3 ms·

It's important to remember they're only testing and using these cars in areas that don't have winters. The results don't extrapolate to driving in non-cherry pi

It's important to remember they're only testing and using these cars in areas that don't have winters. The results don't extrapolate to driving in non-cherry picked areas. This article should only be considered re: the context of local SF politics and the recent Cruise revocation.

SmarsJerry 3y ago
If you are comparing insurance rates then you are already comparing a similar area. If insurance companies would give you a discount if your car had self driving enabled that seems like a pretty sure fire guarantee that self driving is better.
- choppaface 3y ago
  Except humans drive all the time and AVs are usually turned off during extreme weather. Additionally, heavy rain days can cause a lot of people to dump Amazon delivery and do Uber for the day because rideshare demand spikes. Even if the AVs are left on, the relative population of the human drivers increases substantially (and the humans are also now also driving in less safe conditions).
- cheeselip420 3y ago
  Cheaper != safer or better... But at least it's observable
  - yreg 3y ago
    In case of insurance it does mean 'expected to be safer'.
  - kuchenbecker 3y ago
    Actuaries are cold and calculating, cheaper is safer all things being equal.
- superkuh 3y ago
  I guess I didn't get my meaning across very well. I am not saying anything about the relative safety of driving in cherry-picked environments like SF. The data can talk about that and in this case is supports autonomous cars as being safer in SF. I am saying that you can't extrapolate that data and say that autonomous vehicles will be safer than human drivers in, say, NYC or Minneapolis. They won't be. The environmental challenges that exist in most roads in the USA do not exist in the SF test environment. That's part of the reason why tests are always done in SF or Arizona, etc. They have much easier and unchanging road conditions and visibility. So the takeaways from this article only really apply and make sense as responses to the current local SF politics. They do not show autonomous cars are better drivers in general.

SmarsJerry 3y ago

If you are comparing insurance rates then you are already comparing a similar area. If insurance companies would give you a discount if your car had self driving enabled that seems like a pretty sure fire guarantee that self driving is better.

choppaface 3y ago
Except humans drive all the time and AVs are usually turned off during extreme weather. Additionally, heavy rain days can cause a lot of people to dump Amazon delivery and do Uber for the day because rideshare demand spikes. Even if the AVs are left on, the relative population of the human drivers increases substantially (and the humans are also now also driving in less safe conditions).
cheeselip420 3y ago
Cheaper != safer or better... But at least it's observable
- yreg 3y ago
  In case of insurance it does mean 'expected to be safer'.
- kuchenbecker 3y ago
  Actuaries are cold and calculating, cheaper is safer all things being equal.
superkuh 3y ago
I guess I didn't get my meaning across very well. I am not saying anything about the relative safety of driving in cherry-picked environments like SF. The data can talk about that and in this case is supports autonomous cars as being safer in SF. I am saying that you can't extrapolate that data and say that autonomous vehicles will be safer than human drivers in, say, NYC or Minneapolis. They won't be. The environmental challenges that exist in most roads in the USA do not exist in the SF test environment. That's part of the reason why tests are always done in SF or Arizona, etc. They have much easier and unchanging road conditions and visibility. So the takeaways from this article only really apply and make sense as responses to the current local SF politics. They do not show autonomous cars are better drivers in general.