What's your model of rational regret? Imagine that Tesla solves level 4 autonomy tomorrow and 10x's the subscription price. So far it's looking very good for your lump sum payment. Maybe that was even predictable and everyone who didn't purchase FSD was genuinely being stupid. Then, still imagining, the day after that a meteor destroys your car while you're sleeping. Now you "regret" your purchase, but only in hindsight. People predicting NO in this market were the ones who were correct.

So that's the question: Will you evaluate the decision ex ante or ex post? Ie, do you ask "was this the correct decision given what I knew at the time?" or do you ask "was this the correct decision knowing what I know now?"?

@dreev good question, i’d do ex post, which is “was this the correct decision knowing what I know now?”. The markets purpose is to evaluate whether I would be happy with my current decision in the future. A disaster that destroys my car will likely make me not happy with the decision because fsd is tied to the car, however, this seems like a low probability scenario. So far, I am very happy with the purchase as it made my road trips much easier.

Longer timeframe

Why would you pay $8k upfront when you can pay $99/month? It would take almost 7 years to start having a positive value, and that's not counting the opportunity cost of investing the $8k. Also you don't even get to keep the FSD if you crash and total your car. It seems like an abysmally terrible deal to me.

@xjp Well you are getting 15% yield per year on the 8k if you treat the $99/month as the "yield" from the purchase, which is a pretty good ROI compared to most investments. Elon mentioned the subscription price may increase in the future. And the -EV from crashing/totaling the car is relatively low due to low odds. After considering these factors, I felt it was a good purchase at the time, despite being impulsive.

I am not sure if I will feel the same way later, though, if I were to actually crash/total my car at some point in the future. Or if they require additional subscriptions for more features in the future.

@Rice It can't be considered anything like a 15% yield because unlike e.g. a bond at the end of its lifespan you don't get your $8k back, you get nothing. It's a sharply depreciating asset.

GPT says that factoring in a 5% interest rate you would have to drive the car for 8.3 years to merely break even. Add on top of that the risk of crashing and the uncertainty of technological change or change in life circumstances and there's no way it's worth it in my opinion.

By technological change I mean, for example, what if in a few years Tesla's pure vision FSD has failed and competitors offer you level 4 autonomy for cheap? Or for that matter, even if Tesla succeeds at level 4 what if market competition forces them to offer it for free or much cheaper to sell cars?

@xjp I don’t necessarily agree this should not count towards the yield as there is some inherent value to the technology whether it depreciates or not. Sure, Tesla could fail and this technology becomes obsolete, but it’s also possible Tesla beats out competitors and this technology becomes more expensive to acquire instead of cheaper because there’s so much more demand. There’s risk and reward associated with this purchase similar to other investments. If the failures you described comes true though, then I will certainly “regret” my purchase. Maybe I ll create longer term markets for this.

@Rice things don't get more expensive with time, they get cheaper, esp. with something like FSD, where you're "unlocking" a feature. There's no supply shortage.

@Vergissfunktor Subscription services don't always get cheaper, the price is dependent on demand.