This is a future that I believe I can see with some considerable clarity. I was discussing Tesla's ability and intent to roll out AHAAS (Autonomous Hardware As A Service) or AI EV fleet as I would term it long before Jonas of Morgan Stanley first queried Musk about it. For a point of reference Jonas thinks this is a good reason for TSLA to have a PT of $465.
IMO it is too early for AI EV fleet to determine or affect a current PT but seeing as you asked, the ability to operate a service like this completely destroys the economics of gasoline and internal combustion. Musk is correct by stating that a non AI car will be like a horse - sentimental value only - but to put a finer point on it, this assumes that the AI car is an EV. The reason for that is because the AI EV will have a cost per mile including amortization of the vehicle of approximately the cost of gasoline with no car, no maintenance and of course no driver. The cost of Gasoline alone kills the ability to compete - even at $1.80 per Gallon there is nothing left to buy a car to run on it if you want to compete with a Tesla AI EV in a fleet setting. Musk put it succinctly in the recent BBC interview: 'Electric Vehicles and Autonomy are the two biggest innovations since the moving production line and they are both happening in parallel'.
As stealthology suggested "I think the majority of autonomous driving in the future will be SAAS" (I guess he meant AHAAS) - yes that is true and here is why.
The AHAAS AI EV will extend private transportation in cars to people that could not previously afford a car (which is a lot of people). It will also serve anyone that is not particularly interested in spending money on cars and would prefer to spend their money on something else - which will be possible with no loss of freedom to get from A to B in your own privacy whenever you want either on your own or with whoever you want (and without the risk, cost and intrusion of a stranger in the driver's seat - or the loss of a seat).
Even with such a service in place representing effectively unlimited demand for what Tesla can make and no existing car maker can compete with - there will remain demand for exclusive use of a private car. If you can afford it and you want to spend your money that way, the market for the luxury of individual ownership will exist and a market for luxurious types cars will exist too.
There will also be a blurring of the boundaries. One of the perks of privately owning a Tesla in future will be the ability to turn it over to Tesla's fleet management while you are at work, asleep or out of town. Then it can start paying you back for owning it. This ability to earn value from a car that is primarily 'yours' will I think give rise to novel leasing structures too.
Considering such a AHAAS / AI EV Fleet service will start off scarce and proceed over a period of years and hundreds of millions and ultimately billions of vehicles before saturating demand it would make sense to offer it as Uber is known to do in a service marketplace where price is set by the market. Considering Uber is charging typically $2+ per mile and Tesla will be operating at a cost of $0.06 cents per mile the profit potential of this service is utterly phenomenal while operating on a cost basis that undercuts any other means of getting around except perhaps walking (even that is questionable - how many miles can a person walk for the cost of one Snickers bar?). A second hand ICE car costing $500 definitely cannot compete with a $40,000 AI EV on a cost per mile basis after fuel and maintenance and riding in the $40K car will be much nicer. This is significant because the AI EV will not just crush the economic viability of new ICE production it will also hasten the legacy ICE fleet to the scrap yard. This in turn will eviscerate residual values of ICE vehicles and hasten the lost viability of purchasing one. This will literally trigger a vortex-dump of the Oil and ICE paradigm decades sooner than the majority of people appreciate.
Competition. I don't think competition will come from the existing majors. I think instead they will go bust. There is no realistic way of getting out of ICE manufacturing or the dealership business model as fast as this will occur and an AI EV fleet has no need for middlemen to process sales and is most unlikely to benefit from the existence of a third party service industry. In Tesla's case, no. The service function needs to be networked into the fleet and to be incentivized to reduce fleet costs, not profit motivated to increase them.
In the USA incumbent auto makers cannot even deploy an OTA network because it is against franchised dealership rules (GM Bolt helped clarify that one) - and they cannot get out of the dealerships. It is completely impossible to operate an AI EV fleet without an OTA network, a service network and a charging network - Tesla has all of these, ICE auto makers have none of them. For example you cannot have Autonomous cars roaming the streets with passengers in them without continuous remote service status monitoring. Instead competition will come from Apple, Samsung, LG, maybe Faraday + collaborators, and from other well-funded startups. The production lines of today's big auto I think will be up for grabs in Chapter 11 and similar around the world but all of these competitors will need batteries and in that regard I think Tesla will look a lot like Saudi Aramco does today (the world's largest oil company). Tesla: The world's supplier of Gigafactories.
So bottom line reality: The actual Tesla business plan that it is on track to execute is ridiculously more positive for a net present DCF valuation than even the most forward-thinking TSLA Bull has ever suggested, except perhaps TSLA's largest shareholder.