5 ms·

There are going to be Autopilot crashes and people should disable it if they uncomfortable with putting their lives in the hands of Tesla's software. I would be

There are going to be Autopilot crashes and people should disable it if they uncomfortable with putting their lives in the hands of Tesla's software. I would be. But I see no reason to not believe to not believe what they have on https://www.tesla.com/VehicleSafetyReport https://www.tesla.com/VehicleSafetyReport It shows Teslas have much fewer accidents than the average car, and that Teslas with autopilot enabled are substantially safe per mile than with autopilot disabled (25-50% more miles driven per accident). The article says a Prius hit the same place as this Tesla the week before. But the NTSB isn't investigating Toyota.

stiglitz 7y ago
Teslas are more expensive than the average car (hence might be driven more carefully), and autopilot is more likely to be enabled in predictable driving environments like highways.
- dmix 7y ago
  > more likely to be enabled in predictable driving environments like highways. Which is what it was designed for and which is a much easier problem to solve than inner-city driving. Which IMO I feel like more self-driving companies should have focused on first to perfect their systems. We used feature phones before smartphones for a reason, hardware/software is best developed incrementally.
- tonyedgecombe 7y ago
  hence might be driven more carefully I don't know where you get that idea from, when I'm out cycling it's generally the people in expensive cars that don't pay much attention.
  - stiglitz 7y ago
    You’re more likely to notice expensive cars, and also more likely to notice cars with unsafe drivers.
  - heavyset_go 7y ago
    If you look at the stats, luxury cars are involved in less crashes per miles driven.
    - tonyedgecombe 7y ago
      Yet: https://www.thesun.co.uk/motors/9202047/bmw-drivers-most-likely-driving-conviction/ https://www.thesun.co.uk/motors/9202047/bmw-drivers-most-lik...
platz 7y ago
assuming those numbers hold when scaled up, you are compressing the actual distribution into a single number (an average). Overall that high-level number may be correct. But at the same time, the kinds of people that get into accidents will be a different distribution. How you drive factors a lot into how likely you are into getting in an accident, even allowing for the other drivers. If the distribution becomes essentially random then you might actually have a higher chance of getting into an accident than before, if you were a very careful driver. (Obviously those that benefit the most are the bad drivers since they'll no longer be bad) Just remember that whenever you summarize a distribution into a single number like an average, you loose a lot of information in the process.
ce4 7y ago
I'd be careful to compare those numbers - some sort of confirmation bias. AP is only used in "AP-easy situations" the harder to drive parts are steered by the human driver.
benzor 7y ago
Your claim relies entirely on the veracity of Tesla's safety report. Unfortunately I am inclined to mistrust any such vehicle safety claims directly from Tesla's own website. There is a conflict of interest in that they are incentivized to publish the most favorable numbers. We should be rightly cynical and rely only on numbers from an independent third party, as we would normally do for other companies with a less favorable reputation. Until we have such independently verified stats, we can't take those claims at face value. I agree that anyone who feels uncomfortable with Tesla's ADAS should simply turn it off. But don't you think it's unfair for the company to place an unfinished product in the hands of a consumer and pass off the risk-management responsibility to them? Tesla's cavalier attitude towards autonomous vehicle safety is concerning to me. They seem to adopt the "move fast and break things" approach, whereas teams like Waymo and Cruise are releasing things in a slower and more controlled manner. And not coincidentally, they've have far less accidents that way.
- skolsuper 7y ago
  If they truly believe that the autopilot driving is safer than unassisted, they might reasonably consider it an immoral act not to ship it. Given the number of lives lost in auto accidents every day, the issue is not as clear-cut as you're suggesting. It's a real-world trolley problem.
  - reallydontask 7y ago
    > If they truly believe that the autopilot driving is safer than unassisted, they might reasonably consider it an immoral act not to ship it. Then Tesla should be given it away, not charging extra for it. To do otherwise would be immoral.
    - dmode 7y ago
      That’s an absurd suggestion. Every car company charges more for advanced safety features
      - reallydontask 7y ago
techie128 7y ago
The AutoPilot is marketed as a safety feature like your link asserts. The issue though is that Tesla's software defect(s) is what killed him. NTSB would be investigating Toyota if their hardware or software defect(s) would've killed people. In Tesla's case, AutoPilot's lane keeping feature did not keep the car in the lane. Instead it drove the car into concrete at 71mph killing the driver. More over anti-collision feature did not stop the car from colliding with the concrete barrier. See where the issue lies? Will you tell the people who faced safety issues with their Takata airbags to get rid of them? or the Toyota owners whose vehicles had unintended acceleration that they should rip out their accelerators?
BoorishBears 7y ago
Accidents without AP vs Accidents with AP isn't what's interesting. Accidents without AP's overreaching half-baked features meant to serve as a marketing gimmick, but with safety technology that's been iterated for almost 20 years like LKAS and AEB, and is now getting affordable, is what's interesting. Now a Corolla, the most conservative car in the history of mainstream cars, can return you to your lane and warn you if it feels you're making too many mistakes due to being tired. These features save lives when they work, and more importantly, do not kill you when they malfunction, and do not lull you into a false sense of safety nearly as actively as a system literally named "Autopilot". AP could literally be the same system, not pretend that lane centering is so easy to do, be much more aggressive about disengaging like other manufacturers (see SuperCruise, which literally tracks your eyes to ensure focus is maintained), and be infinitely more palatable an experiment than it currently is - If it seems I'm being aggressively anti-AP here, it's because I didn't sign up to be a part of this beta, yet I have to drive on the road with these people. I like that a bored developer can't push updates to the lane guidance in my cars. AP has had regressions that brought back deadly bugs. I'll take a crusty system that was put through thousands of hours of testing then never "improved" again over that any day.
truncate 7y ago
I do believe autopilot has the potential to be safer than driving manually. But I'm not sure what to take out of that safety report. (1) Comparison against all other cars in US (and across all states) just introduces lots of sampling bias. Teslas are like much more common in states like CA, than say Tennessee. Tennessee accounts for much more accidents in general than California and that adds up in overall stats, but not in Tesla stats. (2) What kind of variation do we have in Tesla drivers vs other car drivers. As comment above mentions, Tesla drivers may be more responsible drivers. (3) What accounts as accident? Minor collisions (fender benders) vs Major collisions. More specifically, I want to know, how does Autopilot react when it encounters something it has never seen before (like the crash in article). Humans can handle things pretty well IMO given they are not way beyond speed-limit, and all kind of random shit happens on road. (4) Autopilot vs Manual Tesla stats is promising IMO. However, a little more information would be nice to have, like proportion of 1-billion miles logged as autopilot miles.
- spartanscrub 7y ago
  I can’t remember if it was just for fires (comparing ICE to electric), but I believe the rates are much closer when you compare to newer cars. Tesla’s are obviously newer and the report does not take this into consideration.
- Brakenshire 7y ago
  There’s also a sampling bias for the kinds of roads where Autopilot is available, and the kinds of circumstances where people are most likely to trust and use it.
absherwin 7y ago
The report is misleading because AP tends to be used far more on controlled access roads which have fewer crashes per mile. There are other biases too but this alone could easily explain more than the observed 1/3 risk reduction.
Luc 7y ago
> The article says a Prius hit the same place as this Tesla the week before. But the NTSB isn't investigating Toyota. Makes sense to investigate Tesla since it's always the same driver causing the accidents. It still makes sense even if this driver is safer than everyone else.
adrr 7y ago
Accidents happen more frequently in terms of miles driven on city streets where Autopilot isn't being used. Tesla is playing with numbers to make themselves look good. This false sense security is part of the problem. Autopilot needs to be closely monitored and isn't a substitute for a human a driver. I use autopilot everyday and there are cases where if i wasn't monitoring it, it would ran me into a stalled car or merged me into a concrete barrier.
keanzu 7y ago
> It shows Teslas have much fewer accidents than the average car Tesla’s Driver Fatality Rate is more than Triple that of Luxury Cars (and likely even higher) https://medium.com/@MidwesternHedgi/teslas-driver-fatality-rate-is-more-than-triple-that-of-luxury-cars-and-likely-even-higher-433670ddde17 https://medium.com/@MidwesternHedgi/teslas-driver-fatality-r...
bagacrap 7y ago
Autopilot is only going to be enabled on freeways, where accident rates will always be much lower than surface streets. Not a useful statistic, even if true.

stiglitz 7y ago

Teslas are more expensive than the average car (hence might be driven more carefully), and autopilot is more likely to be enabled in predictable driving environments like highways.

dmix 7y ago
> more likely to be enabled in predictable driving environments like highways. Which is what it was designed for and which is a much easier problem to solve than inner-city driving. Which IMO I feel like more self-driving companies should have focused on first to perfect their systems. We used feature phones before smartphones for a reason, hardware/software is best developed incrementally.
tonyedgecombe 7y ago
hence might be driven more carefully I don't know where you get that idea from, when I'm out cycling it's generally the people in expensive cars that don't pay much attention.
- stiglitz 7y ago
  You’re more likely to notice expensive cars, and also more likely to notice cars with unsafe drivers.
- heavyset_go 7y ago
  If you look at the stats, luxury cars are involved in less crashes per miles driven.
  - tonyedgecombe 7y ago
    Yet: https://www.thesun.co.uk/motors/9202047/bmw-drivers-most-likely-driving-conviction/ https://www.thesun.co.uk/motors/9202047/bmw-drivers-most-lik...

platz 7y ago

assuming those numbers hold when scaled up, you are compressing the actual distribution into a single number (an average). Overall that high-level number may be correct. But at the same time, the kinds of people that get into accidents will be a different distribution. How you drive factors a lot into how likely you are into getting in an accident, even allowing for the other drivers. If the distribution becomes essentially random then you might actually have a higher chance of getting into an accident than before, if you were a very careful driver. (Obviously those that benefit the most are the bad drivers since they'll no longer be bad) Just remember that whenever you summarize a distribution into a single number like an average, you loose a lot of information in the process.

ce4 7y ago

I'd be careful to compare those numbers - some sort of confirmation bias. AP is only used in "AP-easy situations" the harder to drive parts are steered by the human driver.

benzor 7y ago

Your claim relies entirely on the veracity of Tesla's safety report. Unfortunately I am inclined to mistrust any such vehicle safety claims directly from Tesla's own website. There is a conflict of interest in that they are incentivized to publish the most favorable numbers. We should be rightly cynical and rely only on numbers from an independent third party, as we would normally do for other companies with a less favorable reputation. Until we have such independently verified stats, we can't take those claims at face value. I agree that anyone who feels uncomfortable with Tesla's ADAS should simply turn it off. But don't you think it's unfair for the company to place an unfinished product in the hands of a consumer and pass off the risk-management responsibility to them? Tesla's cavalier attitude towards autonomous vehicle safety is concerning to me. They seem to adopt the "move fast and break things" approach, whereas teams like Waymo and Cruise are releasing things in a slower and more controlled manner. And not coincidentally, they've have far less accidents that way.

skolsuper 7y ago
If they truly believe that the autopilot driving is safer than unassisted, they might reasonably consider it an immoral act not to ship it. Given the number of lives lost in auto accidents every day, the issue is not as clear-cut as you're suggesting. It's a real-world trolley problem.
- reallydontask 7y ago
  > If they truly believe that the autopilot driving is safer than unassisted, they might reasonably consider it an immoral act not to ship it. Then Tesla should be given it away, not charging extra for it. To do otherwise would be immoral.
  - dmode 7y ago
    That’s an absurd suggestion. Every car company charges more for advanced safety features
    - reallydontask 7y ago

techie128 7y ago

The AutoPilot is marketed as a safety feature like your link asserts. The issue though is that Tesla's software defect(s) is what killed him. NTSB would be investigating Toyota if their hardware or software defect(s) would've killed people. In Tesla's case, AutoPilot's lane keeping feature did not keep the car in the lane. Instead it drove the car into concrete at 71mph killing the driver. More over anti-collision feature did not stop the car from colliding with the concrete barrier. See where the issue lies? Will you tell the people who faced safety issues with their Takata airbags to get rid of them? or the Toyota owners whose vehicles had unintended acceleration that they should rip out their accelerators?

BoorishBears 7y ago

Accidents without AP vs Accidents with AP isn't what's interesting. Accidents without AP's overreaching half-baked features meant to serve as a marketing gimmick, but with safety technology that's been iterated for almost 20 years like LKAS and AEB, and is now getting affordable, is what's interesting. Now a Corolla, the most conservative car in the history of mainstream cars, can return you to your lane and warn you if it feels you're making too many mistakes due to being tired. These features save lives when they work, and more importantly, do not kill you when they malfunction, and do not lull you into a false sense of safety nearly as actively as a system literally named "Autopilot". AP could literally be the same system, not pretend that lane centering is so easy to do, be much more aggressive about disengaging like other manufacturers (see SuperCruise, which literally tracks your eyes to ensure focus is maintained), and be infinitely more palatable an experiment than it currently is - If it seems I'm being aggressively anti-AP here, it's because I didn't sign up to be a part of this beta, yet I have to drive on the road with these people. I like that a bored developer can't push updates to the lane guidance in my cars. AP has had regressions that brought back deadly bugs. I'll take a crusty system that was put through thousands of hours of testing then never "improved" again over that any day.

truncate 7y ago

I do believe autopilot has the potential to be safer than driving manually. But I'm not sure what to take out of that safety report. (1) Comparison against all other cars in US (and across all states) just introduces lots of sampling bias. Teslas are like much more common in states like CA, than say Tennessee. Tennessee accounts for much more accidents in general than California and that adds up in overall stats, but not in Tesla stats. (2) What kind of variation do we have in Tesla drivers vs other car drivers. As comment above mentions, Tesla drivers may be more responsible drivers. (3) What accounts as accident? Minor collisions (fender benders) vs Major collisions. More specifically, I want to know, how does Autopilot react when it encounters something it has never seen before (like the crash in article). Humans can handle things pretty well IMO given they are not way beyond speed-limit, and all kind of random shit happens on road. (4) Autopilot vs Manual Tesla stats is promising IMO. However, a little more information would be nice to have, like proportion of 1-billion miles logged as autopilot miles.

spartanscrub 7y ago
I can’t remember if it was just for fires (comparing ICE to electric), but I believe the rates are much closer when you compare to newer cars. Tesla’s are obviously newer and the report does not take this into consideration.
Brakenshire 7y ago
There’s also a sampling bias for the kinds of roads where Autopilot is available, and the kinds of circumstances where people are most likely to trust and use it.

absherwin 7y ago

The report is misleading because AP tends to be used far more on controlled access roads which have fewer crashes per mile. There are other biases too but this alone could easily explain more than the observed 1/3 risk reduction.

Luc 7y ago

> The article says a Prius hit the same place as this Tesla the week before. But the NTSB isn't investigating Toyota. Makes sense to investigate Tesla since it's always the same driver causing the accidents. It still makes sense even if this driver is safer than everyone else.

adrr 7y ago

Accidents happen more frequently in terms of miles driven on city streets where Autopilot isn't being used. Tesla is playing with numbers to make themselves look good. This false sense security is part of the problem. Autopilot needs to be closely monitored and isn't a substitute for a human a driver. I use autopilot everyday and there are cases where if i wasn't monitoring it, it would ran me into a stalled car or merged me into a concrete barrier.

keanzu 7y ago

> It shows Teslas have much fewer accidents than the average car Tesla’s Driver Fatality Rate is more than Triple that of Luxury Cars (and likely even higher) https://medium.com/@MidwesternHedgi/teslas-driver-fatality-rate-is-more-than-triple-that-of-luxury-cars-and-likely-even-higher-433670ddde17 https://medium.com/@MidwesternHedgi/teslas-driver-fatality-r...

bagacrap 7y ago

Autopilot is only going to be enabled on freeways, where accident rates will always be much lower than surface streets. Not a useful statistic, even if true.