AJ Investment Research@alojoh
I think the best leading indicator for Tesla FSD adoption is how many miles people who own it drive it daily since:
1) If people were unhappy they would drive it less and fewer people would sign up for it.
2) It's inconceivable that usage would go up if people were not happy with it since...
3) Attrition rate of existing subscribers would outpace new subscribers (since existing subscribers span all vehicle cohorts and not just current quarter).
4) Propensity to buy/sub FSD is likely vastly higher for people at time of purchase a new Tesla (at least currently).
5) I think the earliest adopters (people who will buy anything shiny) has been baked into these numbers for long and this bucket won't be able to contribute as high a share as it may have in the past. Hence, the latest increments cannot be explained by these earliest adopters. This would suggest Tesla starts to penetrate other - and most importantly - much bigger customer groups.
Based on Tesla's disclosure I estimate Q3 saw not only a new all-time high but a significant sequential increase from 9.5 million miles driven daily on FSD to 13.2 million miles daily.
Additionally, I estimate based on Tesla's disclosure and comments during the earnings call that Tesla FSD is driven 12-13* million miles daily in October.
1) Using a good old pixel count we can see Tesla accumulated about 5.75 billion FSD miles as of September 30, 2025. (see attached chart)
2) Ashok stated during the earnings call that "as of yesterday" (call took place on October 22, 2025) Tesla hit the 6.00 billion FSD miles mark.
3) This implies that between Q3 end (Sep 30) and Oct 21 (21 days) Tesla accumulated another 0.25 billion miles.
4) This implies Tesla is currently adding daily about 12-13* million FSD miles.
Bottom line: this data implies a significant uptick in adoption is imminent.
The next part is for the nerds:
I think a useful mathematical model (although not perfect) would be that of diffusion in physics with regard to particulate count where one can describe the spread of particles from regions of higher concentration to regions of lower concentration due to random motion. Applying this model to FSD adoption we can see how a higher concentration (higher daily usage) drives (diffuses) incremental adoption.
The weakness of this high level analogy is that adoption is not random, market saturation is nonlinear, and - unlike particulates eg - attrition doesn't exist (except for some edge cases).
[*seasonal factors need to be accounted for since the U.S. has several peak driving seasons]