you get cool driver assisted features and the promise by the CEO that one day "soon", you will be able to make tens of thousands of dollars by using your car as a autonomous driving taxi service. There is no current deadline/timeline for when this tax service will materialize.
This may be an important point. If you want to rent out your car, or several cars, as robotaxis, then they're a business, and you have a choice between hiring drivers (or working as a taxi driver yourself, if it's just one car) and paying for level 5 autonomy. In this scenario, level 5 autonomy is probably worth at least the $15,000 that it will soon cost, since that would be
much less than the cost of a driver over the lifetime of the car.
For a typical personal user of a car, though, I'm not at all sure that even a full level 5 version of FSD is worth that much money. In fact,
most people couldn't afford it; the cheapest Tesla today is the SR+ Model 3, which would be $61,990 when equipped with FSD at its soon-to-be price of $15,000. That's awfully close to a full year's income in the US; in 2019, the
median income was $68,703. Tesla's current customer base is more affluent than average, although I don't have figures at hand, so many of us can afford to throw away money on the FSD package, but that still doesn't make it a good or appealing deal. As I understand it, the FSD take rate has been dropping since 2019, probably because of a failure to meet promises and rising prices.
Thus, whether even a fully-functioning level 5 autonomy system is "a bargain" (or whatever Musk called it) at $15,000 really depends on the future of transportation. If we replace personal automobiles with robotaxis, then an expensive level 5 package may be worth a lot of money, and help those who get in at the "ground level" of operating fleets of robotaxis get rich. If the personal automobile retains its place as a fixture of modern life, then such an expensive system will never take off. It'll either never emerge as a viable technology or prices will drop (compared to what Tesla is currently charging). Even if you believe that you'll be able to rent out your Tesla as a robotaxi in the future, the timeline for all this is important. Supposing a 10-year usable life for a robotaxi (which I'm guessing is long, since taxis see a lot of miles), if it takes 15 years before the technology is approved for use, you'll reap no financial rewards from buying it now. If it comes to pass in 5 years, then you'll get 5 years of useful rentals, and you'll almost certainly recoup a $15,000 investment today. Renting one car, though, would likely be a side hustle for most current owners. I suspect that most robotaxis, if and when they become real, will be owned by fleets -- Uber, Lyft, Hertz, etc.
Personally, I'm skeptical that Tesla will be able to turn current-production Teslas into cars with level 5 autonomy.
Maybe it'll happen, but I'm not going to gamble $12,000, much less $15,000, on that happening. Even if it does happen, I'd be quite surprised if it were to happen in five years or less. I've seen steady but slow progress on the problem that falls
well short of Tesla's promised timelines. Even if the next version of FSD was
much better, there's the question of regulation, which involves government bureaucracies --
state-by-state government bureaucracies, unless Congress were to pass some overriding law, if I understand correctly.
All of which is to say that Tesla's current pricing is high. Charging $15,000 when they're on the verge of getting legal approval to run a robotaxi service might make sense, but they're pretty far from that.
Another point is that, because there's such an economic disparity between robotaxi use and personal vehicle use, I could see a pricing model that distinguishes between the two. That is, I as a private individual who only wants the car I own to drive me places, might pay some lesser amount for the feature (say, $5,000), whereas somebody who participates in Tesla's robotaxi fleet might need to pay much more (maybe $25,000). Another way to do it might be to pay by subscription, with a price charged per mile that the feature is active. This sort of differentiation is possible given the way Tesla's software works, and it's entirely possible we'll see such a split in the future, if and when robotaxis become a reality (or close to it).
Finally, there's the question of what the competition does. Other companies are working on level 5 autonomy, and some are even operating limited (geofenced) prototype services. If GM, Ford, VW, or whoever releases a level 5 system as an option on their cars for $5,000, then Tesla will have a hard time charging significantly more than that. Right now, AFAIK, Tesla's FSD-on-city-streets beta program is the closest anybody has to a system that's for sale for the general public, so Tesla is able to charge what they want (albeit with dropping take rates as they raise the price to ridiculous levels). I'm guessing that Tesla is using this income to fund R&D on FSD, whereas other companies are funding it in other ways. Once there are final products and competition between them, market forces will determine the price.