Elon Musk: Always A Douche

[Admiral Yi]

Elon apparently claims that Tesla is not a car company but an AI company. I came across a thing that said that Tesla stocks is being evaluated as such.

So perhaps his strategy is to continually pivot Tesla onto new bubbles as they appear? Bubbleriding, as it were.

Or perhaps his strategy has always been to look for additional high potential ways to monetize the car space.  There's a difference between a bubble, in which people pay no attention to the fundamentals and care only about whether some other sucker will bid up the price, and growth stocks, in which the fundamentals come into play at some point in the future and it's very difficult to model the future with any precision.  For example with AI, no one really knows what its potential upside is.  But I think it's self evidently immense.

[OttoVonBismarck]

I am skeptical of Tesla's valuation being justifiable even based on assessing it as a growth stock. The economics of the car market are damn well understood, and it is well understood companies that make cars simply aren't as valuable as Tesla is.

I think it is all but undeniable that a significant portion of Tesla's valuation is simply speculators who buy into Musk, and expect or expected him to continually find ways to keep juicing the stock up. It has never meaningfully moved in relation to the company's real fundamentals, and it is incredibly unlike a company that isn't a major player in software is meaningfully positioned on AI. The big software players are all already there, and they have all the top researchers and software guys in that space.

Musk isn't going to build a Microsoft or Google-class software company to then stand up an AI effort.

[Zanza]

Elon apparently claims that Tesla is not a car company but an AI company. I came across a thing that said that Tesla stocks is being evaluated as such.

So perhaps his strategy is to continually pivot Tesla onto new bubbles as they appear? Bubbleriding, as it were.

Or perhaps his strategy has always been to look for additional high potential ways to monetize the car space.  There's a difference between a bubble, in which people pay no attention to the fundamentals and care only about whether some other sucker will bid up the price, and growth stocks, in which the fundamentals come into play at some point in the future and it's very difficult to model the future with any precision.  For example with AI, no one really knows what its potential upside is.  But I think it's self evidently immense.

AI is an area where Tesla so far has not delivered. They take much more risks than other carmakers with their driver assistance systems, but they have no significant technology advantage over their competitors. At least not yet. Their latest innovation in that area is an own supercomputer which is supposed to bring their autonomous driving capability forward. Let's see.

I think the single most meaningful innovation of Tesla was its supercharger network. That has even set an industry standard in North America now. But physical infrastructure is very capital intensive at best comparable with tech company investments into data centers. But the valuation of the tech companies is from their software, not their hardware.

[Admiral Yi]

AI is an area where Tesla so far has not delivered. They take much more risks than other carmakers with their driver assistance systems, but they have no significant technology advantage over their competitors. At least not yet. Their latest innovation in that area is an own supercomputer which is supposed to bring their autonomous driving capability forward. Let's see.

I think the single most meaningful innovation of Tesla was its supercharger network. That has even set an industry standard in North America now. But physical infrastructure is very capital intensive at best comparable with tech company investments into data centers. But the valuation of the tech companies is from their software, not their hardware.

When I said the upside of AI is immense I was not referring to Tesla specifically but rather AI in general.  I haven't the slightest idea how AI adds value to Tesla.

I do know that Tesla has managed to produce software upgrades that customers are willing to pay for.  Did Tesla invent that?  You would obviously know better than me.  But that is not part of the traditional auto revenue stream.

As Cathy Wood put it, with Tesla you're buying a car company and you also get call options on things like self driving and battery tech etc.  The correct pricing of those options is of course completely up for debate.

I also think it's self evident that the EV share of the total market will only continue to grow.

[Barrister]

Elon apparently claims that Tesla is not a car company but an AI company. I came across a thing that said that Tesla stocks is being evaluated as such.

So perhaps his strategy is to continually pivot Tesla onto new bubbles as they appear? Bubbleriding, as it were.

Or perhaps his strategy has always been to look for additional high potential ways to monetize the car space.  There's a difference between a bubble, in which people pay no attention to the fundamentals and care only about whether some other sucker will bid up the price, and growth stocks, in which the fundamentals come into play at some point in the future and it's very difficult to model the future with any precision.  For example with AI, no one really knows what its potential upside is.  But I think it's self evidently immense.

AI is an area where Tesla so far has not delivered. They take much more risks than other carmakers with their driver assistance systems, but they have no significant technology advantage over their competitors. At least not yet. Their latest innovation in that area is an own supercomputer which is supposed to bring their autonomous driving capability forward. Let's see.

I think the single most meaningful innovation of Tesla was its supercharger network. That has even set an industry standard in North America now. But physical infrastructure is very capital intensive at best comparable with tech company investments into data centers. But the valuation of the tech companies is from their software, not their hardware.

To the extent Tesla is an "AI" company it is in the area of self-driving cars, which certainly is an area of AI technology.  I do know Tesla has sunk a tremendous amount of resources into that area.  Whether that has given them any kind of technological edge over its competitors is way beyond my knowledge.

*IF* Tesla was able to perfect true self-driving well in advance of its competitors that would be a huge competitive advantage.  I'm so far unimpressed.

[Zanza]

When I said the upside of AI is immense I was not referring to Tesla specifically but rather AI in general.  I haven't the slightest idea how AI adds value to Tesla.

There are lots of potential AI use cases in a car. Autonomous driving is the best known, but you also have navigation, infotainment, predictive maintenance, other assistance than the core driving and of course AI powered subsystems like intelligent high beams etc.
But Tesla has not really shown significant advantage over their competitors in these yet. They take significant risks with their autonomous driving, which is why their systems seem more capable than others even if they are not from the underlying technology.

I do know that Tesla has managed to produce software upgrades that customers are willing to pay for.  Did Tesla invent that?  You would obviously know better than me.  But that is not part of the traditional auto revenue stream.

The actual innovation here was not to sell software based services, but having an own operating system on own chips, which allowed OTA updates. Traditional car companies had typically left the software in distributed control units provided by suppliers. That made OTA updates virtually impossible as it is very hard to test all HW/SW configurations of a car.
But even in that area Tesla is quickly being outshone by Chinese competition - at least in China itself, which is the biggest car market by some margin.

As Cathy Wood put it, with Tesla you're buying a car company and you also get call options on things like self driving and battery tech etc.  The correct pricing of those options is of course completely up for debate.

Sure. Let's see if those options come to fruition.

I also think it's self evident that the EV share of the total market will only continue to grow.

Yes, even if not as fast as many predicted two or three years ago.

[Grey Fox]

I expect the first real self-driving car to be marketed to every day people will be Apple's.

[Barrister]

I expect the first real self-driving car to be marketed to every day people will be Apple's.

"Project Titan" is coming up on being 10 years old by late 2024.

[Jacob]

Did some people pre-pay for the self-driving feature?

[HVC]

Did some people pre-pay for the self-driving feature?

They did. 5k IIRC. There's a class action lawsuit out there.

[The Minsky Moment]

The reason Elon is making these claims is that that automotive sales, both present and projected, are far too insufficient to support the current stock valuation.  The current valuation is only supportable if "Full Service Driving" (FSD) completely succeeds and brings in multibillion revenue amounts by 2030.

But while Tesla has managed to earn some subscription revenue off of FSD - perhaps in the 1-3 billion range - at present it appears that the chief beneficiary of Autopilot/FSD is the class action plaintiff's bar, and underemployed regulators at the NHTSA.

Tesla's key competitive advantage has been range, but that advantage has been eroding. Lucid and some of the Mercedes range have passed Tesla - these are very pricey models but the trend is concerning.  Tesla isn't going to collapse but it faces increasingly tough competition in the space.

The revenue potential of FSD depends on the capability matching the name.  To validate the valuation everything has to go right, namely:
+ Self-driving cars must reach the stage where they are actually self-driving and do not require full human attention and both hands on the wheel.  Otherwise it's just glorified cruise control.
+ Tesla's implementation of the tech must lap the field to the extent it can earn significant licensing revenue as well as inhouse subs.

Both propositions seem questionable.  True self-driving autos are the transportation equivalent of nuclear fusion - the tech is always X years out in the future.  Five years ago analysys were saying the tech would be ready by now; now they are saying it's another 5 years.  At the same time, there is still a lot of competition and I've seen nothing to suggest that Tesla has overwhelming technical or developmental advantage.

[PJL]

Both propositions seem questionable.  True self-driving autos are the transportation equivalent of nuclear fusion - the tech is always X years out in the future.  Five years ago analysys were saying the tech would be ready by now; now they are saying it's another 5 years.  At the same time, there is still a lot of competition and I've seen nothing to suggest that Tesla has overwhelming technical or developmental advantage.

Any bets on what will become available commercially first? Honestly, I think nuclear fusion power might actually beat self driving autos.

[crazy canuck]

Both propositions seem questionable.  True self-driving autos are the transportation equivalent of nuclear fusion - the tech is always X years out in the future.  Five years ago analysys were saying the tech would be ready by now; now they are saying it's another 5 years.  At the same time, there is still a lot of competition and I've seen nothing to suggest that Tesla has overwhelming technical or developmental advantage.

Any bets on what will become available commercially first? Honestly, I think nuclear fusion power might actually beat self driving autos.

The AI folks I know say self driving is a long way off and may never be practical. 

The same might be said for fusion.

[Grey Fox]

I expect the first real self-driving car to be marketed to every day people will be Apple's.

"Project Titan" is coming up on being 10 years old by late 2024.

It's coming. Apple's buying all kinds of IR sensors.

[Baron von Schtinkenbutt]

The AI folks I know say self driving is a long way off and may never be practical. 

The same might be said for fusion.

FSD is a very complex problem.  Road infrastructure is already bad for human drivers in most places and needs to be improved for them (but isn't).  I think it's impossible to make any machine with less than human-level intelligence and sensory capability work in the current environment, and I don't think the promise of FSD is enough to motivate major infrastructure improvements that are already needed but not happening.  There's also the resistance to car culture in general.  While that doesn't directly affect the development of the technology, it does affect the will to implement needed infrastructure changes to support its deployment, as well as the will to allow experimentation on actual roads.  In short, FSD is a hard technical, political, and social problem.

Nuclear fusion, on the other hand, is largely a very hard technical problem.  It's possible there could be political and social problems created by a specific implementation down the road (specifically with regards to safety), but actually achieving sustainable fusion is purely science and engineering (and the willingness to fund it).  I suppose it's possible that the fundamental physics just won't allow a self-sustaining reaction at the relatively microscopic scale we'd need for power generation.  Barring that, though, I thik it's actually more practical than FSD.