"And if, for some reason, this sort of coverage hurts you, offends you, makes you angry and bitter and affronted, then maybe take a moment to close your eyes, really feel the hurt, and then try to determine why focused reporting on an important innovation made by an important company is something you feel personally."
Because it is still safer than the cars without it, and by casting it as awful, Jalopnik is killing people.
I don't even drive, I just ended up processing data on "road deaths"
There is no data published to fully back up the claim that autopilot is safer than humans. The problem on the surface is:
Autopilot is used mainly on limited access roads but safety is compared to all roads.
A Tesla is an expensive, new car and is compared to all cars of all ages and maintenance.
The drivers of above cars are also skewed statistically versus all drivers.
It's disappointing that the same people quick to claim to "follow the data" or "follow the science" are so comfortable doing nothing of the sort and simply trusting Tesla -- who obviously can't be objective here.
Do they compare like-for-like driving conditions and roadways? If it's just autopilot versus not-autopilot, that's flawed. People don't turn on autopilot in urban environments, for example, where accidents are more likely.
No it doesn't... autopilot can only be safely operated on highways, on roads with traffic lights and stop signs autopilot often times will refuse to even start, and when it does, it can only be used up until the next traffic stop, it can't be compared to a human driver under typical driving conditions.
The interest in these crashes is because technically, it's AI that has killed these people. That is fascinating in of itself. Are there any other automakers that have deaths from a similar cause to autopilot?
I'd be interested in quality data about the fatal crashes. I know Tesla states that the death rate is lower than with human drivers. But I've also heard that they use deaths per mile globally to compare and that may not be comparable. Is there any federal requirement for companies to report total autonomous miles driven?
Almost all the autonomous miles are on sunny days driven on the same circuit hour after hour. So the same mile driven millions of times, and even then they crash.
Source: lived along one of the Mountain View circuits, would see 3 autonomous vehicles from the same company in a row doing their "miles" with no other traffic. Even if they crashed into each other, it would be from the same company, so why report it?
As per HN guidelines please do not post comments insinuating that the person you replied to has not read the article. It does not lead to fruitful discussion.
Given that this problem has existed for decades with plane crashes I don‘t expect this to get better.
In fact US road deaths are increasing but the media has become numb to it because it‘s so routine. At best it‘s “accidents happen”, at worst it’s “that pedestrian was on their phone and should’ve worn bright reflective clothes, they had it coming”.
Seems to me that Yes, you can trick the car into thinking you are in the front seat while AP is on. But it takes some effort on your part.
You could also just push the gaspedal to the floor and crash that way. Would that also be Teslas fault?
> Also, they hyperbolic analogy of the Mustang and the brick would only make sense if Ford was selling a $10,000 Mustang Brick option that they claimed would let their car just about drive itself.
Here, they are responding to someone who asks “if I put a brick on the accelerator and crashed my car, would it be the manufacturer’s fault?”
But here Tesla isn't selling a $10,000 dollar brick; they're selling a $10,000 dollar system and reviewers are bringing their own Brick to circumvent safety systems built in.
When the driver has to willfully circumvent built-in safety features to do something clearly stupid; the line comes in on the side of "That person is an idiot" not "The system is inherently broken". Example: Ford could prevent people from doing stupid burnouts leaving car shows because mustangs fly into crowds multiple times a year.
I don't see a difference, stupid driver is stupid driver.
Quick Q: why is there such a confusion between Autopilot(Fancy ADAS/Level-2/Free on all Teslas iirc) and FSD(not yet complete/costs $10K/Many features still in access-limited beta)
iirc Tesla has been very clear to differentiate between the two; all the hands-free stuff is always linked to FSD and not Autopilot, like in the video they link on the article, the title of the video says FSD, not Autopilot.
Autopilot is a fancy cruise control, why is its abuse Tesla's headache?
If the issue was FSD related deaths, yes I would be concerned, but I feel any autopilot related deaths are like cruise control deaths; sad but ultimately something in the court of the user.
Autopilot isn’t a very good name for a fancy driver assist system.
Most regular folk associate the term autopilot with “the plane flies itself, the pilot is just there to guide it on landing”. We’ve had decades of this use of the term and got used to the idea that “autopilot” means the vehicle is self guided and the operator can take a nap, have a meal, etc.
1- would removing the word auto and just calling it "TeslaPilot"(TM) make it any better/worse? (in the same line of naming as Nissan ProPilot etc)
2-Aeroplanes famously have a co-pilot, and while one is eating/napping, the other is at the helm. Similar to a person riding shotgun, they can eat/nap while the driver drives, and then switch over after sometime. Does that make the autopilot name any better/worse?
19 comments
[ 62.3 ms ] story [ 759 ms ] threadBecause it is still safer than the cars without it, and by casting it as awful, Jalopnik is killing people.
I don't even drive, I just ended up processing data on "road deaths"
Autopilot is used mainly on limited access roads but safety is compared to all roads.
A Tesla is an expensive, new car and is compared to all cars of all ages and maintenance.
The drivers of above cars are also skewed statistically versus all drivers.
It's disappointing that the same people quick to claim to "follow the data" or "follow the science" are so comfortable doing nothing of the sort and simply trusting Tesla -- who obviously can't be objective here.
I'd be interested in quality data about the fatal crashes. I know Tesla states that the death rate is lower than with human drivers. But I've also heard that they use deaths per mile globally to compare and that may not be comparable. Is there any federal requirement for companies to report total autonomous miles driven?
Source: lived along one of the Mountain View circuits, would see 3 autonomous vehicles from the same company in a row doing their "miles" with no other traffic. Even if they crashed into each other, it would be from the same company, so why report it?
that's your cue to tell me that you actually did read it and that I'm only commenting to get attention.
What I find the media does not make any effort to do is is to communicate the level of risk.
It may be an unintentional bias, but it really is there.
I think they try harder when it's something like the J&J vaccine risk vs the risk of getting covid.
I'm uncertain what the high road looks like.
In fact US road deaths are increasing but the media has become numb to it because it‘s so routine. At best it‘s “accidents happen”, at worst it’s “that pedestrian was on their phone and should’ve worn bright reflective clothes, they had it coming”.
> Also, they hyperbolic analogy of the Mustang and the brick would only make sense if Ford was selling a $10,000 Mustang Brick option that they claimed would let their car just about drive itself.
Here, they are responding to someone who asks “if I put a brick on the accelerator and crashed my car, would it be the manufacturer’s fault?”
When the driver has to willfully circumvent built-in safety features to do something clearly stupid; the line comes in on the side of "That person is an idiot" not "The system is inherently broken". Example: Ford could prevent people from doing stupid burnouts leaving car shows because mustangs fly into crowds multiple times a year.
I don't see a difference, stupid driver is stupid driver.
iirc Tesla has been very clear to differentiate between the two; all the hands-free stuff is always linked to FSD and not Autopilot, like in the video they link on the article, the title of the video says FSD, not Autopilot.
Autopilot is a fancy cruise control, why is its abuse Tesla's headache?
If the issue was FSD related deaths, yes I would be concerned, but I feel any autopilot related deaths are like cruise control deaths; sad but ultimately something in the court of the user.
Most regular folk associate the term autopilot with “the plane flies itself, the pilot is just there to guide it on landing”. We’ve had decades of this use of the term and got used to the idea that “autopilot” means the vehicle is self guided and the operator can take a nap, have a meal, etc.
2-Aeroplanes famously have a co-pilot, and while one is eating/napping, the other is at the helm. Similar to a person riding shotgun, they can eat/nap while the driver drives, and then switch over after sometime. Does that make the autopilot name any better/worse?