Hacker News Reader

HNHacker News

Top New Best Ask Show Jobs

Comment by sschueller | Hacker News Reader

Full thread

sschueller·

"which lowers overall insurance costs by factoring in the vehicles' Autopilot safety features"

Are they saying the autopilot makes the care safer? Hasn't the opposite occurred? [1]

[1] https://www.ntsb.gov/news/events/Documents/2017-HWY16FH018-B...

View on HN

No, the organization that put that slideshow together concluded at the end of its investigation that the primary causes of the crash were the truck failing to yield properly and the Tesla driver not paying attention for a long time. Their only recommendation to Tesla was to make Autopilot nag drivers more strictly if it doesn't detect that they're paying attention, which Tesla followed. Another agency investigating the crash, the NHTSA, concluded that they could not find a safety risk with Autopilot, they could not find any defects warranting a recall, and they noted that in broad terms Autopilot appears to have increased driver safety.

https://www.ntsb.gov/news/events/Documents/2017-HWY16FH018-B...

https://static.nhtsa.gov/odi/inv/2016/INCLA-PE16007-7876.PDF

No. Perhaps in a couple of cases, but on average, I believe it has shown to be safer.

From the original article:

"The National Highway Traffic Safety Administration found that crash rates for Tesla vehicles have plummeted 40% since Autopilot was first installed."

Are there actually enough crashes involving Teslas for this to be statistically useful? It’s not a particularly common car brand.

Good point. The sample size of one crash cited a few comments above is way too few to show autopilot makes cars more dangerous.

You don't need that big of a sample size for the results to be significant. A lot of people overestimate the sample size you need.

The safety benefit is from the automated emergency braking system, which Tesla has coupled with the Autopilot package. It's not from Autopilot itself.

Reply on news.ycombinator.com