You might want to take your own advice. If you were listening to Elon's words you would know that he's considered autonomous driving to be a solved problem for a long time.
See
this old post for a selection of assertions from Elon from 2018 and earlier
he also thought he'd be able to do a cross country drive very shortly then too.
And the year after.
And so on...
Elons been provably wrong pretty often on this one.
In fact, the post you link to is example after example after example of
proving exactly that
Or to sum up- self driving is, factually,
not a solved problem as of today
OK, this is semantics now, but by solved I mean Tesla cracked the technique, cracked the methodology, cracked all the key ingredients, there are no potential showstoppers.
But again- this is factually untrue.
They THINK that will be the case.
They also thought there were no showstoppers with HW2.0 and the first gen HW2 camera filters. And said so in public.
They also thought there were no showstoppers with HW2.5 and the 2nd gen HW2 camera filters. And said so in public.
They also thought there were no showstoppers with HW3 and the previous-gen code branch they ended up having to to a large scale rewrite on when they found local maximums.
They won't
know there's no showstoppers until it's working.
It's not yet.
(and Green has mentioned several other potential ones, largely HW limitations with things like frame rates, blind spots, etc, that might still come into play).
Again it's not a solved problem yet.
And Elon has told us we shouldn't believe him when he gives a date on something being done until he's actually done it.
Elon Musk on 60 minutes said:
People should not ascribe to malice that which can easily be explained by stupidity." (LAUGHTER) So-- so it's, like, just because I'm, like, dumb at-- at predicting dates does not mean I am untruthful. I don't know-- I-- we've-- I never made a mass-produced car. How am I supposed to know with precision when it's gonna get done?
Sub "FSD" for "mass-produced car" and same deal.
Or "reusable rocket that lands upright"
As we've seen, once he does it the first time, he's excellent at getting even faster and better at it iteratively.
But telling us exactly when it'll be ready the first time? Not so much.
So if they get it solved in the US (which it'll almost certainly be, every new advanced driving feature was released there first so far, likewise the current FSD beta) it'll roll elsewhere surprisingly fast.
But it's certainly not solved
yet