Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[Mengy]

It's both.

Auto margins are rather small. Software margins are huge.

Anybody can make a car, but the overhead is massive. It's a very tough business.

If only Tesla has the software for Ford's robotaxis, then Tesla gets most of the profits. Only Tesla can make Ford's robotaxi viable.

Even if robotaxi never happens and Tesla is just licensing FSD to Ford, Tesla will still get most of the profit from every car Ford sells. Even if the license cost is only, say, $8K, that's a lot more than Ford makes per car and it's almost 100% profit for Tesla.

Will customers pay an extra $8K for a car that drives itself? I think they will. The cost of an EV is going to come down by more than that in the next few years. And nobody will want a car that you have to drive manually.

Like I said earlier today, this is why selling Windows was a much more profitable business than selling the PCs that Windows runs on. The value is in the software, not the hardware.

But wouldn't that price Tesla out of it entirely? If Tesla is charging so much per vehicle for an OEM to license FSD then why would the OEM bother in the first place? Tesla's pricing has to be fair and equitable or no one will take them up on licensing at all. The OEM's need to make profit to survive as well.

Remember Tesla's mission is not to make the most profits possible, it's to transition the world to EV's and sustainable living. Look at how Tesla is folding the OEM's into Supercharging, it isn't super lucrative for Tesla at all, they are going easy on the OEM's without punishing them financially. It would stand to reason FSD licensing will follow the same pattern simply because Tesla likely won't be greedy with it.

Or do most people really think Tesla will fleece the OEM's for huge profits?

 

[Krugerrand]

Dude - none of us have including you. But we get to comment and criticize. Thats what this board is for. If you can't tolerate mild criticism of your hero, get your own board.

The way you talk - no one should criticize a president (except for ex-presidents) and senators (except ex-senators) ?!

Not so respectfully this time since you can’t seem to extend any in return. This THREAD is for discussing long term investing in Tesla. You’ll want one of those toxic cesspool threads to criticize Elon and jabber about unrelated politics.

To the point, though, you are straight up wrong. Some of us do know a wee bit having actual experience working in the car industry and related fields (you’d know that if you weren’t so busy criticizing my hero and passing your opinion off as fact).

We even have the odd ex-employee show up here on occasion and provide some insight. Looking at Disco for example.

Of course you are free to comment and criticize just as I’m free to correct you. Dude.

 

[EVNow]

I see "between two growth waves' as "sharpening the axe", the components are:-

1. Developing Robotaxi ready FSD, and getting sufficient regulatory approval,

Hmmmm .... Developing a robotaxi would be more like cold fusion, not "sharpening the axe".

 

[EVNow]

Not so respectfully this time since you can’t seem to extend any in return. This THREAD is for discussing long term investing in Tesla. You’ll want one of those toxic cesspool threads to criticize Elon and jabber about unrelated politics.

No - you want people who are absolutely and uncritically supportive of everything and anything Tesla/Elon does.

Where did I say anything about politics in that post ?

Exactly.

You are the one blabbering something about "you can't criticize because you have not run a large company like Tesla". Total BS argument.

 

[Musskiah]

Essentially once Y was done - they should have started building the next gen model plant in Mexico.

By choosing not to build out GigaMex they are prudently reducing CapEX to remain lean and keeping their best engineers around the pilot line of the new model.

CT is not the millions a year model

Ah-hem...it's not?

 

[2daMoon]

But wouldn't that price Tesla out of it entirely? If Tesla is charging so much per vehicle for an OEM to license FSD then why would the OEM bother in the first place? Tesla's pricing has to be fair and equitable or no one will take them up on licensing at all. The OEM's need to make profit to survive as well.

Remember Tesla's mission is not to make the most profits possible, it's to transition the world to EV's and sustainable living. Look at how Tesla is folding the OEM's into Supercharging, it isn't super lucrative for Tesla at all, they are going easy on the OEM's without punishing them financially. It would stand to reason FSD licensing will follow the same pattern simply because Tesla likely won't be greedy with it.

Or do most people really think Tesla will fleece the OEM's for huge profits?

I agree.

Tesla could practically give away the licensing to the OEM, and make bank on the ongoing support for the product (subscription only).

Tesla could even share some of the take with the OEM and still be rolling in the dough. Allowing the OEM to make more BEVs.

This would meet both the "transition" goal and the business goals while keeping costs to the customer at what the market will bear. (AKA a fair price)

 

[Usain]

But wouldn't that price Tesla out of it entirely? If Tesla is charging so much per vehicle for an OEM to license FSD then why would the OEM bother in the first place? Tesla's pricing has to be fair and equitable or no one will take them up on licensing at all.

Remember Tesla's mission is not to make the most profits possible, it's to transition the world to EV's and sustainable living. Look at how Tesla is folding the OEM's into Supercharging, it isn't super lucrative for Tesla at all, they are going easy on the OEM's without punishing them financially. It would stand to reason FSD licensing will follow the same pattern simply because Tesla likely won't be greedy with it.

Or do most people really think Tesla will fleece the OEM's for huge profits?

Tesla charges $12K to its own customers right now. That's for just the software as the hardware is already included.

So if Tesla only charged an OEM $8K for the software and the FSD computer, plus engineering support, that sounds like a very fair price to me.

Remember that in just a few years, the cost of EVs will come down by more than $8K. So the cost of the car plus FSD will still be very reasonable.

As for the mission, I hope that Tesla stipulates that FSD is only to be used in EVs. So now all EVs are even more compelling. If you want an ICE car, you have to drive it manually.

 

[Musskiah]

Not so respectfully this time since you can’t seem to extend any in return. This THREAD is for discussing long term investing in Tesla. You’ll want one of those toxic cesspool threads to criticize Elon and jabber about unrelated politics.

To the point, though, you are straight up wrong. Some of us do know a wee bit having actual experience working in the car industry and related fields (you’d know that if you weren’t so busy criticizing my hero and passing your opinion off as fact).

View attachment 1035312

We even have the odd ex-employee show up here on occasion and provide some insight. Looking at Disco for example.

Of course you are free to comment and criticize just as I’m free to correct you. Dude.

Musskiah (our Hero) reaction x1

 

[Mengy]

I agree.

Tesla could practically give away the licensing to the OEM, and make bank on the ongoing support for the product (subscription only).

Tesla could even share some of the take with the OEM and still be rolling in the dough. Allowing the OEM to make more BEVs.

This would meet both the "transition" goal and the business goals while keeping costs to the customer at what the market will bear.

Exactly. I model OEM licensing in my spreadsheet as a negligible up front fee, but rather a small percentage of the ride revenue from non-Tesla Robotaxis. And I do mean small because eventually even the Robotaxi market will become competitive, which means margins will come down to reasonable levels.

I just don't see Tesla getting greedy with FSD if they end up being the one to supply FSD to the majority of the RT market. On the contrary, given Tesla's history of being very generous with open patents and very affordable tech deals, I'd wager Tesla will make FSD licensing to OEM's very inexpensive and easy for them to justify.

In other words, I don't think it's going to be the massive huge immense profit stream many others do. Sure it will be very profitable, but not ridiculously so. IMHO.



It's going to be incredibly interesting to see how this story plays out.

 

[Musskiah]

I’m sure the quality of Tesla Customer Service varies by location, but I have to give a shout out to the Tesla team in Indianapolis. I’ve had four Teslas over the last eight years and the service in Indy has been exceptional, far and away better than any other car company I’ve dealt with. This, for me, combined with the utility of the Tesla app and convenience of mobile service stands head and shoulders above the service I’ve received from legacy Automakers.

Same. Incredible in Ohio. I have experience with all locations in the state.

 

[Usain]

I agree.

Tesla could practically give away the licensing to the OEM, and make bank on the ongoing support for the product (subscription only).

Tesla could even share some of the take with the OEM and still be rolling in the dough. Allowing the OEM to make more BEVs.

This would meet both the "transition" goal and the business goals while keeping costs to the customer at what the market will bear. (AKA a fair price)

Yes, staying with a subscription-only model and sharing the profits with the OEM would work as well. I suspect that the profits would be similar to the other model.

I don't think the profits would be shared evenly with the OEM. Tesla took the risk of spending billions to develop FSD. Tesla deserves a larger share of the FSD profit.

 

[ggr]

The way you talk - no one should criticize a president (except for ex-presidents) and senators (except ex-senators) ?!

Where did I say anything about politics in that post ?

Hmmm, thank dog that I'm taking a break from moderating.

 

[MC3OZ]

We also know that Tesla is consistently sending out employees in cars to get data. They did this on Chucks UPL for months and it's gotten better, but still not perfect. It's not as easy as they have millions of cars and they just let video flow.

IMO the process might be something like:-

1) Capture edge cases from user interventions.
2) Send requests to the fleet for data relevant to the highest priority edge cases.
3) Send employees out to capture data on the highest priority edge cases...
4) Retrain and hopefully advance in the march of 9's.
5) Repeat 1)..4) as many times as is needed.

As long as V12 mostly improves each time through the cycle, that is a likely path to a viable solution..

Robotaxis will effectively be geofenced starting in a particular city. It makes sense for initial locations to be cities were FSD performs well. It also makes sense to do more cycles, and not rush it out before it is ready.

California is a likely starting location, but I think they need to register some sort of official trial as the first step?

How good were Cruise and Waymo when they started?

Tesla can start out with a "safety driver" in the car.

 

[EVCollies]

... literally none of that makes any sense my man. You can't "reroute processing" to the infotainment-- that's some Star Trek technobabble.

First- they're entirely different chip architectures running entirely different code.

Second, to my knowledge the infotainment computer by design for safety reasons can not give commands to the driving components of the vehicle like the brakes, accelerator, or steering wheel. (Note: this, among the 3 reasons I list, IS a software one and COULD be changed- but wouldn't help you even if you did because of reasons 1 and 3)

Third-even if it could both run some simple "pull over safely" code AND command the car controls to do it- guess where all the camera data that ANY computer would need to be able to pull over safely goes via direct plug in? The driving computer. Which loses networking if node A dies. So it can't provide any data to the MCU to "drive" with even if the other reasons didn't exist (which- they do).

What you're suggesting would require additional HW that does not exist (at least up through HW3 cars--- HW4 appears to add a second actual network interface into the other node).




You haven't actually described any that are physically possible though so I'm not sure there's "many"... or any.




Nobody's moving anything. The 2020 issue of using both nodes for a single instance of the software is still true today.

But in addition to that there's HW limitations that make true redundancy impossible even if that software problem was magically gone.

I agree they have meetings about this. Which is probably why HW4 has dual network connections to solve the exact hardware limitation I'm describing.

If there was some easy software fix Tesla would've just done THAT instead of adding hardware- instead they added hardware. Because it's a hardware limitation you can only fix in hardware.

But that does not help the majority of the fleet on HW3 and older- esp. when Elon already told us those cars won't be upgraded to HW4.







FWIW, every actual service tech I've ever dealt with from Tesla has been excellent.

The people who do all their service communication are terrible.

A couple weeks ago I finally was able to schedule the trunk harness recall a few years after it was issued... I set up an appointment that the app then told me was BOTH:
A mobile service appointment
and
Scheduled to take place at the Tesla service center in Raleigh


Now, it was 10 days away, I figured this should be quick/easy to resolve... the only communication method is the app of course--- I messaged them pointing out the concern- and can they clarify if it's really a service center appointment-- or if it's a mobile one, and if it's a mobile one, can they please correct the address to my actual home. Which I also gave them even though I know they have it on file.

3 days go by and nothing. I message again- they reply a day later saying yes, it is a mobile appointment.

And say nothing about the address issue- and it still lists the SC in Raleigh as the location.

I message again asking them to fix this to correct the address for the service location.

3 days go by and nothing. I message again. They tell me THEY can't actually change it- but they've sent a request to the mobile team who can.

3 more days and nothing has changed... my appointment is now tomorrow and I still don't know where it's happening... I message once again, and finally they get back to me saying ok , we updated it. (so I guess magically they can now?).

Next day- Tech shows up bright and early-- he's polite, professional, friendly, does a quick and great job-- had him change my cabin filter while he was there as I've got an older 3 where it's a PITA.

Great service- terrible communication as always.

As a 2 Tesla family for almost a decade living far away from a Tesla shop (till recently) relying on Ranger service I will share two great anecdotes:

"Hi guys, was in the neighborhood and noticed on of your cars was due for a recall seatbelt check, took a chance and found the car in the garage, all is well."

"Hi guys, was in the neighborhood and hadn't seen the collies in a while so dropped in for a quick visit with them, hope you don't mind."

Agree, Tesla service is the best!

 

[2daMoon]

Exactly. I model OEM licensing in my spreadsheet as a negligible up front fee, but rather a small percentage of the ride revenue from non-Tesla Robotaxis. And I do mean small because eventually even the Robotaxi market will become competitive, which means margins will come down to reasonable levels.

I just don't see Tesla getting greedy with FSD if they end up being the one to supply FSD to the majority of the RT market. On the contrary, given Tesla's history of being very generous with open patents and very affordable tech deals, I'd wager Tesla will make FSD licensing to OEM's very inexpensive and easy for them to justify.

In other words, I don't think it's going to be the massive huge immense profit stream many others do. Sure it will be very profitable, but not ridiculously so. IMHO.



It's going to be incredibly interesting to see how this story plays out.

Economies of scale dictate lowering the costs once it is in full swing. Today's prices are no reference point at all once there are millions and millions of robotaxis and autonomous vehicles out and about.

Then, a little further down the road, after TE has solved for the Earth having been one of those "Batteries not included" planets, the cost of energy becomes very close to free, as Tony Seba and others have calculated.

At that point FSD, robotaxis, and most everything else (UBI?) might be provided by collection of minuscule portion of a tiny sales tax or corporate tax, as most anything needed for basic living (food, shelter, transport) could easily be almost completely subsidized.

Once the cost of energy at all levels is removed from the equation for everything needed to survive reasonably well, only the nicer things will have a price.

It might then become a TISTAAFL world (There Is Such Thing As A Free Lunch) where those who are creative can supplement their standard of living by bringing something to the party. (art, invention, skills, etc.)

I don't think most people really grok how this could be possible, yet. Though I think that Elon does.

 

[2daMoon]

I don't think the profits would be shared evenly with the OEM. Tesla took the risk of spending billions to develop FSD. Tesla deserves a larger share of the FSD profit.

I figured it would be more like Tesla will be the one continuing to support the operation for all autonomous BEVs, creating updates, etc. and will need to cover those costs, and slip a little something to the shareholders too.

 

[Robertj]

Regarding FSD Ashoks latest statement about solving the trolley problem , refers to conflicting risks when making decisions while driving

In the context of Full Self-Driving (FSD) technology, the "trolley problem" could refer to a difficult decision that an autonomous vehicle's AI system might face. For example, in a situation where an accident is unavoidable, the AI system might have to choose between two potential outcomes: one where it swerves to avoid hitting a group of pedestrians but potentially causes harm to the vehicle's occupants, or one where it maintains its course to protect its occupants but potentially harms the pedestrians. This dilemma is similar to the trolley problem, as it involves making a difficult ethical decision with potentially life-or-death consequences.

Interesting dilemma to solve , one that a computer can make quicker than a human most likely

 

[Jim Holder]

I don’t disagree with you. I’m just saying it’s not actually up to us what the SP does or would do given this information.

I mean, if you toddle over to the options thread you’ll see a bunch of confuzzled individuals. People who have spent months, even years, studying Tesla, TSLA, TA, options, et al and simply can not understand how TSLA didn’t crater to fill ‘that’ gap yesterday. I haven’t caught up today to see if they’ve figured out why TSLA is so far green today or if they still are confuzzled.

SP not in our control. Not in Tesla’s control.

Hey, I hang out in that “other” thread mostly and am indeed curious why TSLA is green and didn’t fill ‘that’ gap. What can you share about that?

 

[SO16]

Hey, I hang out in that “other” thread mostly and am indeed curious why TSLA is green and didn’t fill ‘that’ gap. What can you share about that?

I have yet to see a good reason why “filling the gap” is even a thing.

 

[Robertj]

FSD Deals

FSD licensing

1. Allow OEM to integrate it into any car for free or at cost. Car owner pays for FSD license

2. Active safety & basics comes free

3. Leverage Supercharger deals. Offer same pricing as Tesla cars for EVs that integrate FSD hardware & share data

Musk on X says happy with those terms