It's not that simple. I believe most folks are being lulled into a false confidence by Elon Musk.
.
What is your evidence for "most folks being lulled into a false confidence"? I say "most folks quickly appreciate the benefits and limitations of AP and stay within them." I have exactly as much evidence.
Look at that "Drunk Tesla" video. The system is nowhere near ready for the public to be Guinea pigs in that updated version.
I have, and the driver is not using the AP in a situation where it is intended, but he is using it while paying attention. There is only one line marked on the road and being detected by the system (as you can see on the dash display). When the car starts veering, you take control, just as the driver did. If you find this happening repeatedly, you stop using AP. That's how it works. Where there are 2 lines clearly marked, the car does very well, but even then we all know it's not perfect. If you drive expecting it to be perfect, you'll crash, but there are multiple feedback systems and warnings telling you that it's not perfect and that you must remain in command of the vehicle. It is clearly not designed for surface roads and that is made quite clear.
And unfortunately NHTSA is deferring to Elon. I believe this is because they have no exposure or experience in to other industries so they buy in to the mantra that this is the best way to do it. It is not. Not even close.
Again, where's your evidence? I read the NHTSA review and it was quite thorough, I don't see them "deferring" to anyone, and why would they. Hell, they won't even allow the freaking cameras to replace rear view mirrors! They have a lot more to lose by allowing AP than by restricting it.
The reason 6 billion miles needs driven is that most of it is repeat data. To get unique scenario data, accident data, in a method that requires folks to stumble on it, that many miles is needed. If the public driving around to gather data. ACCIDENT data, is the best method and Tesla needs 6 billion miles of it, what happens with upgrades or changes? Redo 6 billion or so miles to re-stumble on the regression scenarios?
There is a lot of value to all those miles and I can think of no better way to collect it. And you are totally making it up that it is "accident data" that is being collected--zero evidence for that at all. They do collect accident data, but not necessarily as part of the AP program, they've been doing that since well before AP.
Again I am for this technology and realize nothing is perfect. However there is a far better way that is actually much faster and puts far fewer lives at risk
Data so far suggests lives are being saved by the technology, can you produce one iota of evidence to the contrary? What are these faster and better ways? You throw that out there like it is obvious, but maybe I'm not too smart and need it spelled out. I don't see how you can get road data for Nixa, MO county roads and where the AP will have difficulty and needs improvement other than real-time data collection like they're doing.
I'm all for building a better safer mouse trap. But you told me off-line that you've never even driven a Tesla with AP and even said that you would not! You're totally barging in to this web site, screaming about the sky falling, and have zero first hand evidence, you've made multiple claims that you can't support and there is good evidence that your data is wrong and that lives are being saved and accidents avoided by the use of AP, even in it's current immature form.