As a sidebar, we'll discuss 'when FSD will be released as v1.0?' This is the date that you could nap in the car or watch movies as it drives you around.
In 2020, there was only one way to have the FSD feature in your car: you had to buy it. However, at customer request, Elon Musk committed that in early 2021, the option to lease FSD would be available (see tweet below).
This means that when the Robotaxi service starts, there are three FSD-related states a car could be in:
- FSD Purchased
- FSD Subscription
- No FSD
Some in the Tesla community have asserted that unless you are in category 1 or 2, you won't be allowed to participate in the Robotaxi network. With the subscription model, you can make a shorter-term commitment to FSD, but as I write this, the terms of the subscription have not yet been released (will it be annual, monthly, trip or distance-based...?).
I'm proposing that Tesla will want to allow as many owners as possible to participate. That would mean allowing cars in all three categories. But this brings up a dilemma, cars in the network will be in FSD mode, but category 3 owners have not paid for FSD. This is a solvable problem. For a car that has connectivity and advanced software, this is an easily solvable problem.
Here's how I propose it could work: When a car that has not purchased FSD or subscribed to FSD joins the Tesla network, FSD would be enabled while (and only while) the car is participating in the Tesla network. Additionally, there would be two owner payment tiers for vehicles that are participating in the network. For vehicles that have FSD, they would receive the highest tier payment. Vehicles that didn't already have FSD, would pay an FSD micro-lease payment with every trip they make.
For example, say a fair for a trip was $5. In the highest tier, some portion of this would go to Tesla (let's assume 40% or $2) and the remaining funds (60% or $3 in our example) would go to the car owner.
Well, if the car has to micro-lease FSD, then the percentages would be a little different. Let's see how the split might look for category 3 cars with our made-up revenue split of a $5 fare. Tesla network cut (again assuming 40% or $2), FSD micro-lease (assume 10% or $0.50), and the remaining funds (50% or $2.50) would go to the owner.
This allows all of the Tesla vehicles (that have FSD hardware) to participate in the network. This is important since you'd want a network with good coverage or it would limit the number of people interested in using it. It also rewards the Tesla owners that have paid for FSD with a little more revenue when they join the network, while still allowing people that have not purchased or leased FSD to profitably participate.
You might be asking yourself why Tesla should get 40% when your car is the one picking customers up and dropping them off. First, a reminder that 40% is just a number that I made up for an example, but Tesla will have expenses too. That 40% (or whatever the number turns out to really be) goes towards the app development, the servers to run the app backend, some profit margin for Tesla, and (the big one) insurance for your car while it's in the network.
There you have it, my reasoning as to why and how non-FSD vehicles will be allowed into the Tesla Robotaxi Network. Please let me know where you think I got it right and where I'm off base.
Before we go, I promised a sidebar on when I think the Tesla Robotaxi service will start. There are the nay-sayers that don't think FSD will happen this century. They are clearly wrong. The FSD beta is out now and there have been some impressive videos with it navigating tricky situations, so it is coming. The only question is when. On the other side, there are enthusiasts that think the steering wheel can be removed from Teslas before 2022. I fall in between these two camps. The FSD beta is impressive and this allows the long march of the nines to begin. The car will quickly be a 99% effective driver and this might seem like we're at "mission complete" but a crash every 1000 miles or so not good enough. To be a true FSD system, it has to be better at driving than the average human, better than the average taxi driver, even better than you. That will mean that it has to be at 99.99999%. It will have to be able to handle strange cases of trees falling across the road, things flying off trucks, double-parked cars, drunk drivers... The list of things that you handle by experience and intuition are difficult but not impossible for an AI to learn given enough training and corner-case examples. Considering all of this, my estimate of when FSD will be ready and legally allowed nationally in the US is 2027. I hope I'm wrong and it happens sooner, but if I had to bet, that's where I'm placing my chips.
Disclosure: I'm long Tesla stock
http://ts.la/patrick7819