My impression is that you're not very familiar with what the law mandates. What you'd need to do to be compliant and offer the functionality you mention, is something a one-person company could handle with no issue. Not to mention that odds are a small business like that would buy an off the shelf webstore, which would include the required functionality and the documentation of it.
It doesn't take much to be compliant. Roughly speaking: informing the user, having them opt-in, letting the know what data is collected, and to delete it if they should so desire.
It is definitely good. A Mom and Pop shop in the states selling homemade soap can't afford to have a DPO
Good thing they wouldn't need one, then. There are criteria for when you'd need one (e.g. your business is mass storage or processing of personal data), and the odds of a tiny shop meeting any of them would be extremely slim. Heck, we're a multinational company and we don't need one. For that matter, there's no requirement to _hire_ someone, it's a role that could be assigned to any employee with sufficient knowledge of privacy laws and best practice.
if an EU resident visits a site, the site has to comply.
Not quite. If your site collects personal data about a EU resident, the site has to comply. If your site does not collect personal data, GDPR does not apply.
access to people's email inboxes in order to data-mine the contents for competitive intelligence -- and controversially flog the gleaned commercial insights to the likes of Uber
It's almost as if that's exactly the sort of undisclosed behavior the GDPR is designed to combat...
Granted, I suppose my subject is a bit unfair. If violating privacy is your primary business model, I guess "can't" is technically accurate.
Hm, having written that it occurs to me that we're talking about events here, which probably means security searches, which makes ticket processing a negligible part of the time consumption.
Expediency. If, say, one in a hundred gets denied, that means 99 people got to just walk through instead of holding up the line while being checked, and more will keep going through while that one person is off to the side getting manually processed. Also, if going by that figure, reducing the number of people needing manual processing by 99%, means fewer staff needed to do just that.
Aye. Considering Tesla's autopilot is little more than a slightly amped up version of lane keeping assist and adaptive cruise control, I'd say he was making a pretty good application for the Darwin awards.
I can't answer for anyone else but, while I have no interest in running a full Linux desktop at this time, I do like some of the tools available in the Linux terminal.
I did, in the past, try to run Linux as my main desktop, it failed miserably (don't get me started). I tried booting into it for the few tools that I preferred on the platform, it proved to be much too cumbersome to be worth it. Getting easy access to them from within Windows in a fashion that is not Cygwin, is all good in my book.
So, to answer your question, I guess WSL exists for people like me.
Steam publishing aggregate information would not be an issue from a GDPR perspective, as long as the source data is compliant (e.g. that it's collected because it is necessary for providing the service, and that the user has given consent).
Steam doesn't, though, so this effectively means the end of the type of information SteamSpy provides.
In that case, I'd be inclined to agree with those who feel this change is mainly to lock down easy access to the sort of aggregated information Steam Spy provides. If it was out of privacy concerns, it would make sense for a lot more to be hidden by default. In particular friend list and group memberships.
It's not only a good idea, in the EU it will be law then the GDPR goes into effect in May. If Valve did not make changes, they'd be liable for some pretty hefty fines for violating the regulation. They couldn't even just hide some legalese blurb way down in an EULA, when it comes to personal data it must be opt-in (and the opt-in must be a user action; a pre-checked consent checkbox is not sufficient), and information must be written in "clear and plain language".
You illustrate the very point I was making, that in the US you don't need any reason to own a gun, and you don't need to show that you are responsible enough to do so. If you don't think that's a problem, that's fine. You can't vote in my country, so it has no effect on me. For the sake of the people of the US, I just hope that you one day will be in the minority. There seems to finally be some movement in that direction, so there's hope.
That does not seem to be true. The specifics depend on your definition of "gun", but if assuming you mean handgun, then google suggests you can definitely own that in Australia. Not in the US "I need this... because.. I just do, ok, gief all the guns!" fashion, but as an active member of a club you can apply for a permit. If "gun" includes rifles and shotguns, then you can own those in England too, as far as I can tell.
From where I sit, regulation seems to be the crucial factor, with bans applying only to categories of firearms. Along with regulation you seem to get a culture where guns are treated with proper respect as opposed to as throwaway items anyone can possess without question or training and toss loaded into their nightstand drawers.
I don't believe there's a single solution that will work for every nation, there are so many factors at play, but it seems fairly self-evident that a good start is to do something different from what the US has been doing.
The worst I've experienced was driving through really slushy winter weather last spring. Wet snow splashed up and coated the car. The parking sensors got obstructed and the car whined endlessly about it, but the cruise control worked fine. In that case, I used it to test if it would work, not because I was going to trust it under those conditions.
I drove my previous car for 8 or so years without any electronics issues, I don't expect this to be any different myself. It's one reason I buy Toyotas. They're boring and overpriced for the equipment level, but they tend to just keep on going.
I don't really worry about such a scenario myself. If I drive in bad weather, I keep the cruise control distance set to maximum (if I use it at all, it depends on the conditions). If the sensor should fail in a way that makes it no longer see the car in front and thus increase speed, the acceleration would be obvious to me long before it became a danger. Same thing would be true if I caught up to a car, I'd notice the cruise control failing to slow down (and the dash not indicating the car in front being there) long before any danger.
Having said that, I'm pretty sure all these sensors are able to detect that they are obstructed and will disengage. To get in an accident because of this would require a triple event. Sensor being obstructed, its ability to detect that it is obstructed failing, and the driver not paying attention for a fair amount of time. Come to think of it, if the collision prevention system uses the parking sensors and/or the camera, there's even an extra layer of fault prevention there. So I relax, but keep my eyes on the road.
If there's a general consensus in the English speaking world that autopilot is synonymous with autonomous, I agree the name was poorly chosen. That's not my impression of the general understanding of the word – we expect a pilot to remain in the cockpit and alert when the plane we're on is on autopilot, after all – but if the data shows otherwise then I'd be the first to argue for the feature to have its name changed.
I guess our brains are very different. I find it really relaxing to just pay attention to the road and steering, and let the car deal with the speed. Since I'm obviously watching the road anyway, it takes away some (literal) footwork without adding any additional workload.
If it was unreliable, it would just add extra anxiety of having to quickly correct for it all the time. But in the year I've had the car, it's been rock solid, even in the worst of slushy winter conditions.
On really winding roads, I do tend to drive manually, though. The desired speed varies too much between straights and corners for cruise control to be convenient, and even if there is a car in front for the cruise control to speed match, that car momentarily disappears out of sight around sharper corners. But for regular highway travel, I've found it to be bloody awesome and never want a car without it again. I like it _that_ much.
Then putting Autopilot in a vehicle is illogical. You don't put something in a vehicle to steer for drivers while totally failing at relieving any kind of duty of driving. Eventually they will get sidetracked, it's just human.
It makes perfect sense to me, along the lines of adaptive cruise control. It makes for a more relaxing drive in that I don't have to actively maintain speed and distance myself, but it does not relieve me from the need to pay attention in order to be able to intervene should the need arise.
Additionally, if I should have a momentary lapse of attention at an inopportune time, odds are that it does not happen at the exact moment my car fails to notice that the car in front slowed down, so it adds safety.
For something like autopilot, it's very common knowledge that it's pretty decent on some types of roads (larger roads with visible painted lines), yet prone to getting it wrong in a lot of situations (e.g. something as simple as a deceleration lane veering off from the lane you are in can make it follow the rightmost line). In its current iteration it should be treated as a souped up combination of lane departure alert and adaptive cruise control.
In other words, autopilot _will_ get you in an accident if you don't pay attention, but in the vast majority of cases the driver will spot where autopilot is likely to have an issue long in advance of the car getting there.
The main issue, to me, seems to be that a lot of people fail to understand the distinction between autopilot (souped up cruise control) and an autonomous (entirely self-driving) vehicle. Perhaps there should be a quiz or something that you have to pass before you are allowed to enable it the first time...
This has become somewhat of a pet peeve of mine. In the vast majority of cases where I see AI used today, it seems to me that the proper term is really "machine learning". According to the dictionary, machine learning is a branch of AI. Sure, I'll grant it that. But that's not what the general public thinks of when AI pops up in articles. If "we fear AI", it's the Ex Machina kind, not the "Google Photos can recognize some types of objects in an image" kind.
Whenever I see e.g. the Google Assistant referred to as AI, I can but roll my eyes. Sure, I suppose it's technically correct in that there's machine learning behind the speech to text it uses to figure out what you're saying. But that's not what people are implying when they call it an AI. Anyone technically inclined who's ever used one of those things quickly realized that what happens next is no more advanced than a pretty limited number of IF statements.
Basically, we're currently using the term AI for everything from very basic machine learning, to sapience, where the general public tends to interpret the term more towards the latter than the former. I don't think I like that, and it turns out I'm currently bored enough to write about it. Sorry I just wasted seconds of your life by reading this rant.:p
1919, not 1917, but I'm not sure I see your point. Did you reply to the wrong post? I wrote that we have decades worth of confirmation. I considered going with "about a hundred years", but considering gravitational lensing took longer to observe, I decided to avoid a possible pedant response pointing that out. I guess there's just no winning the internet.:p
You can't call it science when you use scientific theories to make discoveries? That seems to me to be what you're saying.
E.g. if we take Einstein's hypothesis for how light should behave near massive objects, and the predictions this provides, and confirm those, we can't call it science if we then use the resultant theory to make observations/discoveries?
Case in point, that mass bends light isn't some wild speculation, but a fact of reality with decades worth of confirmation at this point. There's a reason we call it the theory of general relativity, not the hypothesis. Using microlensing to make discoveries is thus perfectly valid science. Now, media has a tendency to ignore any kind of reservation that the actual scientists make. In this case, the obtained data seems open to interpretation. So, interesting and exciting, but not sufficient to actually announce a discovery. But claiming it's not science strikes me as ludicrous.
My point was that it did not answer the question. We know nothing of the woman's connectivity, or whether she cares about privacy issues. The "answer" offered no actual suggestions for solutions, just some related personal opinions phrased as absolutes.
Slashdot Mirror
My impression is that you're not very familiar with what the law mandates. What you'd need to do to be compliant and offer the functionality you mention, is something a one-person company could handle with no issue. Not to mention that odds are a small business like that would buy an off the shelf webstore, which would include the required functionality and the documentation of it.
It doesn't take much to be compliant. Roughly speaking: informing the user, having them opt-in, letting the know what data is collected, and to delete it if they should so desire.
It is definitely good. A Mom and Pop shop in the states selling homemade soap can't afford to have a DPO
Good thing they wouldn't need one, then. There are criteria for when you'd need one (e.g. your business is mass storage or processing of personal data), and the odds of a tiny shop meeting any of them would be extremely slim. Heck, we're a multinational company and we don't need one. For that matter, there's no requirement to _hire_ someone, it's a role that could be assigned to any employee with sufficient knowledge of privacy laws and best practice.
if an EU resident visits a site, the site has to comply.
Not quite. If your site collects personal data about a EU resident, the site has to comply. If your site does not collect personal data, GDPR does not apply.
access to people's email inboxes in order to data-mine the contents for competitive intelligence -- and controversially flog the gleaned commercial insights to the likes of Uber
It's almost as if that's exactly the sort of undisclosed behavior the GDPR is designed to combat...
Granted, I suppose my subject is a bit unfair. If violating privacy is your primary business model, I guess "can't" is technically accurate.
Hm, having written that it occurs to me that we're talking about events here, which probably means security searches, which makes ticket processing a negligible part of the time consumption.
Expediency. If, say, one in a hundred gets denied, that means 99 people got to just walk through instead of holding up the line while being checked, and more will keep going through while that one person is off to the side getting manually processed. Also, if going by that figure, reducing the number of people needing manual processing by 99%, means fewer staff needed to do just that.
Aye. Considering Tesla's autopilot is little more than a slightly amped up version of lane keeping assist and adaptive cruise control, I'd say he was making a pretty good application for the Darwin awards.
I can't answer for anyone else but, while I have no interest in running a full Linux desktop at this time, I do like some of the tools available in the Linux terminal.
I did, in the past, try to run Linux as my main desktop, it failed miserably (don't get me started). I tried booting into it for the few tools that I preferred on the platform, it proved to be much too cumbersome to be worth it. Getting easy access to them from within Windows in a fashion that is not Cygwin, is all good in my book.
So, to answer your question, I guess WSL exists for people like me.
Steam publishing aggregate information would not be an issue from a GDPR perspective, as long as the source data is compliant (e.g. that it's collected because it is necessary for providing the service, and that the user has given consent).
Steam doesn't, though, so this effectively means the end of the type of information SteamSpy provides.
In that case, I'd be inclined to agree with those who feel this change is mainly to lock down easy access to the sort of aggregated information Steam Spy provides. If it was out of privacy concerns, it would make sense for a lot more to be hidden by default. In particular friend list and group memberships.
The summary states the information is now hidden by default, are you saying that is incorrect?
Wasn't the whole point that they are now changing to an opt-in system with opt-out being the default?
It's not only a good idea, in the EU it will be law then the GDPR goes into effect in May. If Valve did not make changes, they'd be liable for some pretty hefty fines for violating the regulation. They couldn't even just hide some legalese blurb way down in an EULA, when it comes to personal data it must be opt-in (and the opt-in must be a user action; a pre-checked consent checkbox is not sufficient), and information must be written in "clear and plain language".
You illustrate the very point I was making, that in the US you don't need any reason to own a gun, and you don't need to show that you are responsible enough to do so. If you don't think that's a problem, that's fine. You can't vote in my country, so it has no effect on me. For the sake of the people of the US, I just hope that you one day will be in the minority. There seems to finally be some movement in that direction, so there's hope.
That does not seem to be true. The specifics depend on your definition of "gun", but if assuming you mean handgun, then google suggests you can definitely own that in Australia. Not in the US "I need this... because.. I just do, ok, gief all the guns!" fashion, but as an active member of a club you can apply for a permit. If "gun" includes rifles and shotguns, then you can own those in England too, as far as I can tell.
From where I sit, regulation seems to be the crucial factor, with bans applying only to categories of firearms. Along with regulation you seem to get a culture where guns are treated with proper respect as opposed to as throwaway items anyone can possess without question or training and toss loaded into their nightstand drawers.
I don't believe there's a single solution that will work for every nation, there are so many factors at play, but it seems fairly self-evident that a good start is to do something different from what the US has been doing.
The worst I've experienced was driving through really slushy winter weather last spring. Wet snow splashed up and coated the car. The parking sensors got obstructed and the car whined endlessly about it, but the cruise control worked fine. In that case, I used it to test if it would work, not because I was going to trust it under those conditions.
I drove my previous car for 8 or so years without any electronics issues, I don't expect this to be any different myself. It's one reason I buy Toyotas. They're boring and overpriced for the equipment level, but they tend to just keep on going.
I don't really worry about such a scenario myself. If I drive in bad weather, I keep the cruise control distance set to maximum (if I use it at all, it depends on the conditions). If the sensor should fail in a way that makes it no longer see the car in front and thus increase speed, the acceleration would be obvious to me long before it became a danger. Same thing would be true if I caught up to a car, I'd notice the cruise control failing to slow down (and the dash not indicating the car in front being there) long before any danger.
Having said that, I'm pretty sure all these sensors are able to detect that they are obstructed and will disengage. To get in an accident because of this would require a triple event. Sensor being obstructed, its ability to detect that it is obstructed failing, and the driver not paying attention for a fair amount of time. Come to think of it, if the collision prevention system uses the parking sensors and/or the camera, there's even an extra layer of fault prevention there. So I relax, but keep my eyes on the road.
If there's a general consensus in the English speaking world that autopilot is synonymous with autonomous, I agree the name was poorly chosen. That's not my impression of the general understanding of the word – we expect a pilot to remain in the cockpit and alert when the plane we're on is on autopilot, after all – but if the data shows otherwise then I'd be the first to argue for the feature to have its name changed.
I guess our brains are very different. I find it really relaxing to just pay attention to the road and steering, and let the car deal with the speed. Since I'm obviously watching the road anyway, it takes away some (literal) footwork without adding any additional workload.
If it was unreliable, it would just add extra anxiety of having to quickly correct for it all the time. But in the year I've had the car, it's been rock solid, even in the worst of slushy winter conditions.
On really winding roads, I do tend to drive manually, though. The desired speed varies too much between straights and corners for cruise control to be convenient, and even if there is a car in front for the cruise control to speed match, that car momentarily disappears out of sight around sharper corners. But for regular highway travel, I've found it to be bloody awesome and never want a car without it again. I like it _that_ much.
Then putting Autopilot in a vehicle is illogical. You don't put something in a vehicle to steer for drivers while totally failing at relieving any kind of duty of driving. Eventually they will get sidetracked, it's just human.
It makes perfect sense to me, along the lines of adaptive cruise control. It makes for a more relaxing drive in that I don't have to actively maintain speed and distance myself, but it does not relieve me from the need to pay attention in order to be able to intervene should the need arise.
Additionally, if I should have a momentary lapse of attention at an inopportune time, odds are that it does not happen at the exact moment my car fails to notice that the car in front slowed down, so it adds safety.
For something like autopilot, it's very common knowledge that it's pretty decent on some types of roads (larger roads with visible painted lines), yet prone to getting it wrong in a lot of situations (e.g. something as simple as a deceleration lane veering off from the lane you are in can make it follow the rightmost line). In its current iteration it should be treated as a souped up combination of lane departure alert and adaptive cruise control.
In other words, autopilot _will_ get you in an accident if you don't pay attention, but in the vast majority of cases the driver will spot where autopilot is likely to have an issue long in advance of the car getting there.
The main issue, to me, seems to be that a lot of people fail to understand the distinction between autopilot (souped up cruise control) and an autonomous (entirely self-driving) vehicle. Perhaps there should be a quiz or something that you have to pass before you are allowed to enable it the first time...
This is one instance where I am fine with an off by one error.
This has become somewhat of a pet peeve of mine. In the vast majority of cases where I see AI used today, it seems to me that the proper term is really "machine learning". According to the dictionary, machine learning is a branch of AI. Sure, I'll grant it that. But that's not what the general public thinks of when AI pops up in articles. If "we fear AI", it's the Ex Machina kind, not the "Google Photos can recognize some types of objects in an image" kind.
Whenever I see e.g. the Google Assistant referred to as AI, I can but roll my eyes. Sure, I suppose it's technically correct in that there's machine learning behind the speech to text it uses to figure out what you're saying. But that's not what people are implying when they call it an AI. Anyone technically inclined who's ever used one of those things quickly realized that what happens next is no more advanced than a pretty limited number of IF statements.
Basically, we're currently using the term AI for everything from very basic machine learning, to sapience, where the general public tends to interpret the term more towards the latter than the former. I don't think I like that, and it turns out I'm currently bored enough to write about it. Sorry I just wasted seconds of your life by reading this rant. :p
In that case, support appreciated. :)
1919, not 1917, but I'm not sure I see your point. Did you reply to the wrong post? I wrote that we have decades worth of confirmation. I considered going with "about a hundred years", but considering gravitational lensing took longer to observe, I decided to avoid a possible pedant response pointing that out. I guess there's just no winning the internet. :p
You can't call it science when you use scientific theories to make discoveries? That seems to me to be what you're saying.
E.g. if we take Einstein's hypothesis for how light should behave near massive objects, and the predictions this provides, and confirm those, we can't call it science if we then use the resultant theory to make observations/discoveries?
Case in point, that mass bends light isn't some wild speculation, but a fact of reality with decades worth of confirmation at this point. There's a reason we call it the theory of general relativity, not the hypothesis. Using microlensing to make discoveries is thus perfectly valid science. Now, media has a tendency to ignore any kind of reservation that the actual scientists make. In this case, the obtained data seems open to interpretation. So, interesting and exciting, but not sufficient to actually announce a discovery. But claiming it's not science strikes me as ludicrous.
My point was that it did not answer the question. We know nothing of the woman's connectivity, or whether she cares about privacy issues. The "answer" offered no actual suggestions for solutions, just some related personal opinions phrased as absolutes.