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

California DMV accuses Tesla of making false claims about Autopilot and FSD

theverge.com

Post reply on HN
Print view View original

51–60 of 85 posts

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#51

Post by Volundr »

BoorishBears wrote: post #44
toolz wrote: post #35

Earlier quoted context omitted.

Tesla isn't hiding data, they simply only have autopilot data on highways, which their autopilot data suggests it's safer than human pilots on highways. Which to me suggests we should be putting social effort behind making autopilot (on the highway, at least) the norm. Perfection is not a requirement for advancement.

> Tesla isn't hiding data No they are. They won't break out FSD and they 100% could > their autopilot data suggests it's safer than human pilots on highways Their autopilot data compared to every single type of car, with every single type of driver, only on highways. So we're comparing every 1994 Corolla with a teenage driver to every 2021 EV with a 30 yr old driver. The reality is AP should really be compared to mod…

> Their autopilot data compared to every single type of car, with every single type of driver, only on highways.

Is there new data I missed? Last I was aware Telsa had published data comparing autopilot to human drivers, but not adjusting for road type, weather, road condition, construction, and any number of other factors in addition to the ones you mentioned.

So basically Telsa determined that autopilot on easy-mode driving (highway, clear enough conditions for autopilot to be willing to operate), with a human backup, has less crashes per-mile than humans in the full range of conditions (including those where autopilot will opt out) with no backup.

If that's not an example of "lies, damn lies, and statistics" I don't know what is.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#52

Post by _ea1k »

misiti3780 wrote: post #29
belkinpower wrote: post #25

Earlier quoted context omitted.

Words have meanings. If you call something “full self-driving” you can’t get upset when people assume it means full self-driving.

So just so I understand, you think there are Tesla owners that think they have Full self driving because the Tesla website calls it FSD?

No, but I do think people who don't own one sometimes buy the car thinking that they are getting FSD.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#53

Post by _ea1k »

andrewmcwatters wrote: post #43

Tesla is the only company that seems to have this issue considering many big players have some form of ACC and LKAS. You just don’t hear Toyota calling their self-driving “Toyota Autosteer.” This all seems to be driven by arrogance and marketing.

Other companies do not fully collect the data on crashes while autosteer type features are active and how many miles they are driven. Comparison is difficult.

NHTSA is making headway, but its going to take time to reach accurate conclusions.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#54

Post by _ea1k »

culi wrote: post #20

Before Tesla came out with autopilot, Google had done extensive user research on "autopilot" cars and found that no matter how many times you tell users that they must be fully attentive, people WILL go on their phones or stop paying attention. It's the entire reason Google didn't go to market with their technology before Tesla Then Musk came along and completely ignored the research (which he must've been really awa…

As tempted as I am to agree, this isn't the direction that anyone in the industry is going. There are lots of lane keeping assistant systems coming onto the market now with similar capabilities to AP.

After watching thorough video reviews, the differences between the best systems are fairly minor.

Making it all about Tesla will do the industry a real disservice. Thankfully NHTSA seems to be aware of the issues and is taking their time to understand fully before making adjustments to regulations.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#55

Post by ra7 »

nradov wrote: post #46
ra7 wrote: post #30

Earlier quoted context omitted.

Here is the video of that Google experiment you mentioned: https://www.youtube.com/watch?v=6ePWBBrWSzo Google had hands-free driving working on highways a decade ago, but decided they didn't want to take the incremental approach of L2 -> L3 -> L4 exactly for the reasons you mentioned and jumped straight to making L4 (fully driverless) work.

GM has taken the incremental approach and has been selling the L2 Super Cruise feature for several years. So far it seems to have a good safety record. They plan to start selling an "Ultra Cruise" system next year with autonomous driving on a wider range of roads, but it will still be L2. Mercedes-Benz had hands-free driving working on highways in 1987 (before Google even existed). However, their approach relied on e…

IIRC, Super Cruise hasn’t been around for long and is only available in certain GM vehicles. Their usage is certainly not at the level of Autopilot in terms of miles. I think when they hit that scale GM will also find out L2 systems are only as good as the driver’s attention span.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#56

Post by mikestew »

gamblor956 wrote: post #50
toolz wrote: post #36

Earlier quoted context omitted.

They release quarterly data that shows it works as promised and is much safer than human drivers on highways.

I have to drive with Teslas every day on my way to work. Teslas drive like drunk drivers, whether using AP or FSD. Tesla fans might feel like they are safer, but from the perspective of someone else on the road, they are objectively less safe, unless you consider swerving, randomly slamming on the brakes, and tailgating to be examples of safe driving. And Tesla's AP system has more recorded fatalities than all of the…

...and tailgating

Is that a AP/FSD thing? I ask only because just this weekend there was a Tesla on my rear bumper in a 30mph zone, and I had the cruise set to 34mph. I did wonder whether it was the computer, or if the driver was just an ass. Regardless, after a number of miles I grew tired of it, let's give 'em a healthy brake tap. Just enough to get the hint, I don't want anyone spilling coffee on themselves. So I give the brakes a good jab, and immediately back off.

Next thing I know, I hear the tire chirp of ABS kicking in, and I look in the rearview to see the Tesla stopped in the middle of the road. Oops, sorry (not sorry?), not exactly what I was going for. OTOH, maybe you or the car will quit riding people's asses now. And I'll admit I got a good chuckle out of it.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#57

Post by panick21_ »

culi wrote: post #20

Before Tesla came out with autopilot, Google had done extensive user research on "autopilot" cars and found that no matter how many times you tell users that they must be fully attentive, people WILL go on their phones or stop paying attention. It's the entire reason Google didn't go to market with their technology before Tesla Then Musk came along and completely ignored the research (which he must've been really awa…

> Google was in negotiations to buy Tesla

Did that happen? I thought that was just a private thing between Musk and Brin.

> Tesla's still have problems. They made up 273 accidents involving driver-assistance systems in 2021. 70% of all such accidents.

Is that broken down anywhere in terms of actual miles, and even more importantly high way miles compared to non high way miles.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#58

Post by panick21_ »

adoxyz wrote: post #14
dekhn wrote: post #8

Earlier quoted context omitted.

True, but a few years ago they had a huge problem with Model3 and ModelY quality- you had to be very careful and inspect the whole car for serious problems before accepting delivery.

lol if you look at any Tesla forums, this is still very much the case with all cars they make. The refreshed Model X seems to be hot garbage from a quality control stand point.

Tesla forum is literally the perfect example of selection bias.

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#59

Post by chmod600 »

simion314 wrote: post #33
chmod600 wrote: post #26

Earlier quoted context omitted.

Are Teslas more lethal per mile than other cars?

We can't know, Tesla is secret about the real data. From what we see in videos where people are testing FSD and the human had to save teh day many times in 30 minutes. So for just that 1 car you would have had 1 death if the human was not there. IF we would have the number of how many times the driver would have been hurt if he did not take control then we could have a better guess how many kilometers a Tesla would s…

That's not needed to answer my question. Regardless of FSD, are Teslas more lethal per mile than other cars?

Re: California DMV accuses Tesla of making false claims about Autopilot and FSD

#60

Post by s1artibartfast »

culi wrote: post #20

Before Tesla came out with autopilot, Google had done extensive user research on "autopilot" cars and found that no matter how many times you tell users that they must be fully attentive, people WILL go on their phones or stop paying attention. It's the entire reason Google didn't go to market with their technology before Tesla Then Musk came along and completely ignored the research (which he must've been really awa…

This is another version of the "guns don't kill people, people do" argument.

Even without any driving assistance, people will go in their their phones and have lapses of attention.

The data shows that people are more irresponsible with the autopilot technology. Does this fault fall on the drivers or manufacturer.

Post reply on HN