of brutally refined and executed implementations of not terribly novel concepts (which isn't a bad thing, especially when the competition mostly has dubiously refined and sloppy implementations; but isn't "innovation").
Didn't Bell labs get spun off in some fashion? I have the vague memory that they are huddled under the Alcatel/Lucent corporate umbrella these days while AT&T finances basic research in contract-obfuscation theory by the nation's top theoretical lawyers...
On the plus side, emulating an AMD x86 and GPU is likely to be considerably easier (especially since AMD's current or near-future PC parts are likely to be extremely similar in most respects, though you will probably have to go up a few speed grades to deal with the emulator running on top of a full OS) than emulating either the relatively fast PPCs of the previous generation (PPC-on-x86 is done; but doing that really fast is another story) or the slow-but-somewhat-esoteric-and-absolutely-every-oddity-was-used-and-abused architectures of the older consoles.
On the minus side, the odds are good that both new consoles (especially the Xbox, given MS's software side; but probably the PS as well) contain a lot of software that, while not integral to the tightly-optimized-graphics-twiddling aspects of the games, will probably have to be given a fairly precise "WINE-like" treatment to avoid breaking things all over the place. Not necessarily impossible (as WINE itself demonstrates); but definitely a different game than the 'emulate the hardware and let the ROM do as it will' emulators that work for older consoles.
On the very minus side, it would not be out of character for either MS or Sony to have added some nasty copy-protection-related cryptographic goodies that will be very hard to emulate. MS, given their PC background, might well have gone for a TPM. Architecturally, emulating one of those would be cake by the standards of what the emulation scene has taken on, except for minor matters like the endorsement key. A TPM emulator that emulates a TPM loaded with the 2048-bit RSA private key of your choice? Sure, no problem. The correct private keys? That might be an issue.
Even if your CEO isn't supposed to be a visionary, and was basically hired to keep the good ship BeanCounter on an even keel, I'm pretty sure that he isn't supposed to make an ass of himself, and make the company sound reactionary and uncreative, in public. That's what I don't understand about the whole thing.
Even if drones are nonsense as a delivery platform, their PR/advertising utility in the 'we ship your shit crazy fast' narrative that Amazon has been trying to build around 'Prime' would seemingly be obvious.
All arguments about whether drones for package delivery are viable or not aside, I am honestly baffled that eBay's CEO would open his mouth on this one...
Even if it's 100%-aw-hell-no-never-going-to-happen, Amazon's work so far has likely been fairly inexpensive and has certainly stirred up as much attention as a decent sized ad campaign (the sort of thing that might actually cost as much or more to produce and buy airtime to run), so it isn't as though they are wallowing in shame and loss right now.
Under those circumstances, what possible benefit is there to a not-terribly-clever rubbishing of the opposition that just makes you look unhip and non-innovative? Especially when that is basically true; direct connection of buyers and sellers worldwide, in an easy-to-use, comparatively safe, framework may have been pretty cool when ebay hit the scene, but they hit the scene quite some time ago and have mostly been ratcheting up the transaction costs since then.
I personally have strong doubts about the viability of drone delivery; but that made me interpret the Amazon stuff as a lighthearted ad piece, done as relatively cheap PR; but probably emerging from a broader 'theorizing about new stuff to sell and new ways to sell it' project that usually operates more quietly, and probably also has more mundane, but practical, notions on the burner. A "Bah, here at Ebay we only do incremental modifications based on short-term considerations, sonny!" response is... tone deaf... to say the least.
People don't like those who risk. From where I stand, creative people risk resources, no matter how trivial.
It has to be more visceral than rational risk management. Creative people may have a wider risk/reward spread than others; but so do some financial instruments that even fairly stodgy investment types like just fine (so long as they can be aggregated to moderate a given portfolio's exposure to any one of them). Either people suspect that 'creative people', even as a class, cost more than they are worth, or they are irrationally leaving potential gains on the table.
The mention of Steve Jobs as an "innovator" makes the article suspect. E.G. the author does not know what she is talking about.
It's arguably worse than that: Jobs (and Apple generally) don't really do 'innovative', in the sense that nearly everything they produced had some sort of less-well-refined immediate antecedent elsewhere, or was purchased, or or the like. However, Jobs is quite notable indeed for his willingness to take successful products out and shoot them in order to make room for something new(even when the new thing is still not a safe bet in competition with the older; but cheaper, widely adopted, and widely accepted thing), to tell people who demand backwards-compatible whatever where they can file their futile protests, and other behaviors that, while not innovative in themselves, are more or less required to take an innovation from 'tech demo' to 'product' in a reasonable amount of time. On the other hand, of course, his enthusiasm for ruthless focus would likely have been a very poor fit indeed for a 'blue skies' R&D operation(and indeed, stodgy old Microsoft is the company that has one of those, and seems to carefully avoid applying what it comes up with to anything they actually sell...)
If you want to look at 'innovation' in an institutional context, he isn't a good example of it; but characters like him are clearly relevant to how the broader institutional context interacts with 'creative' or 'innovative' people.
Resistance to new ideas and ways of doing things is also habitual; we also enjoy the comfort of predictability and simplicity because it's less stressful than constantly trying to adapt to new situations. This requires neural plasticity, which decreases with age, hence the difficulty of many elderly folks to "change with the times". Likewise, studies on teens have shown that they are less risk averse than older folks. They tend to rate reward much higher than risk, and each new generation seems to bring with it new ideas and change.
EU agricultural policy strongly reflects French agricultural interests. The tricky bit has been trying to figure out what sort of special-pleading magic can be employed to continue policy designed to subsidize French farmers without getting stuck applying it to the (much poorer; but of less interest to France) farmers in some of the new EU members to the east...
True, though (while truly ghastly results have been substantially reduced by sanitary handling equipment and standardized yeast strains) those are exactly the sort of things frequently made from deeply undistinguished jug or box wines.
On the plus side, they don't seem to have invented wine coolers or 'flavored fortified wines' so they can count themselves blessed on those counts.
That sounds like it would be good for bolstering legionary morale: if the alternative is drinking it, fighting to the death against whatever enemies are available starts to sound substantially more attractive...
Do you add 'honey, mint, cedar, tree resins and cinnamon bark' to your wine because your technique is "sophisticated" or because you are trying to restore some semblance of drinkability to the result of a really dreadful fermentation process?
Why not handle it like OS X's Keychain, where your passphrase unlocks the encrypted secret... while the secret and the data store are on the same device?
The trouble is that you end up storing your secret and your data on the same device as your big, complex, modern OS, your web browser, and all the other neat network connected stuff you may have installed. Anything goes wrong with all that, and it isn't a secret anymore.
If your ciphertext must be stored in such a fashion, why bother? Properly encrypted data should be able to fall into the hands of an attacker, that's the whole point.
Because you want to avoid trusting the computer on which you are entering the password to also handle decryption duties. You do want the encrypted data to be useless without the key; but if you are planning on decrypting the data yourself, your key is going to be living in some computer's memory, at least briefly. If you are using a suitably compromised computer, it won't be a private key for long.
Slashdot Mirror
See the 'or simply emotional majoritarianism' option. Anything the majority (or loud minority that pines for the good ol' days) approves of isn't 'offensive' in the sense used to argue against things; because it is simply right and proper that they exercise their power; but such things are quite frequently calculated to offend those who don't have the power to do anything about it.
Having an implant that can 'patch over' damaged brain areas should make lesion studies more subtle and precise. They can certainly tell us something already, at least in broad strokes, about what functions go where; but it's hard to shake the question of 'if you damage area X, does function Y suffer because area X handles it, or because it depends on connectivity through area X between areas W and Q?' If we have a technique for replacing a functional area with a mere transmission line, that gives us greater ability to differentiate between an area with a functional role in some function and an area with a merely connective role (presumably, there are also areas that are mostly connective; but apply some amount of signal processing between input and output. In the future, maybe we will be able to write arbitrary signal processing filters and patch them in, in software, between the input and the output of this 'bridge' device. That'd be extra neat).
If anything, it's the opposite: Experimentally, we observe a variety of effects as we prod the black box in various ways, examine it with various clever inferential techniques, and so on. In some cases, we are able to develop reliable "If you poke it there, it does that, reliably" type rules based on repeated observation. However, our prodding of the system has not yet provided enough information to posit the underlying structure, that would unify all our disparate observations of patterns. So, we don't have one. We might have some theorists, on a parallel track, tinkering with candidate unified theories; but operationally we only have our pattern inferences.
If it were faith based, it'd be way easier: we could just posit a Grand Unified Theory, and then bodge like crazy to force any observations into agreement with it.
Instead, we start with a whole bunch of tedious empiricism, observational inference of patterns, and hope that eventually somebody comes up with something that will elegantly explain the mess. It's Newton's old "Hypotheses non fingo": He had no idea how or why universal gravitation worked the way it did; but accorded with astronomical and physical observation, so he proposed it, without any metaphysical entanglements, as a model.
The evangelicals really should have remembered the Danbury Baptists...
Exercising the temporal power of fire and sword against your religious enemies is fun and all; but (even when you are on top) tends to be corrupting, and when you aren't, it opens the door to being at the mercy of every different group out there.
Plus, even among people who would ordinarily be inclined to treat your choices of faith as purely personal and let you believe as you will, nothing sours toleration quite like making it clear that you are ready and willing to impose what you believe on everyone else. Suddenly, and wholly because of your actions, your beliefs are now everybody's business; because everybody will suffer for them. That's when the gloves come off (most notably among atheists: 'god-not-existing' is something that isn't even worth mentioning, except that people who believe otherwise keep pushing the matter. In absence of pressure from theists, the nonexistence of god is about as interesting as the nonexistence of Russel's teapot.)
The Jews with true Jew-fu have 613 commandments, though they do overlap with the reduced-commandment-set religions that were derived from them.
"Offensive" is an unhelpful criterion because it's a subjective assessment, and one that varies from person to person. Unless you couple it with a statistical cut-off of some kind, you probably couldn't do anything without offending somebody. Even deciding whether to go with grass or masonry paving for the public space wouldn't be uncontroversial...
"Offensive", in practice, is either meaningless (since everything is, to somebody) or simply emotional majoritarianism (if you only count as 'offensive' things that offend large and influential groups of people). Lousy criterion.
If you think atheists drive evangelical conservatives nuts, you ain't seen nothing yet.
Which is a trifle ironic, because 'satanists' (to the degree that they actually take the stuff seriously, and aren't just into heavy metal and upsetting their parents), are far closer, in terms of opinions on metaphysics, to Christians than atheists are. Especially to some of the protestant outfits that are practically Manichean in their emphasis on the power of satan in the world...
Though, given how much they like Muslims, who are closer still, I suppose that it may be a matter of hating your competitors even more than people in a different industry altogether.
No reason to worry, with our exciting advances in military robot technology, we can have robots solve the unemployment problems that other robots create!
Not to worry! With new FungusLease Managed Robotic Workforce technology, businesses mired in boring old 'making physical objects' industries can experience the joys of The Cloud by replacing an increasing number of aspects of their production process with robots whose hardware is leased from me, and whose software and configuration data are licensed in exchange for monthly fees in perpetuity! Don't worry, the part of the license granting me a perpetual nonexclusive license to all the configuration data and job parameters is, um, just legal boilerplate, not something I'd ever use to sell services to your competitors!
I suspect that one skilled in the art could perform a physical attack. Decapping chips and analyzing both their logical structure and their non-volatile memory contents isn't trivial; but it's a commercially available service with today's technology. TPMs are usually FIPS 140-2 compliant; but tend not to mention whether they are level 1, 2, 3, or 4 devices (which makes one suspect that the answer is unflattering).
The big question would be whether the DRM system has some sort of revocation support (blacklisting on XBL, 'burn' lists on game disks authored after the compromised unit is discovered, etc. The design of AACS is an example of how this sort of thing can be done). If not, it might cost a couple hundred thousand, more or less depending on the difficulty of the chip; but all you'd need is one private key. If there is a revocation mechanism, that makes physical attacks substantially less helpful, since MS(or any other party in the position of DRMed-platform vendor) can force you to pay the price, again and again, every time they catch you.
In my (admittedly layman's) understanding, the difficulty with the XB360/PS3 isn't so much that their cores are all that cryptic (the xbox, and the PS3's main core, are all basic PPCs, and the PS3's 'SPE' elements are weird; but IBM talked a lot about them in the course of trying to build interest in using them as accelerator cards for compute applications); but because they aren't x86; but are clocked as high as contemporary x86s, which makes it difficult to get emulation at anything remotely resembling usable speed, much less support anything that does some fancy timing-dependent tricks.
That's fair, Back when Apple sold "Apple"-line computers, they had a number of neat tricks up their sleeve. Once Woz got the shove, though, they still produced some very refined products(and some messes); but all in the direction of brutally refined and executed implementations of not terribly novel concepts (which isn't a bad thing, especially when the competition mostly has dubiously refined and sloppy implementations; but isn't "innovation").
Didn't Bell labs get spun off in some fashion? I have the vague memory that they are huddled under the Alcatel/Lucent corporate umbrella these days while AT&T finances basic research in contract-obfuscation theory by the nation's top theoretical lawyers...
On the plus side, emulating an AMD x86 and GPU is likely to be considerably easier (especially since AMD's current or near-future PC parts are likely to be extremely similar in most respects, though you will probably have to go up a few speed grades to deal with the emulator running on top of a full OS) than emulating either the relatively fast PPCs of the previous generation (PPC-on-x86 is done; but doing that really fast is another story) or the slow-but-somewhat-esoteric-and-absolutely-every-oddity-was-used-and-abused architectures of the older consoles.
On the minus side, the odds are good that both new consoles (especially the Xbox, given MS's software side; but probably the PS as well) contain a lot of software that, while not integral to the tightly-optimized-graphics-twiddling aspects of the games, will probably have to be given a fairly precise "WINE-like" treatment to avoid breaking things all over the place. Not necessarily impossible (as WINE itself demonstrates); but definitely a different game than the 'emulate the hardware and let the ROM do as it will' emulators that work for older consoles.
On the very minus side, it would not be out of character for either MS or Sony to have added some nasty copy-protection-related cryptographic goodies that will be very hard to emulate. MS, given their PC background, might well have gone for a TPM. Architecturally, emulating one of those would be cake by the standards of what the emulation scene has taken on, except for minor matters like the endorsement key. A TPM emulator that emulates a TPM loaded with the 2048-bit RSA private key of your choice? Sure, no problem. The correct private keys? That might be an issue.
Even if your CEO isn't supposed to be a visionary, and was basically hired to keep the good ship BeanCounter on an even keel, I'm pretty sure that he isn't supposed to make an ass of himself, and make the company sound reactionary and uncreative, in public. That's what I don't understand about the whole thing.
Even if drones are nonsense as a delivery platform, their PR/advertising utility in the 'we ship your shit crazy fast' narrative that Amazon has been trying to build around 'Prime' would seemingly be obvious.
All arguments about whether drones for package delivery are viable or not aside, I am honestly baffled that eBay's CEO would open his mouth on this one...
Even if it's 100%-aw-hell-no-never-going-to-happen, Amazon's work so far has likely been fairly inexpensive and has certainly stirred up as much attention as a decent sized ad campaign (the sort of thing that might actually cost as much or more to produce and buy airtime to run), so it isn't as though they are wallowing in shame and loss right now.
Under those circumstances, what possible benefit is there to a not-terribly-clever rubbishing of the opposition that just makes you look unhip and non-innovative? Especially when that is basically true; direct connection of buyers and sellers worldwide, in an easy-to-use, comparatively safe, framework may have been pretty cool when ebay hit the scene, but they hit the scene quite some time ago and have mostly been ratcheting up the transaction costs since then.
I personally have strong doubts about the viability of drone delivery; but that made me interpret the Amazon stuff as a lighthearted ad piece, done as relatively cheap PR; but probably emerging from a broader 'theorizing about new stuff to sell and new ways to sell it' project that usually operates more quietly, and probably also has more mundane, but practical, notions on the burner. A "Bah, here at Ebay we only do incremental modifications based on short-term considerations, sonny!" response is... tone deaf... to say the least.
People don't like those who risk. From where I stand, creative people risk resources, no matter how trivial.
It has to be more visceral than rational risk management. Creative people may have a wider risk/reward spread than others; but so do some financial instruments that even fairly stodgy investment types like just fine (so long as they can be aggregated to moderate a given portfolio's exposure to any one of them). Either people suspect that 'creative people', even as a class, cost more than they are worth, or they are irrationally leaving potential gains on the table.
The mention of Steve Jobs as an "innovator" makes the article suspect. E.G. the author does not know what she is talking about.
It's arguably worse than that: Jobs (and Apple generally) don't really do 'innovative', in the sense that nearly everything they produced had some sort of less-well-refined immediate antecedent elsewhere, or was purchased, or or the like. However, Jobs is quite notable indeed for his willingness to take successful products out and shoot them in order to make room for something new(even when the new thing is still not a safe bet in competition with the older; but cheaper, widely adopted, and widely accepted thing), to tell people who demand backwards-compatible whatever where they can file their futile protests, and other behaviors that, while not innovative in themselves, are more or less required to take an innovation from 'tech demo' to 'product' in a reasonable amount of time. On the other hand, of course, his enthusiasm for ruthless focus would likely have been a very poor fit indeed for a 'blue skies' R&D operation(and indeed, stodgy old Microsoft is the company that has one of those, and seems to carefully avoid applying what it comes up with to anything they actually sell...)
If you want to look at 'innovation' in an institutional context, he isn't a good example of it; but characters like him are clearly relevant to how the broader institutional context interacts with 'creative' or 'innovative' people.
Resistance to new ideas and ways of doing things is also habitual; we also enjoy the comfort of predictability and simplicity because it's less stressful than constantly trying to adapt to new situations. This requires neural plasticity, which decreases with age, hence the difficulty of many elderly folks to "change with the times". Likewise, studies on teens have shown that they are less risk averse than older folks. They tend to rate reward much higher than risk, and each new generation seems to bring with it new ideas and change.
"Science advances one funeral at a time."
EU agricultural policy strongly reflects French agricultural interests. The tricky bit has been trying to figure out what sort of special-pleading magic can be employed to continue policy designed to subsidize French farmers without getting stuck applying it to the (much poorer; but of less interest to France) farmers in some of the new EU members to the east...
True, though (while truly ghastly results have been substantially reduced by sanitary handling equipment and standardized yeast strains) those are exactly the sort of things frequently made from deeply undistinguished jug or box wines.
On the plus side, they don't seem to have invented wine coolers or 'flavored fortified wines' so they can count themselves blessed on those counts.
That sounds like it would be good for bolstering legionary morale: if the alternative is drinking it, fighting to the death against whatever enemies are available starts to sound substantially more attractive...
Do you add 'honey, mint, cedar, tree resins and cinnamon bark' to your wine because your technique is "sophisticated" or because you are trying to restore some semblance of drinkability to the result of a really dreadful fermentation process?
Why not handle it like OS X's Keychain, where your passphrase unlocks the encrypted secret... while the secret and the data store are on the same device?
The trouble is that you end up storing your secret and your data on the same device as your big, complex, modern OS, your web browser, and all the other neat network connected stuff you may have installed. Anything goes wrong with all that, and it isn't a secret anymore.
If your ciphertext must be stored in such a fashion, why bother? Properly encrypted data should be able to fall into the hands of an attacker, that's the whole point.
Because you want to avoid trusting the computer on which you are entering the password to also handle decryption duties. You do want the encrypted data to be useless without the key; but if you are planning on decrypting the data yourself, your key is going to be living in some computer's memory, at least briefly. If you are using a suitably compromised computer, it won't be a private key for long.