I don't see how that's supported here? I mean, clearly Tesla does its own closed-track testing. I've seen coverage of that in the past. They don't release public software the doesn't pass these kinds of tests. Likewise it's not like Waymo restricts their testing to closed tracks, they have vehicles on the streets too.
> One of the scariest parts about Tesla is that they don't even seem to know what they don't know.
I'm curious what the reference here is? Again, are you taking coverage of Waymo's test environment as evience of its absence at its competitors?
How? The only problem with this page[0] is that they haven’t released 2022 stats. Otherwise:
> To ensure our statistics are conservative, we count any crash in which Autopilot was deactivated within 5 seconds before impact, and we count all crashes in which the incident alert indicated an airbag or other active restraint deployed.
That "vehicle safety report" is 70 words, none of which answer the question of "how many crashes were there" or "how many people were injured or killed." One of the few numbers they give us, the percentage of incidents/hour for driving "with active safety but without autopilot," bounces around by like a factor of two from quarter to quarter, which doesn't make any sense. Literally every provided number is in Tesla's favor, and context is only provided in ways favorable to Tesla.
I'm honestly curious what test tracks Tesla uses. Do they just rent private tracks for their road testing before it gets to public beta? It clearly isn't done onroad by safety drivers.
GM has Milford and Yuma + others, Waymo has Castle, Zoox uses Altamont Raceway, Nuro has a track at the Vegas speedway. I can't imagine Tesla's small Fremont loop and their winter track in Fairbanks are sufficient.
Tesla’s FSD is so terrifyingly bad at routine tasks (I used it for six months before giving up) that it’s natural to assume whatever closed track testing they did was ineffective. Or perhaps they did a lot, I don’t know—but it doesn’t feel like many of those lessons learned made it into the “production” system.
I really like AP in general, but for FSD Beta, I tend to agree. I've seen enough mistakes from even the really careful Youtubers that I don't understand why its still in the field.
And those are the mistakes that they were willing to show! To be clear, I mean situations where the driver needed to take over but either didn't, or didn't in sufficient time to avoid an illegal or dangerous maneuver.
From what I can tell of people posting FSD videos on YouTube, they are actively seeking adversarial conditions with a desire to show where it fails. I’m sure there are some YouTubers that are trying to sugarcoat FSD, but I haven’t seen any.
I partially agree. I mean, I find dirty tesla's videos to be pretty fair.
Having said that, he's also been really clear that the video doesn't always make it obvious just how many aspects of FSD are just plain weird. Even when its not actively failing, it moves in odd ways that are uncomfortable.
He's also said that earlier versions resulted in curbed wheels.
Agreed, that's one way. The other way is multiple downward facing cameras with stereoscopic vision of the area immediately around the car. Mobileye demonstrated a system like that and it worked really well.
Tesla's cameras have huge blind spots near the car and aren't stereoscopic in all directions. They had ultrasonic sensors at one time, but those had blind spots and major limitations too.
> They don't release public software the doesn't pass these kinds of tests.
Do we actually know that this is true? October of last year "internal QA" found regressions on an already public release and they rolled out an update in less than 24 hours after that was published. Both the fact that it was already public when QA found an issue and that they were able to push new public version in such a short time seems to me like they don't necessarily have a release gauntlet for each version, or at least not a very robust one.
> One of the scariest parts about Tesla is that they don't even seem to know what they don't know.
I'm curious what the reference here is? Again, are you taking coverage of Waymo's test environment as evience of its absence at its competitors?