I think the software component is truly underestimated. It's been said that Tesla is really a software company that happens to build cars, and I can believe it. I'm not sure that VW, GM, or Ford will ever be able to compete on the software front, they just don't have it in their DNA.
I kinda view it like Silicon Valley; there's nothing intrinsic about the specific place on the map that makes it a great place to develop software. But other places just can't seem to replicate it. I've heard talk of Silicon Alley in NYC but it doesn't seem to compare. I'm not sure it's possible to recreate SV somewhere else, just like I'm not sure it's possible to recreate Tesla inside GM.
There is a really good example of this in the Q1 report. The chip shortage that has hit a lot of car manufacturers also hit Tesla, but they quickly changed hardware and developed new firmware. Only a software savvy company can do that.
> Diminishing returns come much faster and harder for software in cars. Software not being the main focus of a car.
That used to be true. With self-driving technology I'm not sure that's true anymore. Tesla has even offered hardware updates to improve self-driving of older models. I haven't heard of anyone else doing that yet.
> Especially now with Android Auto/Apple CarPlay.
Eh. That's just infotainement, and cars can still provide significant value without using that. When driving a Tesla I don't miss CarPlay, and I rarely use it in my Kia even though it's better than the built-in infotainment stuff. Mostly just for navigation if I know it's a complex route. But again, I never miss CarPlay for navigation in Tesla cars because the built-in system is very good and always up-to-date through OTA.
> That used to be true. With self-driving technology I'm not sure that's true anymore. Tesla has even offered hardware updates to improve self-driving of older models. I haven't heard of anyone else doing that yet.
Nobody has self-driving tech in production. What they do have is a combo of adaptive cruise control, lane keeping assist, lane changing assist and park assist.
Full self-driving tech is still at the basic research stage, just like cold fusion reactors. It could be out in 5 years, it could be out in 40. Nobody's going to "OTA" that.
Excuse me, what "self-driving" is that? For all of Musk's grand promises of "self-driving in 6 months time, just you see", nothing has ever happened. Why people take anything that conman says at face value is beyond me.
Do you really want software auto-updates with potentially breaking changes in your car though?
Chances are, you will end up with the iPhone situation, where a 10-year-old phone is now essentially a brick - the software 'updates' just made things slower on the older hardware, if they support the older hardware at all. Do you really want that in your car?
For comparison, you can get a 20-year old car today that would work today as well as on the day it was made - mechanical controls still respond instantly, radio still works.
And then suddenly you find that the Teslas are a lot more expensive than they seem, because the depreciation will be more severe. Not only that, but I would question their green credentials if the cars are forced to become obsolete after N years.
I kinda view it like Silicon Valley; there's nothing intrinsic about the specific place on the map that makes it a great place to develop software. But other places just can't seem to replicate it. I've heard talk of Silicon Alley in NYC but it doesn't seem to compare. I'm not sure it's possible to recreate SV somewhere else, just like I'm not sure it's possible to recreate Tesla inside GM.