Just making a blanket assertion that humanity needs to get off this rock to survive is pointless unless one actually has a plan that considers all the logistical hurdles of limitations on technology, reasonable expectations on the limitations of human labor, funding, etc, and still be able to demonstrate that it is realistically achievable. Has Hawking provided such a detailed plan, and been able to show that it would actually work in the real world? I doubt it..... just like every other crackpot who says the same thing.
While I would agree with you that it is probably the case that most of the time people are being inattentive rather than it being a medical emergency where the driver is being unresponsive to warnings by the vehicle to assume to control of the car, an inattentive driver is not going to be any less likely to kill somebody than a vehicle coming to a stop on the highway when it is done in a controlled fashion, and I would suggest that the latter would still be safer, in the long run.
Remember, again, the hazard lights would be flashing at this point, and this is going to be alerting other drivers around that something unusual is up with the car. The vehicle would *not* suddenly stop without giving any warning here... Active sensors in the car would also look at how quickly the vehicle behind is approaching and determine the best rate at which to slow down that still gives the trailing driver adequate time to react. If there had actually been something wrong, coming to a controlled stop in this way is going to be the safest course of action that is at the very least automatable (the safest course of action being to safely navigate to the shoulder and come to a stop there, but the tech doesn't exist to do that yet, at least not in any commercially viable mass-market vehicles).
Further, if the drivers of such cars knew that this was going to happen if they weren't paying attention (and especially if they knew 911 was going to be automatically called), I would suggest they would be less likely to let it get to that point in the first place, and be more willing to assume control of the vehicle whenever requested, making the scenario that the vehicle has come to a stop unnecessarily that much more unlikely.
A full stop even on a highway can be entirely safe if it is done in a certain way... first of all, when the hazard flashers are on, this is going to alert the driver of vehicle behind that something is up, and rather than coming to a sudden stop, the automated vehicle can simply initially only stop delivering torque to its drive axles, allowing it to coast on inertia for a moment while the trailing driver reacts to the fact that the car in front is starting to slow down. The automated vehicle could monitor the distance and relative speed of vehicles behind so as to ensure that it could do this safely, even while on a highway. As the trailing distance increases relative to speed, the car could then start decelerating more quickly, eventually coming to a complete stop. If 911 is automatically called to dispatch to the cars location to advise them of a nonresposive driver, then any medical emergency can be dealt with in a timely manner. If the driver was simply being deliberately inattentive, then he will get caught because 911 will have been called, his asshatery will get him a nice big fine, and if we're lucky, possibly even suspension of his license.
Ideally, of course, the vehicle would pull over to the shoulder before coming to a stop, but as has been pointed out already by many people, Tesla's autopilot is not anywhere nearly sophisticated enough to accomplish this maneuver while in traffic.... especially if it is in lane other than the one that is already nearest to the shoulder. Using the hazard lights to alert other drivers and implementing a controlled deceleration in the lane wherever the car happens to be is the second best option, and is the next most likely thing to prevent a serious accident.
Remember, this course of action would only be taken when the driver has *not* responded to repeated warnings to assume control of the vehicle. Whether it is because they are being deliberately inattentive or if there is a medical emergency, it is a GUARANTEE that the vehicle will eventually get in accident unless the driver is willing and able to assume control of the vehicle when the situation requires it. The only way that stopping on a highway can possibly be just as or any more likely to cause an accident than this is when the driver is somehow going to still be willing and able to assume control of the vehicle when it is absolutely necessary, which is not a valid assumption when the driver is non-responsive to repeated warnings to assume control of the car.
Again, an automated vehicle can monitor approaching cars and control its rate of deceleration to ensure that tailing drivers have enough time to react to the situation. The hazard lights of the vehicle would be flashing, which should alert the driver following that something is going on with the vehicle ahead, and by extension, they would be more prepared for a speed adjustment that will occur.
On the flip side, what if the sensors fail in the steering wheel and you are holding it but the car doesn't recognise it and refuses to move anywhere
The point of my suggestion would be to err on the side of caution. If the sensors aren't functioning, then the vehicle needs to be repaired, plain and simple. Once a driver is aware that the sensors are faulty, they could still use the car by starting it in manual driving mode, and not switching it to autopilot in the first place.
Although he may certainly be right about mass-market consumer devices, there will always remain at least a niche market for devices that are not connected to anything.
I mean heck... what about raw electronic components? Is he suggesting that even basic led's will connect to the internet?
I would suggest that if it doesn't know how to pull over, then the next safest thing would be (when the driver has been repeatedly unresponsive to advisories to assume control of the vehicle) is to turn on the car's hazard lights and stop supplying further torque to the drive axles, regardless of any pressure that may be on the accelerator, and as soon as it deemed safe to do so, based on the relative velocity and distance of any vehicles behind it, decelerating as quickly as can be managed while still givng the driver behind enough time to react to the velocity change and either change lanes or slow down as well. Since the hazards would be flashing at this point, this should alert any driver following that something is up with the vehicle, and they will, in general, be more prepared to deal with the imminent velocity change, even if it is on a highway.
911 could then be automatically called by the vehicle, supplying information obtained via gps for location data to dispatch an emergency vehicle to deal with the unresponsive driver. The call should be put on the car's radio speakers so that if there are any passengers, they will be made aware of what is happening.
If it was just an idiot driver that was being deliberately unresponsive.... well, then, he should be fined for causing 911 to be dispatched for a non-emergency, and also fined for being inattentive to the surrounding conditions while behind the wheel of a moving automobile.
Given that the tech doesn't seem to exist yet for a full autopilot, the next best thing to do when the driver has been unresponsive to warnings to assume control of the vehicle is to stop delivering torque to the vehicle axle, regardless of any pressure on the accelerator, turn its hazard flashers on, and then begin decelerating to a full stop at the fastest rate that is safely possible commensurate with its distance and relative speed to the vehicle behind it. Again, if there were a medical emergency in the vehicle that has caused the driver to be unresponsive, this course of action is less likely to result in a serious accident than while the vehicle is moving at speed.
Not entirely true.... although it is more likely a correlative relationship than a causative one.... a driver that puts a seatbelt on is taking an initial step to dealing with what is an otherwise unforseeable incident while driving, and I would suggest this is corrrelated to how defensively they drive in the first place, which in turn relates directly to the likelihood of even being in an accident, regardless of who caused it.
Before there were laws requiring seatbelts in my area, any driver that cared so little for their own life that they wouldn't put on a seatbelt is not someone I would want to be in a vehicle with.
I don't believe that he was in this case, but I think that you raise a valid point in general... the proper thing for an autopilot to do when a driver has repeatedly ignored warnings to assume control of the vehicle is to pull over and stop as soon as it is possible to safely do so, and probably even call 911.
No... it should pull over and stop as soon as it is possible to safely do so. What if there were a medical emergency in the vehicle that is causing the driver to be unresponsive?
Of course, if you expect drivers that are able to maintain consciousness while having a heart attack to just keep driving instead of pulling over and coming to a stop as soon as they can, then I'm not sure what world you are living in.
I would say so, yes.. the driver could be incapacitated for some reason... What if a driver had a heart attack and is unconscious, for example? If the driver ignores repeated warnings to put their hands on the wheel, the most sensible thing to do is to pull the car over as soon as it is prudent to do so, and call 911 once the car is stopped, with a declaration that the driver of the vehicle appears to be incapacitated and unresponsive to audio or visual cues to control the vehicle..
Seriously... online business is decimating brick and mortar stores but there is absolutely no way I would ever buy something like a sofa without trying it out first.
Cell phones are bad for a parent's pocketbook when the kid isn't responsible enough to keep it from getting lost when they take it somewhere.
It's cheaper in the long run to just wait until they are old enough that they are earning their own money to buy their own phone and pay for a young person's account themselves.
As we were definitely not rich in the 90's, I can most certainly say that assertion is false. All of my kids had Lego when they were growing up in that decade, and a couple of them still play with it now as adults.
The point remains, if there is no electrical connectivity anywhere near where people park their cars, you are looking at an extremely expensive rewiring job just to even *begin*. I realize that this isn't applicable to everybody, but it's not even slightly abnormal in high density living situations.
The only objective difference between them lies in an appeal to emotion.
Note that I did not ever contest the notion that motive for killing someone could ever make a difference (obviously killing in self defense is one example where it such action would be legally justified), but I simply responded to the comment that I quoted above, and I had not previously even responded to another post in this thread, so I cannot be accused of having changed the subject we were discussing.
If you're not doing anything wrong, why do you need to hide your communications with encryption?
So that people with possibly nefarious intentions cannot monitor you. Even ignoring the nonzero possibility that the government itself might be anything less than 100% benevolent, if the government can decrypt your content, then so can others... in fact, if there was any real ability to stop this, then they would have been able to stop unauthorized monitoring in the first place
That was the point I was trying to make as well, above.... Apparently, you phrased it a heck of a lot better than I did, because I certainly got an earful of response about how wrong I am.
Slashdot Mirror
Just making a blanket assertion that humanity needs to get off this rock to survive is pointless unless one actually has a plan that considers all the logistical hurdles of limitations on technology, reasonable expectations on the limitations of human labor, funding, etc, and still be able to demonstrate that it is realistically achievable. Has Hawking provided such a detailed plan, and been able to show that it would actually work in the real world? I doubt it..... just like every other crackpot who says the same thing.
While I would agree with you that it is probably the case that most of the time people are being inattentive rather than it being a medical emergency where the driver is being unresponsive to warnings by the vehicle to assume to control of the car, an inattentive driver is not going to be any less likely to kill somebody than a vehicle coming to a stop on the highway when it is done in a controlled fashion, and I would suggest that the latter would still be safer, in the long run.
Remember, again, the hazard lights would be flashing at this point, and this is going to be alerting other drivers around that something unusual is up with the car. The vehicle would *not* suddenly stop without giving any warning here... Active sensors in the car would also look at how quickly the vehicle behind is approaching and determine the best rate at which to slow down that still gives the trailing driver adequate time to react. If there had actually been something wrong, coming to a controlled stop in this way is going to be the safest course of action that is at the very least automatable (the safest course of action being to safely navigate to the shoulder and come to a stop there, but the tech doesn't exist to do that yet, at least not in any commercially viable mass-market vehicles).
Further, if the drivers of such cars knew that this was going to happen if they weren't paying attention (and especially if they knew 911 was going to be automatically called), I would suggest they would be less likely to let it get to that point in the first place, and be more willing to assume control of the vehicle whenever requested, making the scenario that the vehicle has come to a stop unnecessarily that much more unlikely.
Ideally, of course, the vehicle would pull over to the shoulder before coming to a stop, but as has been pointed out already by many people, Tesla's autopilot is not anywhere nearly sophisticated enough to accomplish this maneuver while in traffic.... especially if it is in lane other than the one that is already nearest to the shoulder. Using the hazard lights to alert other drivers and implementing a controlled deceleration in the lane wherever the car happens to be is the second best option, and is the next most likely thing to prevent a serious accident.
Remember, this course of action would only be taken when the driver has *not* responded to repeated warnings to assume control of the vehicle. Whether it is because they are being deliberately inattentive or if there is a medical emergency, it is a GUARANTEE that the vehicle will eventually get in accident unless the driver is willing and able to assume control of the vehicle when the situation requires it. The only way that stopping on a highway can possibly be just as or any more likely to cause an accident than this is when the driver is somehow going to still be willing and able to assume control of the vehicle when it is absolutely necessary, which is not a valid assumption when the driver is non-responsive to repeated warnings to assume control of the car.
Again, an automated vehicle can monitor approaching cars and control its rate of deceleration to ensure that tailing drivers have enough time to react to the situation. The hazard lights of the vehicle would be flashing, which should alert the driver following that something is going on with the vehicle ahead, and by extension, they would be more prepared for a speed adjustment that will occur.
Indeed, but "auto pilot" is easier to say than the mouthful "Adaptive Cruise Control, Lane Keeping Assist, and Active Collision Avoidance System".
I mean, even if you could successfully pronounce "ACCLKAACAS", nobody would have a fucking clue what you were talking about.
The point of my suggestion would be to err on the side of caution. If the sensors aren't functioning, then the vehicle needs to be repaired, plain and simple. Once a driver is aware that the sensors are faulty, they could still use the car by starting it in manual driving mode, and not switching it to autopilot in the first place.
Although he may certainly be right about mass-market consumer devices, there will always remain at least a niche market for devices that are not connected to anything.
I mean heck... what about raw electronic components? Is he suggesting that even basic led's will connect to the internet?
I would suggest that if it doesn't know how to pull over, then the next safest thing would be (when the driver has been repeatedly unresponsive to advisories to assume control of the vehicle) is to turn on the car's hazard lights and stop supplying further torque to the drive axles, regardless of any pressure that may be on the accelerator, and as soon as it deemed safe to do so, based on the relative velocity and distance of any vehicles behind it, decelerating as quickly as can be managed while still givng the driver behind enough time to react to the velocity change and either change lanes or slow down as well. Since the hazards would be flashing at this point, this should alert any driver following that something is up with the vehicle, and they will, in general, be more prepared to deal with the imminent velocity change, even if it is on a highway.
911 could then be automatically called by the vehicle, supplying information obtained via gps for location data to dispatch an emergency vehicle to deal with the unresponsive driver. The call should be put on the car's radio speakers so that if there are any passengers, they will be made aware of what is happening.
If it was just an idiot driver that was being deliberately unresponsive.... well, then, he should be fined for causing 911 to be dispatched for a non-emergency, and also fined for being inattentive to the surrounding conditions while behind the wheel of a moving automobile.
Given that the tech doesn't seem to exist yet for a full autopilot, the next best thing to do when the driver has been unresponsive to warnings to assume control of the vehicle is to stop delivering torque to the vehicle axle, regardless of any pressure on the accelerator, turn its hazard flashers on, and then begin decelerating to a full stop at the fastest rate that is safely possible commensurate with its distance and relative speed to the vehicle behind it. Again, if there were a medical emergency in the vehicle that has caused the driver to be unresponsive, this course of action is less likely to result in a serious accident than while the vehicle is moving at speed.
Not entirely true.... although it is more likely a correlative relationship than a causative one.... a driver that puts a seatbelt on is taking an initial step to dealing with what is an otherwise unforseeable incident while driving, and I would suggest this is corrrelated to how defensively they drive in the first place, which in turn relates directly to the likelihood of even being in an accident, regardless of who caused it.
Before there were laws requiring seatbelts in my area, any driver that cared so little for their own life that they wouldn't put on a seatbelt is not someone I would want to be in a vehicle with.
I don't believe that he was in this case, but I think that you raise a valid point in general... the proper thing for an autopilot to do when a driver has repeatedly ignored warnings to assume control of the vehicle is to pull over and stop as soon as it is possible to safely do so, and probably even call 911.
No... it should pull over and stop as soon as it is possible to safely do so. What if there were a medical emergency in the vehicle that is causing the driver to be unresponsive?
Of course, if you expect drivers that are able to maintain consciousness while having a heart attack to just keep driving instead of pulling over and coming to a stop as soon as they can, then I'm not sure what world you are living in.
I would say so, yes.. the driver could be incapacitated for some reason... What if a driver had a heart attack and is unconscious, for example? If the driver ignores repeated warnings to put their hands on the wheel, the most sensible thing to do is to pull the car over as soon as it is prudent to do so, and call 911 once the car is stopped, with a declaration that the driver of the vehicle appears to be incapacitated and unresponsive to audio or visual cues to control the vehicle..
Seriously... online business is decimating brick and mortar stores but there is absolutely no way I would ever buy something like a sofa without trying it out first.
Cell phones are bad for a parent's pocketbook when the kid isn't responsible enough to keep it from getting lost when they take it somewhere.
It's cheaper in the long run to just wait until they are old enough that they are earning their own money to buy their own phone and pay for a young person's account themselves.
Or is mobilesyrup.com somehow supposed to be an official spokesperson for the CRTC?
As we were definitely not rich in the 90's, I can most certainly say that assertion is false. All of my kids had Lego when they were growing up in that decade, and a couple of them still play with it now as adults.
The point remains, if there is no electrical connectivity anywhere near where people park their cars, you are looking at an extremely expensive rewiring job just to even *begin*. I realize that this isn't applicable to everybody, but it's not even slightly abnormal in high density living situations.
And in my response, I said:
Note that I did not ever contest the notion that motive for killing someone could ever make a difference (obviously killing in self defense is one example where it such action would be legally justified), but I simply responded to the comment that I quoted above, and I had not previously even responded to another post in this thread, so I cannot be accused of having changed the subject we were discussing.
So that people with possibly nefarious intentions cannot monitor you. Even ignoring the nonzero possibility that the government itself might be anything less than 100% benevolent, if the government can decrypt your content, then so can others... in fact, if there was any real ability to stop this, then they would have been able to stop unauthorized monitoring in the first place
That was the point I was trying to make as well, above.... Apparently, you phrased it a heck of a lot better than I did, because I certainly got an earful of response about how wrong I am.
It has evidently failed to capture your attention that the post to which I had responded above did not mention killing in self defense, at all.
... is that people are not allowed to talk in a language that the powers that be don't know, and talking in such a language is criminally punishable.
You're being ironic, aren't you?
The only objective difference between them lies in an appeal to emotion.