Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Waymo's driverless cars crash less often than people

iihs.org

Post reply on HN
Print view View original

31–36 of 36 posts

Re: Waymo's driverless cars crash less often than people

#31

Post by xnx »

beering wrote: post #19
autoexec wrote: post #4

Unless those people are also driving taxis maybe. Waymo Robotaxis Record Higher Serious Injury and Fatal Crash Rate Than NYC's For-Hire Fleet ( https://www.yahoo.com/news/science/articles/waymo-robotaxis-... ) It may also be worth noting that waymo doesn't report all crashes to the DMV and they went to court to keep crash data secret ( https://techcrunch.com/2022/02/22/waymo-to-keep-robotaxi-saf... )

I looked through the report. It’s a bad report but it couldn’t have been done well because of not having apples to apples data. They compare data from very different environments (all CA Waymo vs NYC taxi) and just kind of shrug aside that difference.

If they weren't motivated to be anti-Waymo, an invalid "study" like that wouldn't be published at all.

Re: Waymo's driverless cars crash less often than people

#32

Post by xnx »

calldacopsidgaf wrote: post #2

"safety" is usually the first point out of the mouth of a self-driving advocate. If safety = preservation of life, I wonder how truly "safe" full Waymo supremacy would be when there are millions of Americans who drive for a living [1], in a country where working age adults need a full-time job to get basic healthcare. Through that lens, I honestly don't know if Waymo's future would lead to preservation of life. I act…

To misquote Milton Friedman: "Oh, I thought you were trying to have a taxi service. If it's jobs you want, we should have 2 drivers per car."

Re: Waymo's driverless cars crash less often than people

#34

Post by vel0city »

dominotw wrote: post #27
vel0city wrote: post #15

Earlier quoted context omitted.

Its a common way of analyzing these kinds of statistics especially on large scales. What other divisor makes sense to you?

same divisor as human comparision. let me ask you the opposite, what divisor(if any ) would make you question it.

VMT is a common divisor used in human traffic comparisons. Like when you're trying to see if a given area is safer than another area to drive, you'll often compare by VMT to have a more fair comparison.

As for bad divisors, accidents per car? Accidents per paint color? Accidents per average number of donut shops in the area? I dunno, tons of things.

Re: Waymo's driverless cars crash less often than people

#35

Post by brokensegue »

dominotw wrote: post #27
vel0city wrote: post #15

Earlier quoted context omitted.

Its a common way of analyzing these kinds of statistics especially on large scales. What other divisor makes sense to you?

same divisor as human comparision. let me ask you the opposite, what divisor(if any ) would make you question it.

Per hour?
Post reply on HN