History isn't encouraging, though. They've been pursuing the dream of one windows to rule them all since the days when that involved smearing a crude layer of flayed win95 across winCE and pretending it was a good fit for PDAs.
Now that hardware has advanced they have a much better shot at architectural unification (if memory serves, NT has basically edged out everything else except for whatever CE support they provide for legacy customers); but UI? That won't go well.
The CPU and GPU are fine, reports on battery life seem to vary, with excellent results if you turn the 'dynamic perspectives' thing off and don't firefly much, tepid ones if you actually try to use those features; but the screen is genuinely disappointing for a phone in that price range, although the internal storage is better than usual.
There's nothing offensively wrong with it; but the price tag befits a device that is genuinely compelling in some way, which it isn't.
Unfortunately, while stock android is undeniably Google's little fiefdom, 'FireOS' is a pretty much point-for-point a replacement of hegemonic Google with hegemonic Amazon (somewhat more, arguably, given the 'silk' browser's extensive default reliance on server-side processing).
It is pretty stark how lousy de-googled Android is; but Amazon isn't really here to change the dynamic of effectively closed control of the platform, just who controls it. Getting models with a 'clean' AOSP firmware(typically excluding drivers; because it seems to be blobs all the way down on the mobile side) isn't hard; but using them is fairly grim.
Did anyone *really* expect a crapp to have any sort of security whatsoever?
It's a trifle surprising given that the usual 'eh, let's just wrap our shit mobile website in a UIWebView and call it a day' school of 'app' development would likely have inherited SSL through sheer laziness, while whatever attempt at app development CNN attempted is apparently so dysfunctional as to be markedly worse than the state of website logins in general, and apparently so incoherent that the phone and tablet versions don't share login behavior...
That seems like the sort of thing that takes effort to screw up.
The one complicating factor, though, is that discrete sound cards are being squeeze from both sides: With even integrated GPUs offering HDMI and displayport audio, and even all but the most spartan (usually super-cheap and/or strictly business oriented) onboard audio supporting S/PDIF, the option of an external DAC or receiver becomes much more attractive, especially if you already have one that you like or want to be able to use other audio sources with a relatively expensive piece of high quality audio gear.
The performance of the analog components of onboard audio is, indeed, going to be more 'endurable' than 'good'; but digital logic is crazy cheap and (mostly) either works or doesn't, and basic boring onboard audio often has less ghastly driver mess than the cards trying to 'value add'(Creative, specifically, being a ghastly offender).
It's not as though they are a ghastly scam or anything, if you want to be able to plug a nice pair of headphones straight into your PC rather than an outboard module that's totally understandable; but they do occupy a slightly precarious middle ground between mostly-competent onboard audio and the full array of audio gear that accepts digital input.
Apropos of that... With at least certain Realtek chipsets(I'm afraid I don't have the model number handy, this was a couple of years ago) I ran into an issue with audio output from a program that made atypically demanding use of MIDI:
The audio worked, and was free of obvious noise problems; but the pitch and playback speed kept changing, sometimes correct, sometimes badly off. With a bit of fiddling, it turned out that putting the system under heavy load made it work properly, and the deviations only cropped up when it was lightly loaded. If CPU power saving was disabled(and so clock speed kept constant) the problem never occurred regardless of load. As best I was able to tell, some part of the sound system was using the CPU (instead of all those fancy system timers that were added because using the CPU clock is something best left to the bad old days of Turbo buttons) as a timebase; but not accounting for the fact that it only actually ran at the maximum frequency when load demanded it.
It was an entertaining bug, as they go; but not confidence inspiring. The $10 USB thing that replaced it had no such issues.
The/. writeup sounds like audiophile wank to me. I would be surprised if this Soundblaster could justify its price in a proper double blind study on real world data (music, games, movies, etc...) vs. the built in audio on your mobo.
It doesn't help that the squeeze is really coming from both sides: On the low end, the performance of onboard audio has improved(SNR may still make the golden-eared cringe; but horrors of the old days are mostly banished, so adequate performance, usually with fairly well behaved default drivers, is rarely a problem).
On the high end, odds are good that the user already has a preferred DAC and amplifier which will skip the cheap and electrically noisy PC entirely. Even fairly nasty onboard sound often has digital out, and with HDMI and displayport including audio support, so do most graphics cards, even integrated GPUs.
Unless you are trying to drive a touchy and analog only device, maybe a nice pair of headphones or an older amplifier or receiver, there just isn't an obvious need for what creative is selling.
A phone with fairly tepid specs being sold for a flagship price (and AT&T SIM-locked, only) supported by one largely useless gimmick and a dedicated 'buy stuff on amazon' button.
They probably have some amount of TRS stuff quietly rotting in a back room somewhere; but the idea that they actually care about those services seems implausible at best.
Verizon does do dedicated lines of various flavors, if you pay them enough; but that's more or less irrelevant to the duel over how finely commodity ISP customers can be diced up and double billed. Nor could one seriously imagine even the most grandiose promises of fast-lanes actually making life-critical applications over cheapy links seem like a good idea.
I'm not surprised, alleging that the telegenic interests of assorted groups just so happen to be aligned with your bottom line is an old strategy; but this is pretty incoherent even by the low standards of the genre.
Yes, if there were a fast lane, one could theoretically put special-deaf-packets in it (or just as easily shove them into the slow lane, if they can't afford to pay); but this ignores the more pressing question of "What, pray tell, is currently suffering for want of special bandwidth and how demanding must it be if your existing service can't cope?".
I can imagine that certain disabilities might drive modestly higher bandwidth demands (the deaf, presumably, don't get much use out of VOIP, which is lower bandwidth than video good enough to make lip reading or signing an option; but last I checked uploading and downloading video wasn't exactly a niche case, even if it is one where Verizon can't seem to get Netflix working...); but nothing that exceeds the current or near-term demands of most internet users.
They obviously won't prefer this interpretation; but just how awful is Verizon planning to make the non-fast lane if these special disabled services will need to be fast-laned to work? Anyone?
How do you defend yourself against accusations like that as a man? We are extremely sensitive to being criticized by women, can you really say thats not true without becoming another "point of proof" that they have?
Well, the most obvious step is to distinguish between "That's not true of me" and "That's not true". The first statement(while not always accurate) is much easier to confirm or deny. Plus, you aren't immediately put in the position of having to 'win' the debate in order to lay out your own position. If you immediately conflate population-level complaints with personal complaints, you end up taking on a markedly larger and more challenging position.
It may also be true that you suspect the harassment to be the work of a vocal and dedicated minority(and it would actually be rather interesting to see what the logs say about troll distribution in various internet locations) rather than a general thing; but you still gain nothing by tying the desire to defend yourself with the desire to defend a population.
Wiring costs can add up as well. At low voltages the amount of copper or aluminum you have to throw at the problem to get resistive losses down can be pretty significant.
Every time I hear about a terrifyingly invasive means of "improving performance" its targeted at developers. Is it just selection bias, or does the world actually hate us?
Mostly because they are a newer profession and a trickier one to quantify.
Time and motion studies, along with 'scientific management' were already a serious hit in terrifyingly invasive performance enhancement for blue collar labor around the turn of the 20th century(Taylor and the Gilbreths being the poster children, with many successors). The workers who haven't been replaced by robots yet are likely still subject to a descendant of it. Though less amenable to automation, service sector jobs are also rationalized more or less as tightly as available technique allows.
Software development is still a work in progress because it only started existing comparatively recently and because it takes more technology to dismiss any "Oh, what we do here is unquantifiable skilled craftsmanship" positions.
It is selection bias, in that you apparently haven't heard of it happening to basically everyone it can reach; but the world does actually hate you, and is actively working on making software development absolutely as soul crushing as seems economically desirable.
So, a canvas randomizer is needed, isn't it? Or a means to get many, many machines to all appear identical.
Unfortunately, since this technique is almost certainly being used alongside a suite of others, it's tricky to know what tactic is most privacy-maximizing. Canvas randomization would ensure that your browser's canvas fingerprint does not remain stable; but if the attacker is able to determine that you are randomizing(by making multiple runs, possibly even from different domains, that ought to be identical but won't be if your canvas is randomized), that may also be a behavior distinctive enough to be useful.
Depending on what you mean by 'block', there may or may not be a properly satisfactory answer:
'Block' as in 'make this specific mechanism fail' is the relatively easy question. If the attacker can't manipulate a canvas element and read the result, it won't work. So the usual javascript blockers or more selective breaking of some or all of the canvas element (the TOR browser apparently already does this for methods that can be used to read back the contents of a canvas element, so you can still draw on one but not observe your handiwork) will do the job.
Unfortunately the attacker doesn't actually care about making your browser draw a picture, they care about achieving as accurate a UID as they can. Given that, you might actually make yourself more distinctive if your attempt to break a given fingerprinting mechanism succeeds. In the case of the TOR browser, for instance, attempts to read a canvas will always be handled as though the canvas is all opaque white. This does prevent the attacker from learning anything useful about font rendering peculiarities or other quirks of your environment's canvas implementation; but it's also a behavior that, for the moment at least, only the TOR browser has. Relatively uncommon. Possibly less common than the result that you'd receive from an unmodified browser.
That's the nasty thing about fingerprinting attacks. Fabricating or refusing to return many types of identifying information is relatively easy (at least once you know that attackers are looking for them); but unless you lie carefully, your fake data may actually be less common (and thus more trackable) than your real data.
This is (largely) true; but the question is why?. It is expected that cheap phones will suffer from somewhat inferior hardware; but it is less clear why they should suffer from inferior software, doubly so if the very same vendor or the AOSP has software without whatever flavor of broken is causing the issue. It's also particularly weird with something like autocorrect making dumb mistakes: that's far too high level to be a 'well, we went with the cheapest SoC vendor, and you wouldn't believe what total shit their BSP is...' problem, it's not something that the guy buying the expensive phone is going to be spared because he has a faster CPU and more RAM, and it's not something where there's any good reason for the vendor to be trying to roll their own.
I suspect that the thesis about 'hard to quantify' stuff getting squeezed first is true, and one would be foolish to expect market mechanisms to work in the absence of good information, which 'hard to quantify' largely assures; but it still surprises me that cheap hardware (and even some expensive hardware) is routinely shipped with software that actually cost somebody money to make worse than 'stock'. Carrier shitware on cheap phones, I understand, because carriers exert most of the control over what phones will be made available 'free' with contract, and so OEMs will suck it up and preinstall whatever they demand; but any other area where the experience is worse than stock android of the equivalent version just seems weird.
If you don't mind looking ridiculous, the helicopter market has had this for ages (since there's nothing quite like sitting under a propeller going fast enough to keep you in the air when it comes to noise...) Nice, sturdy, over-the-ear headphones with substantial protection from outside noise, along with a mic which gets piped to everyone else's headphones so they can hear you as though you were speaking in a more normal environment(the ability to mute individual users would, of course, be vital in broad application).
This is an airport we are talking about: the kiosk AIs have more rights than the passengers interacting with them. Rates of assault will likely be low.
The trouble with progress as a cure for stress is twofold:
One, expectations tend to grow as fast, sometimes faster, than capabilities. Unless you are traveling without any connecting flights and on a very leisurely schedule, everyone's assumptions about where you'll be and when will be calibrated to 'your flight; but on time', so delays that would have faded into the noise historically will now throw you off.
Perhaps more fundamentally it appears to be the powerlessness rather than the absolute time that stresses people out, and being at the mercy of complex systems run by other people is beautifully designed to rub your face in powerlessness. Technology has, of course, increased our absolute level of power by chiseling away at the domain of 'nope, go try placating the spirits or something'; but all those places where it used to be that nobody had any control, now somebody; but not you, has control and you can't quite shake the impression that they are jerking you around.
Aside from all that, it isn't clear why adding a shallow emulation of a talking human head is even going to improve the terminal experience:
If you are dealing with a routine matter, you aren't really trying to convey that much data (and none of the data you are trying to convey are subtly emotionally nuanced or anything, it's basically an "I want to be on this flight, ideally in this seat, here's the billing info" operation, not a sonnet) and existing text and graphic based interfaces, while often questionably thought out, are at least as competent as a natural-language dialog for anyone who isn't illiterate or otherwise handicapped.
If something or someone is fucked up and/or deeply confused, the computer won't be able to help you because it will just format and present the garbage you are trying to sort out. You need someone who can understand an edge case or error and has the power to give a good hard shove to whatever fields aren't cooperating.
I'd bet nontrivial money that the effect of this 'advance' will be to make the experience worse: The licensing fees will be calibrated to be lower than human salaries; but the underlying system will still be far dumber and less flexible than the humans who it will replace (because why do we need so many desk staff now that our kiosks are so user friendly!?); so users whose problems were already solved will have, at best, a slightly more pleasant interaction, and the users with real problems will have to wait in a longer line for a more harried human to fix it.
We all know how totally peachy-keen 'interactive voice recognition' systems have made interacting with call centers, and this is basically the same old shit with an animated face.
The high-heritability hack only works on sexual reproduction; but horizontal gene transfer mechanisms do not.
The heritability hack wouldn't directly cause more horizontal transfers than usual; but it would ensure that the introduced gene spreads quickly through the target population(increasing the odds that a gene transfer event from that population will include the gene in question) and if it is successfully transferred, it will be more likely than usual (if the transfer target reproduces sexually) to spread into the new host species rather than dying with the individual who received the transfer, increasing the odds that a single gene transfer event would end in a population-level change, rather than just one genome that will be taking a dirt nap soon enough.
I am not impressed with the state of coal fired emissions regulation (sulfur compounds are down; but fly ash certainly isn't something that cures what ails you, and the general 'Eh, old stuff just gets grandfathered because we can't fight the incumbents' model of regulation is broken); but your snarking about the poor reactors being treated as unnatural is rather flawed.
The further your coal gets from being pure carbon, the more dire some of the potential aerosolized-and-spread-hither-and-yon materials are; but the process is just conventional chemistry, you aren't going to emit anything you didn't dig up(except the added oxygen). A nuclear reactor; shockingly enough, is not subject to this limitation, and fairly aggressively shoves assorted fissionables down the decay chain.
Aside from the one (known) incident at Oklo, the crust isn't seeing much in the way of activity above background decay rates, and it follows that anything with a short half life is going to be extremely scarce. Something that's been dug up, concentrated, and carefully stewed in its own neutrons, by contrast, will have a very different collection of isotopes, some remarkably scarce anywhere else.
This doesn't mean that coal power is good for you, or restricted in what it contributes to our air supply; because that is very unlikely; but it's just silly to pretend that reactor products are isotopically similar to what you'll find in the ground; the 'power' in 'nuclear power' is only there because they aren't.
Slashdot Mirror
But where will the US get its supply of mutant kangaroos?
History isn't encouraging, though. They've been pursuing the dream of one windows to rule them all since the days when that involved smearing a crude layer of flayed win95 across winCE and pretending it was a good fit for PDAs.
Now that hardware has advanced they have a much better shot at architectural unification (if memory serves, NT has basically edged out everything else except for whatever CE support they provide for legacy customers); but UI? That won't go well.
The CPU and GPU are fine, reports on battery life seem to vary, with excellent results if you turn the 'dynamic perspectives' thing off and don't firefly much, tepid ones if you actually try to use those features; but the screen is genuinely disappointing for a phone in that price range, although the internal storage is better than usual.
There's nothing offensively wrong with it; but the price tag befits a device that is genuinely compelling in some way, which it isn't.
Unfortunately, while stock android is undeniably Google's little fiefdom, 'FireOS' is a pretty much point-for-point a replacement of hegemonic Google with hegemonic Amazon (somewhat more, arguably, given the 'silk' browser's extensive default reliance on server-side processing).
It is pretty stark how lousy de-googled Android is; but Amazon isn't really here to change the dynamic of effectively closed control of the platform, just who controls it. Getting models with a 'clean' AOSP firmware(typically excluding drivers; because it seems to be blobs all the way down on the mobile side) isn't hard; but using them is fairly grim.
Did anyone *really* expect a crapp to have any sort of security whatsoever?
It's a trifle surprising given that the usual 'eh, let's just wrap our shit mobile website in a UIWebView and call it a day' school of 'app' development would likely have inherited SSL through sheer laziness, while whatever attempt at app development CNN attempted is apparently so dysfunctional as to be markedly worse than the state of website logins in general, and apparently so incoherent that the phone and tablet versions don't share login behavior...
That seems like the sort of thing that takes effort to screw up.
The one complicating factor, though, is that discrete sound cards are being squeeze from both sides: With even integrated GPUs offering HDMI and displayport audio, and even all but the most spartan (usually super-cheap and/or strictly business oriented) onboard audio supporting S/PDIF, the option of an external DAC or receiver becomes much more attractive, especially if you already have one that you like or want to be able to use other audio sources with a relatively expensive piece of high quality audio gear.
The performance of the analog components of onboard audio is, indeed, going to be more 'endurable' than 'good'; but digital logic is crazy cheap and (mostly) either works or doesn't, and basic boring onboard audio often has less ghastly driver mess than the cards trying to 'value add'(Creative, specifically, being a ghastly offender).
It's not as though they are a ghastly scam or anything, if you want to be able to plug a nice pair of headphones straight into your PC rather than an outboard module that's totally understandable; but they do occupy a slightly precarious middle ground between mostly-competent onboard audio and the full array of audio gear that accepts digital input.
Apropos of that... With at least certain Realtek chipsets(I'm afraid I don't have the model number handy, this was a couple of years ago) I ran into an issue with audio output from a program that made atypically demanding use of MIDI:
The audio worked, and was free of obvious noise problems; but the pitch and playback speed kept changing, sometimes correct, sometimes badly off. With a bit of fiddling, it turned out that putting the system under heavy load made it work properly, and the deviations only cropped up when it was lightly loaded. If CPU power saving was disabled(and so clock speed kept constant) the problem never occurred regardless of load. As best I was able to tell, some part of the sound system was using the CPU (instead of all those fancy system timers that were added because using the CPU clock is something best left to the bad old days of Turbo buttons) as a timebase; but not accounting for the fact that it only actually ran at the maximum frequency when load demanded it.
It was an entertaining bug, as they go; but not confidence inspiring. The $10 USB thing that replaced it had no such issues.
The /. writeup sounds like audiophile wank to me. I would be surprised if this Soundblaster could justify its price in a proper double blind study on real world data (music, games, movies, etc...) vs. the built in audio on your mobo.
It doesn't help that the squeeze is really coming from both sides: On the low end, the performance of onboard audio has improved(SNR may still make the golden-eared cringe; but horrors of the old days are mostly banished, so adequate performance, usually with fairly well behaved default drivers, is rarely a problem).
On the high end, odds are good that the user already has a preferred DAC and amplifier which will skip the cheap and electrically noisy PC entirely. Even fairly nasty onboard sound often has digital out, and with HDMI and displayport including audio support, so do most graphics cards, even integrated GPUs.
Unless you are trying to drive a touchy and analog only device, maybe a nice pair of headphones or an older amplifier or receiver, there just isn't an obvious need for what creative is selling.
A phone with fairly tepid specs being sold for a flagship price (and AT&T SIM-locked, only) supported by one largely useless gimmick and a dedicated 'buy stuff on amazon' button.
Where do I sign?
They probably have some amount of TRS stuff quietly rotting in a back room somewhere; but the idea that they actually care about those services seems implausible at best.
Verizon does do dedicated lines of various flavors, if you pay them enough; but that's more or less irrelevant to the duel over how finely commodity ISP customers can be diced up and double billed. Nor could one seriously imagine even the most grandiose promises of fast-lanes actually making life-critical applications over cheapy links seem like a good idea.
I'm not surprised, alleging that the telegenic interests of assorted groups just so happen to be aligned with your bottom line is an old strategy; but this is pretty incoherent even by the low standards of the genre.
Yes, if there were a fast lane, one could theoretically put special-deaf-packets in it (or just as easily shove them into the slow lane, if they can't afford to pay); but this ignores the more pressing question of "What, pray tell, is currently suffering for want of special bandwidth and how demanding must it be if your existing service can't cope?".
I can imagine that certain disabilities might drive modestly higher bandwidth demands (the deaf, presumably, don't get much use out of VOIP, which is lower bandwidth than video good enough to make lip reading or signing an option; but last I checked uploading and downloading video wasn't exactly a niche case, even if it is one where Verizon can't seem to get Netflix working...); but nothing that exceeds the current or near-term demands of most internet users.
They obviously won't prefer this interpretation; but just how awful is Verizon planning to make the non-fast lane if these special disabled services will need to be fast-laned to work? Anyone?
How do you defend yourself against accusations like that as a man? We are extremely sensitive to being criticized by women, can you really say thats not true without becoming another "point of proof" that they have?
Well, the most obvious step is to distinguish between "That's not true of me" and "That's not true". The first statement(while not always accurate) is much easier to confirm or deny. Plus, you aren't immediately put in the position of having to 'win' the debate in order to lay out your own position. If you immediately conflate population-level complaints with personal complaints, you end up taking on a markedly larger and more challenging position.
It may also be true that you suspect the harassment to be the work of a vocal and dedicated minority(and it would actually be rather interesting to see what the logs say about troll distribution in various internet locations) rather than a general thing; but you still gain nothing by tying the desire to defend yourself with the desire to defend a population.
As demonstrated by our victory in the war on drugs, clearly we can just step up enforcement to solve the problem!
Wiring costs can add up as well. At low voltages the amount of copper or aluminum you have to throw at the problem to get resistive losses down can be pretty significant.
Every time I hear about a terrifyingly invasive means of "improving performance" its targeted at developers. Is it just selection bias, or does the world actually hate us?
Mostly because they are a newer profession and a trickier one to quantify.
Time and motion studies, along with 'scientific management' were already a serious hit in terrifyingly invasive performance enhancement for blue collar labor around the turn of the 20th century(Taylor and the Gilbreths being the poster children, with many successors). The workers who haven't been replaced by robots yet are likely still subject to a descendant of it. Though less amenable to automation, service sector jobs are also rationalized more or less as tightly as available technique allows.
Software development is still a work in progress because it only started existing comparatively recently and because it takes more technology to dismiss any "Oh, what we do here is unquantifiable skilled craftsmanship" positions.
It is selection bias, in that you apparently haven't heard of it happening to basically everyone it can reach; but the world does actually hate you, and is actively working on making software development absolutely as soul crushing as seems economically desirable.
So, a canvas randomizer is needed, isn't it? Or a means to get many, many machines to all appear identical.
Unfortunately, since this technique is almost certainly being used alongside a suite of others, it's tricky to know what tactic is most privacy-maximizing. Canvas randomization would ensure that your browser's canvas fingerprint does not remain stable; but if the attacker is able to determine that you are randomizing(by making multiple runs, possibly even from different domains, that ought to be identical but won't be if your canvas is randomized), that may also be a behavior distinctive enough to be useful.
Depending on what you mean by 'block', there may or may not be a properly satisfactory answer:
'Block' as in 'make this specific mechanism fail' is the relatively easy question. If the attacker can't manipulate a canvas element and read the result, it won't work. So the usual javascript blockers or more selective breaking of some or all of the canvas element (the TOR browser apparently already does this for methods that can be used to read back the contents of a canvas element, so you can still draw on one but not observe your handiwork) will do the job.
Unfortunately the attacker doesn't actually care about making your browser draw a picture, they care about achieving as accurate a UID as they can. Given that, you might actually make yourself more distinctive if your attempt to break a given fingerprinting mechanism succeeds. In the case of the TOR browser, for instance, attempts to read a canvas will always be handled as though the canvas is all opaque white. This does prevent the attacker from learning anything useful about font rendering peculiarities or other quirks of your environment's canvas implementation; but it's also a behavior that, for the moment at least, only the TOR browser has. Relatively uncommon. Possibly less common than the result that you'd receive from an unmodified browser.
That's the nasty thing about fingerprinting attacks. Fabricating or refusing to return many types of identifying information is relatively easy (at least once you know that attackers are looking for them); but unless you lie carefully, your fake data may actually be less common (and thus more trackable) than your real data.
This is (largely) true; but the question is why?. It is expected that cheap phones will suffer from somewhat inferior hardware; but it is less clear why they should suffer from inferior software, doubly so if the very same vendor or the AOSP has software without whatever flavor of broken is causing the issue. It's also particularly weird with something like autocorrect making dumb mistakes: that's far too high level to be a 'well, we went with the cheapest SoC vendor, and you wouldn't believe what total shit their BSP is...' problem, it's not something that the guy buying the expensive phone is going to be spared because he has a faster CPU and more RAM, and it's not something where there's any good reason for the vendor to be trying to roll their own.
I suspect that the thesis about 'hard to quantify' stuff getting squeezed first is true, and one would be foolish to expect market mechanisms to work in the absence of good information, which 'hard to quantify' largely assures; but it still surprises me that cheap hardware (and even some expensive hardware) is routinely shipped with software that actually cost somebody money to make worse than 'stock'. Carrier shitware on cheap phones, I understand, because carriers exert most of the control over what phones will be made available 'free' with contract, and so OEMs will suck it up and preinstall whatever they demand; but any other area where the experience is worse than stock android of the equivalent version just seems weird.
If you don't mind looking ridiculous, the helicopter market has had this for ages (since there's nothing quite like sitting under a propeller going fast enough to keep you in the air when it comes to noise...) Nice, sturdy, over-the-ear headphones with substantial protection from outside noise, along with a mic which gets piped to everyone else's headphones so they can hear you as though you were speaking in a more normal environment(the ability to mute individual users would, of course, be vital in broad application).
This is an airport we are talking about: the kiosk AIs have more rights than the passengers interacting with them. Rates of assault will likely be low.
The trouble with progress as a cure for stress is twofold:
One, expectations tend to grow as fast, sometimes faster, than capabilities. Unless you are traveling without any connecting flights and on a very leisurely schedule, everyone's assumptions about where you'll be and when will be calibrated to 'your flight; but on time', so delays that would have faded into the noise historically will now throw you off.
Perhaps more fundamentally it appears to be the powerlessness rather than the absolute time that stresses people out, and being at the mercy of complex systems run by other people is beautifully designed to rub your face in powerlessness. Technology has, of course, increased our absolute level of power by chiseling away at the domain of 'nope, go try placating the spirits or something'; but all those places where it used to be that nobody had any control, now somebody; but not you, has control and you can't quite shake the impression that they are jerking you around.
Aside from all that, it isn't clear why adding a shallow emulation of a talking human head is even going to improve the terminal experience:
If you are dealing with a routine matter, you aren't really trying to convey that much data (and none of the data you are trying to convey are subtly emotionally nuanced or anything, it's basically an "I want to be on this flight, ideally in this seat, here's the billing info" operation, not a sonnet) and existing text and graphic based interfaces, while often questionably thought out, are at least as competent as a natural-language dialog for anyone who isn't illiterate or otherwise handicapped.
If something or someone is fucked up and/or deeply confused, the computer won't be able to help you because it will just format and present the garbage you are trying to sort out. You need someone who can understand an edge case or error and has the power to give a good hard shove to whatever fields aren't cooperating.
I'd bet nontrivial money that the effect of this 'advance' will be to make the experience worse: The licensing fees will be calibrated to be lower than human salaries; but the underlying system will still be far dumber and less flexible than the humans who it will replace (because why do we need so many desk staff now that our kiosks are so user friendly!?); so users whose problems were already solved will have, at best, a slightly more pleasant interaction, and the users with real problems will have to wait in a longer line for a more harried human to fix it.
We all know how totally peachy-keen 'interactive voice recognition' systems have made interacting with call centers, and this is basically the same old shit with an animated face.
The high-heritability hack only works on sexual reproduction; but horizontal gene transfer mechanisms do not.
The heritability hack wouldn't directly cause more horizontal transfers than usual; but it would ensure that the introduced gene spreads quickly through the target population(increasing the odds that a gene transfer event from that population will include the gene in question) and if it is successfully transferred, it will be more likely than usual (if the transfer target reproduces sexually) to spread into the new host species rather than dying with the individual who received the transfer, increasing the odds that a single gene transfer event would end in a population-level change, rather than just one genome that will be taking a dirt nap soon enough.
I am not impressed with the state of coal fired emissions regulation (sulfur compounds are down; but fly ash certainly isn't something that cures what ails you, and the general 'Eh, old stuff just gets grandfathered because we can't fight the incumbents' model of regulation is broken); but your snarking about the poor reactors being treated as unnatural is rather flawed.
The further your coal gets from being pure carbon, the more dire some of the potential aerosolized-and-spread-hither-and-yon materials are; but the process is just conventional chemistry, you aren't going to emit anything you didn't dig up(except the added oxygen). A nuclear reactor; shockingly enough, is not subject to this limitation, and fairly aggressively shoves assorted fissionables down the decay chain.
Aside from the one (known) incident at Oklo, the crust isn't seeing much in the way of activity above background decay rates, and it follows that anything with a short half life is going to be extremely scarce. Something that's been dug up, concentrated, and carefully stewed in its own neutrons, by contrast, will have a very different collection of isotopes, some remarkably scarce anywhere else.
This doesn't mean that coal power is good for you, or restricted in what it contributes to our air supply; because that is very unlikely; but it's just silly to pretend that reactor products are isotopically similar to what you'll find in the ground; the 'power' in 'nuclear power' is only there because they aren't.