He has a long history of outright lies that he presents as visionary ambition. For example, at the time he was promising that Tesla's with full self driving, available one year from then, would act as robotaxis and bring their owners enough money to recoup their investment in a year, he was obviously lying through his teeth - he knew full well that nothing like that will ever happen.

Similarly, his lies about starship being used as an earth-to-earth mass transportation system are equally transparent.

Not to mention the time he lied about financial plans for Tesla with the express purpose of manipulating the stock price, such as the Twitter thread that cost him his nominal role as chairman of the board.

I'm not a fan of Elon Musk, but I think his greatest flaw is hubris. I don't think he was consciously lying about things like Tesla FSD. I think he thought he would actually do it in the time he predicted but he overestimated his own abilities and domain knowledge.

Having said that, after so many past exaggerations one would hope he would have more self-awareness by now. I guess not.

> I think he thought he would actually do it in the time he predicted but he overestimated his own abilities and domain knowledge.

If he truly believed that a Tesla would recuperate its value in one year by the profit it would make as a robotaxi, he wouldn't have sold any Teslas, as they would make MUCH more money by running a robotaxi fleet than by selling cars.

That strategy gives you better margins. It doesn't give you better revenue or profits. The biggest problem Tesla had as a fledgling car company was cash flow, and the solution to that is scale and sales. FSD can be converted to immediate revenue by just charging for it, which is more cash-flow efficient than running a centralized taxi service. They have announced they have a dedicated rideshare vehicle in the works, and it's likely something of that nature was always in the works.

You're not totally wrong here. However the same is true of computers - a computer "fleet" makes much more money than the profit on the components, but Intel is not out there operating AWS-style datacenters.

I think there's such a thing as "core competency". It's possible that Musk views Tesla's core competency as building vehicles and vehicle software, not running a logistics-heavy operation like a robot taxi fleet.

Underestimating the complexity of autonomous cars, overestimating his abilities and domain knowledge, continuous lying about it (even if not consciously) sound pretty stupid to me. Like, textbook Dunning-Kruger.