Why is Tesla FSD a Big Deal Even Before Robotaxi?

Every Tesla FSD sale even with the price cuts is $8000 per car or $99 per month. The $99 per month is $1200 per year or $7200 for six years. Tesla has the highest profit per car of all car makers at $7000 per car. This is why Tesla FSD before robotaxi is a big deal.

Tesla FSD is the profit of a car with just the software.

Tesla is coming out with vastly improved FSD 12.4 and reportedly next level FSD 12.5. Those should be here as they get permission to rollout in China. Tesla needs to get permission to send the China driving data back to their 120 Exaflop AI training server cluster in the USA.

Every percentage of FSD adoption rate for existing Tesla owners in North America is 1% of 1.8 million Tesla owners who did not buy FSD when they first got their car. This is 18,000 car equivalent profits. There are about 2.2 million Teslas already in North America but 400,000 have already purchased FSD.

Every percentage of FSD adoption rate for new Tesla buyers in North America is 1% of 700,000. The Tesla sales are increasing. This is 7000 car equivalents in profits in 2024.

Every percentage of FSD adoption rate for existing Tesla owners in China is 1% of 1.5 million Tesla owners who did not buy FSD when they first got their car. This is 15,000 car equivalent profits. Virtually no one in China has gotten FSD. China matching existing US FSD adoption rates with better software would be a quick move to about 20%. This would be 300,000 car equivalent profits.

Why a Tesla Software and FSD Licensing Deal to Another Carmaker Would be Huge

Over half of the worlds EV are sold in China and there are 40+ EV companies there. Now any of them can sign a software, FSD, robotaxi deal with Tesla. Didi and several other Uber equivalents could sign a deal with Tesla.

Once the licensing wave starts that is not revenue until 2-4 years out but it means every analyst must model all-non Tesla cars as possible FSD profits for Tesla. What percentage do a deal with Tesla? What percentage buy Tesla FSD? They also most buy Tesla hardware to make it work and Tesla operating system software to make it work.

If the analyst 10% of all new cars would be 6-8 million cars/year worth of car profits each year. If those were all subscriptions at the current prices then that would be 1 million extra car profits each year. FSD prices do not go up with better versions. IF supervised FSD is shown to be 10-20X safer than without FSD then half of all drivers may be required to get a FSD and insurance bundle to make things safer. This would be a path to 50% adoption.

IF the supervised FSD data in China showed 10-20X more safety then China government could mandate Tesla FSD on all cars (Tesla and non-Tesla), this would be 25-30 million cars per year. China has 300,000 auto deaths each year and millions of injuries each year.

Car accidents cost the Chinese economy over $700 billion per year.

China has the most EVs in the world and will be making the most EVs in the world. Tesla getting a licensing deal with China EV makers could enable Tesla to get car equivalent profits for the 80% of the China EV market that Tesla does not currently have.

8 thoughts on “Why is Tesla FSD a Big Deal Even Before Robotaxi?”

  1. What am I missing? The $1200/year subscription is pure profit? Really? The AI is free? The vehicle’s connection to the AI and back is free? I’m thinking that there are costs and the profit is less. I’m confident Tesla will make a good profit, and hopeful the price will lower as time goes on.

    I wonder what will become of accident insurance (who to sue). The state tax on gas that supposedly all goes to road maintenance (ahem) will have to be collected. Life expectancy will trend a bit higher and lost-work-hours will be lower with FSD. And (being a private person) when will FSD be mandatory with government oversight of the data to keep track of subjects? There will be astonishing changes in society as this goes forward. Thank you for the updates, Brian!

  2. The China angle might be something.

    Chinese accident avoidance systems are commonly criticized in car reviews. Licensing there could solve a real problem.

    Or the software could get licensed for one generation then replaced by a local copy.

  3. I can imagine some pranksters having fun once everyone has self driving cars. How about towing a TV screen behind your car showing traffic lights, roads going over cliffs, aliens running into the road etc. Will AI get smart enough to see these as pranks or respond to them as real?

    • Lol, ai will indeed still have a very long time to go before they really get the exceptions we humans understand, and untill that happens nobody with a sensible mind will get into a taxi with no steering weel. Not after the first accidents viewing this problem happen.

  4. With AGI coming soon, carmakers will make their own FSD. In the same way GPT isn’t the only LLM out there.

    • Tesla’s crown jewels is all the data they have. Its hard to get hold of 1 billion miles of driving data.

    • AGI can `make` FSD then all you need is a few billion miles of `safe` driving data to supply to the regulatory bodies to actually prove it works its not enough to say it works you must prove it, then you need to build it into the cars you manufacture, possible but far from easy.

  5. “IF supervised FSD is shown to be 10-20X safer than without FSD then half of all drivers may be required to get a FSD and insurance bundle to make things safer.”
    This only works economically IF the insurance savings offsets the cost of the FSD. Maybe it will in China, where everything car-related is cheaper, but presumably insurance is already cheaper too, so there may not be enough enough savings in the liability part, or if collision insurance is not optional like it is in America, than in the two parts combined.
    China is going to hesitate to require a feature of a Tesla that’s going to cost drivers overall, and may push back at Musk to make FSD even cheaper than whatever discount from $99 he’s already going to offer in China.

    Look closely at the road safety chart. By far the highest percentage of fatalities – >55% – is coming from motorcycle riders. Tesla FSD will do nothing about that. They were called “donercycles” when I was still riding my Honda 450. I gave up motorcycles after crashing that one, luckily suffering no permanent damage even though the bike was totaled.

Comments are closed.