Here is a quote from another thread which beautifully illustrates, from a legal perspective, the legal problems Tesla has:
The quote implies more than actually happened though. Tesla never claimed they had a finished product.
The nearest you get was them promising
level 2 city streets driving.... and I already agreed that was absolutely a valid basis for a potential lawsuit- but that any resolution was likely to be a tiny settlement to each buyer in that period, so long as they'd finally delivered that (at L2) by the time the suit was settled, which seems likely given it's already in beta testing.
It'd be similar to the delayed AP2 suit- couple hundred bucks per customers, and millions to the lawyers.
Most of Elons statements, in contrast to the definitive "L2 by end of 2019" one from Tesla during the sale process, are forward looking/speculative, generally protected by safe harbor rules.... and a lot of folks mix up "Statements made during sale" and "General speculative comments about the possible future of a product"
Also, this claim is flat out untrue:
"Tesla never said "we don't have a working version anywhere and there is no proof we ever will.""
They have always explicitly said any of the not-released-yet features would only be pending software validation to confirm safety in ADDITION to the regulation language....(and if we wanna pretend Elons tweets matter, he's mentioned various bits were NOT fully working yet tons of times too--- see his remarks on smart summon during the last year of development as a great example... "It almost doesn't suck right now" and so on)
The city streets code for example has "existed" for -years- but was not good or safe enough to release to anybody-- it's only now that software validation is approaching the stage they can release to a tiny group of beta testers.
A wider release will then serve as the start of software iteration and validation for higher levels of driving code.
Interesting that the first prediction is likely one of the most reasonable. It seems plausible (to me) that in around 200% of the estimated time, the prediction could be true.
As pointed out in post #8- they're kind of already there- arguably ahead of schedule.
Accident rate of drivers on AP is ~10x lower than the average human driver.
(there's statistical arguments to be made about this data, type of roads, age of fleet, etc... but at that point you're only arguing about if it's "really" 10x safer yet, or only 5x safer, or whatever the real x number is)