One of the odder things about computers, which confuses people who aren't used to them -- the hard part is specifying the problem correctly. Once you've specified the problem, as a programmer, you have practically solved it. (A traditional program is essentially a specification.
...
This accounts for people's confusion about the way progress arrives in computing. They think the hard part is solving the problem but the hard part is actually defining the problem, and the actual solution part goes quickly. . .
Most of this post is correct in my opinion, but you're limiting the case to computing, when the same logic applies to the vast majority of human problems.
Quoting myself from my thesis of more than 50 years ago "statistics is what you use for prediction when you do not know the causes..."
Bluntly, in my view almost all human understanding fails because of lack of definition. After all humans once attributed all infectious diseases as caused by the wrath of God. Because people mdid not define the problem correctly they wasted millennia with dirty water and bleeding people to cure diseases.
Elon Musk managed to make recoverable parts in Spacecraft because he did understand the problem.
Although other people had the concept fo electric cars Elon understood that sex appeal was the crucial missing element. All the endless list of technological achievements were needed to deliver the Model S.
So, in FSD we have a [problem that, in my opinion, Elon understands very well. he is, as usual, overstating speed and simplifying the achievements needed to actually deliver anything remotely like true Level 5. OTOH, they very well might deliver Level 4 with numerous geo-fenced areas with less than 100% availability within those areas, reducing a vast number of necessary edge cases.
That sort of functionality might well be deliverable in a year or so.
As usual in human advance the first market entries are quite limited and continuous improvement. We need not list all the advances.
Somehow in this thread and elsewhere we have completely lost our grasp of reality. Neroden tells us it will be ten years and that the Tesla approach is wrong. Somebody else listens to Elon and actually assumes that he is saying Robotaxis will go anywhere at any time late nexct year and that private cars will be a thing of the past. I cannot wait to take a robotaxi from Beijing to Lhasa.
Realistically we will have highly compact, well organized, perfectly signed and lighted areas served by rototaxis next year. Easy to predict that because such is already happening in Singapore, possibly the best organized urban road structure in the world. We must all remember that we cannot assume Elon Musk actually thinks everything for everybody will happen tomorrow. It will be revolutionary enough to provide level 4 in some major cities in good weather conditions in areas without major construction or other impediments. Places like Moscow, and even parts of places like Rome could have such operations quite soon, not that either one would be first on anybody's list.
Would we not be more wisely evaluating the impact on Tesla and our investments of a gradual entry in areas with congenial surroundings geographically, governmentally, societally and economically?