I make the case for large Tesla profits from very good FSD (Full Self Driving) in an interview on Brighter With Herbert (youtuber Herbert Ong).
Full Self Driving is the name of Tesla software for its driving assistance system.
My case is that Tesla FSD is vastly improving and as good version of 12.4.X and then 12.5 can get to smooth driving with over 2000 miles before interventions. This will be the level where old and new buyers of Teslas in the US and China will have a significant percentage purchases.
This is hugely significant because each full purchase of FSD is the profit of a car. You buy a Model Y for $40,000 and Tesla gets $8000 of profit. Tesla sells FSD for about $8000 and that is the profit of a car. A monthly subscription even at $99 per month will get the profit of a car in about 6-7 years.
When Tesla FSD gets to a threshold of quality where 10% of Tesla owners adopt it in the US and China that would be 400,000 sales or subscriptions and adding 200,000-400,000 each year and growing.
The other half of the case is that Tesla FSD at robotaxi quality will impact Tesla profits before the robotaxis are regulator approved and before the removal of the safety driver.
IF Tesla did not take any money from the operation of the robotaxis in China and just had the car sale and the FSD sales, Tesla would get $16000 of profit per robotaxi with human driver.
In 2025, if there were 20% adoption of FSD this would be $10 billion of profit from FSD sales or $1.5 billion from 1.8 million FSD subscriptions.
At the end 2025, if there were 110 cities mostly in China with robotaxi trials with human safety driver averaging 5000 cars per city this would be 550,000 more car sales with FSD. This would nearly double Tesla valuation from $777 billion today.
In 2026, if there were 250 cities mostly in China with robotaxi trials with human safety driver averaging 10000 cars per city this would be 2.5 mmillion more car sales with FSD. This would 4X Tesla valuation from $777 billion today.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
I think you’re too close to your subject. I came to the article via a link from Instapundit. As I clicked your link I had no idea what FSD meant. After reading your article I had a hunch what FSD was.
I had to do a google search to find out it means Full Self Driving. My recommendation is to put that full explanation in the first paragraph of your articles for us newbies.
HINT: When using an acronym that is not in world-wide usage or recognition, you should do your readers a favor and spell it out the first time.
Two things Brian:
It’s extremely unlikely that Tesla will make 4 million cars in 2026. No new factories are being build right now. It takes 2 years to ramp production
FSD SW has shown no improvement from 11.4 to 12.3.6 as measured in miles between interventions according to Tesla FSD tracker. It’s at 28 miles between interventions. The 12.4.3 seems to be at about the same objective level as 12.3.6. Ie it’s very unlikely that the take rate will jump above 10% any time soon. Let’s hope that the new computer cluster will improve things..
All robotaxi profit projections hinge on people choosing to use them, and the service being priced at the assumed level.
Waymo is offered to the public in a few places, for free in some markets. There is some adoption. It isn’t huge.
Robotaxi getting to huge adoption and profits requires that people choose to use them, and requires that the service is priced at a profitable level. If the price is high enough to achieve the forecast profits, will the price of service be low enough and quality of service be good enough to get mass adaption?
These are obvious statements. Ignoring them doesn’t make them go away.
True, and in majority of countries there will still be legal problems, possible extra takes and of course labour parties. I’m not getting in one for the first ten years as well.
If the insurance liability companies gave Teslas with FSD a discount on their premiums, that would mean a whole lot more than whatever claims between interventions there are, even assuming people believed those claims.
The videos of out-of-control Teslas crashing are all over 60 Minutes stories, the internet in general, and won’t be unseen by an incremental version change. It actually looks like Musk is perpetually trying to reset the clock with all these version changes.
Last time I checked it cost MORE to insure a Tesla than a comparably priced ICE car, mostly because any damage to the battery totals the whole car.
When insurance companies take a chance on lower liability premiums, people will notice.
What do they mean by driver intervention? Is it a “I am going to crash type” or the road conditions ahead are not something the car can handle and the driver must decide how to proceed.
I don’t see how these trackers could be useful because I’m constantly turning FSD on and off. It handles boring routine driving and I drive the fun bits. Every time I turn it off it pops up this thing asking why I intervened and I can leave a voice tag. I say “didn’t like lane position” or “the truck we’re catching up to is drunk or something and I’d rather drive it myself” or whatever. If it was an expected thing I won’t say anything. Sometimes the car drives 100 miles by itself. Sometimes I just wanted to turn it on while I got a stubborn drink open for a few minutes. It doesn’t really have anything to do with the software quality.