3 ms·

Tesla publishes details about accidents per million miles driven and compares that with Autopilot active and without. https://www.tesla.com/VehicleSafetyReport

Tesla publishes details about accidents per million miles driven and compares that with Autopilot active and without. https://www.tesla.com/VehicleSafetyReport https://www.tesla.com/VehicleSafetyReport In the 4th quarter, we recorded one crash for every 4.31 million miles driven in which drivers were using Autopilot technology (Autosteer and active safety features). For drivers who were not using Autopilot technology (no Autosteer and active safety features), we recorded one crash for every 1.59 million miles driven. By comparison, NHTSA’s most recent data shows that in the United States there is an automobile crash every 484,000 miles. Unfortunately, it doesn't look like they split out numbers for FSD, which is important to understand in this context. Assuming FSD is currently as good as a human driver (Tesla suggests it is better), how much better does FSD have to be than a person to be accepted as a safer way to drive?

lern_too_spel 4y ago
> Assuming FSD is currently as good as a human driver That's a massive assumption made without evidence. > Tesla publishes details about accidents per million miles driven and compares that with Autopilot active and without. The problems with Tesla's autopilot (different from FSD) statistics have been pointed out before: https://arstechnica.com/cars/2022/06/teslas-using-autopilot-crashed-273-times-in-less-than-a-year/ https://arstechnica.com/cars/2022/06/teslas-using-autopilot-...
- pedalpete 4y ago
  I wasn't aware of the problems with Tesla's statistics before. I find it of an accusation that you say I've a "massive assumption made without evidence". The "evidence" I have is the data from Tesla itself, which states that FSD is better than a person, so I reduced that to say "as good". The "problems with Tesla's data" as pointed out in the ars technica article doesn't make a comparison to an average driver, or give details on the number of miles driven per accident, etc etc So, how are we to know where a regular human driver falls on the spectrum.
  - lern_too_spel 4y ago
    > The "evidence" I have is the data from Tesla itself, which states that FSD is better than a person, so I reduced that to say "as good". What data has Tesla provided about FSD? The only data I'm aware of is about Autopilot, and that data doesn't look very good when correctly interpreted.

lern_too_spel 4y ago

> Assuming FSD is currently as good as a human driver That's a massive assumption made without evidence. > Tesla publishes details about accidents per million miles driven and compares that with Autopilot active and without. The problems with Tesla's autopilot (different from FSD) statistics have been pointed out before: https://arstechnica.com/cars/2022/06/teslas-using-autopilot-crashed-273-times-in-less-than-a-year/ https://arstechnica.com/cars/2022/06/teslas-using-autopilot-...

pedalpete 4y ago
I wasn't aware of the problems with Tesla's statistics before. I find it of an accusation that you say I've a "massive assumption made without evidence". The "evidence" I have is the data from Tesla itself, which states that FSD is better than a person, so I reduced that to say "as good". The "problems with Tesla's data" as pointed out in the ars technica article doesn't make a comparison to an average driver, or give details on the number of miles driven per accident, etc etc So, how are we to know where a regular human driver falls on the spectrum.
- lern_too_spel 4y ago
  > The "evidence" I have is the data from Tesla itself, which states that FSD is better than a person, so I reduced that to say "as good". What data has Tesla provided about FSD? The only data I'm aware of is about Autopilot, and that data doesn't look very good when correctly interpreted.