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

[Usain]

https://twitter.com/x/status/1788236700709175700

Elon says:

We’re getting into rare, complex situations, for example: going down a narrow, one-way road, encountering a road closure and having to reverse out to find a new route. That closure also needs to be communicated to the rest of the fleet, so you don’t get a whole bunch of Teslas stuck down a road

I believe this is the first time Elon has mentioned that they are working on robotaxi-specific behavior for FSD.

This is just what I was hoping to see, assuming they are indeed working on fleet-wide communication.

 

[durkie]

Autonomy is happening now. Future dominance in FSD and androids, aka humanoid robots, could be established within the next six months to a year.

The "dominance" part I legitimately don't understand. Self-driving is hard right now because no one has done it. Waymo sort of has done it, but they control all the vehicles and the vehicles only work in specific areas.

But if Tesla comes up with a general solution and then releases it out to the world (through software updates to the cars), I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

I think OpenAI even at one point had some clause about not allowing ChatGPT to be used to train other models. And they control the whole stack -- they can actually rate limit / ban people for bad behavior. But the FSD computer is just sitting in all of our cars.

 

[2daMoon]

I don't understand the uber bulls not accepting any criticism of the company or their investment or feel everyone is a troll (which I disagree with obviously). Past performance does not guarantee future returns with investments. It's still a wait and see I feel and I'm fine with seeing what happens in 3/5/10 years.

I think the facts are:
1) Sales YoY, QoQ has dropped which has never happened (* outside of Covid) for Tesla. No growth in sales tend to be a death spell for growth companies. I'm more interested in Q2 numbers rather than 8/8.

Macro economics, not specific to Tesla

2) Elon is in demon mode and to me, he himself is very concerned so why aren't investors or long term people here even slightly concerned? I a lot of people bring back the days of Tesla in 2018, but the stock/company was only worth about $20 and had a market cap of $10 billion (I think). We are at a $550 billion market cap now and Elon himself is very concerned with costs/doing layoffs, telling everyone to go hardcore. You don't do that unless you are worried.

There is more than one possible explanation for this pivot on Elon's part.
You seem to think there is only one way to read this in order to shape the narrative you prefer.

3) There has been very large inventories of cars built up now just sitting for anyone to buy. This was not common until recently. Remember how Tesla's were supposed to be an appreciating asset/car?

You really have no idea what you are talking about here.

Please provide the inventory numbers compared to other auto companies you feel are doing better.

Take into consideration their inventory on dealer lots may not get counted as those have been "delivered" by their accounting, and, that Tesla inventory includes vehicles already purchased and in transit for delivery, as well as demo and showroom vehicles.

4) Unlike the uber bulls who will buy nothing but a Tesla, every other EV is an EV Tesla didn't sell and there is a LOT more options now vs. 2018 which only had the Bolt.

Every other EV is an EV helping Tesla complete the Mission.

5) For folks who thought the SC network was Tesla's main advantage, that advantage is now gone as well.

Please, explain how you came to this conclusion.
Include discourse on how Tesla currently has the largest and most reliable network.

Also, touch on how EV adoption has reached the point where others will finally be compelled to build charging infrastructure.

Tesla can focus on SC manufacturing, servicing, and operating the SC network and step away from the burden of site selection and permitting. This very possibly could accelerate Tesla SC deployment.

6) They have cut prices of the MY (I tend to not care about the M3) by $20k in the US.

Go read Elon's Top Secret document from a decade and a half ago that laid out Tesla's long term plans to consistently reduce prices.

7) Their operating margin was almost 20% in some quarters, but has since dropped to 5.5%.

Meh. Tesla could have any margin they choose. They price to most effectively accelerate the transition to renewable energy, rather than to accelerate the growth of their pile of money. This is what makes Tesla different from companies whose only goal is to pay dividends.

8) Elon has touted they should be viewed as an AI/robotics company and if you don't care for it or without it, you should not invest. That's all fine and good, but folks who believed on the earlier mission of EVs for all has a voice as well as to whether they should invest and if they don't like the FSD/Optimus, are probably getting out.

The earlier mission was, to accelerate the transition to renewable energy, same as the later mission.
The EVs are AI Robots, after all. Did you not get the memo?

9) A few senior execs have left. This is all normal, but the manner of some layoffs has workers concerned and productivity will probably take a short term hit.

When in doubt, zoom out.
In this case, zoom out to look at any of the previous examples where Elon did this, and what then happened.

BTW, these replies are not for you. You have been provided this information before. This is for anyone that might be taken in by your consistent attempts to introduce fear, uncertainty, and doubt via the stale narrative you routinely and repeatedly present.


Sing along with me,

Don't sunworry, be sunhappy

 

[Linkmeup]

https://www.notebookcheck.net/2024-Tesla-Model-3-wins-fastest-charging-in-the-gas-station-stop-test.834919.0.html

"Thanks to the integrated experience of Tesla cars with their Superchargers, the Model 3 Highland managed to add more than 60 miles in the gas station test. That's nearly double the rate of most of its competitors that AMCI tested, even the 800V Hyundais."

Obviously, everyone can charge at home overnight and that's the end goal compared to having to do an apples-to-apples comparison between new tech (supercharger + at-home charging) and gas stations. Though, thought this particular tidbit was interesting about Highland.

 

[Mengy]

That's why I don't sell. I just think Elon 100% mismanaged the last few years and didn't care about shareholders

I wouldn't presume to say Elon has mismanaged Tesla, but I would say he's made some decisions which have not been favorable to TSLA investors over the past few years. I feel much of the stock's pain is his doing, even though I do feel Tesla the company is doing okay. Except for laying off the SC team, I still can't get on board with that one.

Future growth is why I don't sell either, though I don't think that growth will be what many TSLA bulls expect it will be. I think Elon will continue to cut margins on everything, including FSD and eventually Robotaxis to increase adoption rates and grow the company in size, BUT at the expense of profits. I also think he'll price Optimus extremely cheap and appealing to increase adoption to massive levels while making meager profits from said massive growth.

I'm expecting modest gains on my TSLA shares over the next decade from a huge company with razor thin margins and relatively low profits, but I'm okay with that.

 

[Usain]

Elon says:


I believe this is the first time Elon has mentioned that they are working on robotaxi-specific behavior for FSD.

This is just what I was hoping to see, assuming they are indeed working on fleet-wide communication.

Thinking on that some more, I guess that feature would be useful for non-robotaxis as well. He may mean all Teslas when talking about the "fleet" here.

 

[Usain]

The "dominance" part I legitimately don't understand. Self-driving is hard right now because no one has done it. Waymo sort of has done it, but they control all the vehicles and the vehicles only work in specific areas.

But if Tesla comes up with a general solution and then releases it out to the world (through software updates to the cars), I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

I think OpenAI even at one point had some clause about not allowing ChatGPT to be used to train other models. And they control the whole stack -- they can actually rate limit / ban people for bad behavior. But the FSD computer is just sitting in all of our cars.

Whether you are training by mimicking humans or mimicking a Model 3, you still need the training data. Nobody else has that.

But you are correct in that if you did have the training data you could use a Model 3 to build a clean-room version of the system at a much lower expense than Tesla. It would always be about a year behind, but it could be done.

 

[SOULPEDL]

Autonomy is happening now. Future dominance in FSD and androids, aka humanoid robots, could be established within the next six months to a year.

That is the source of the urgency.

Elon hates to lose.

Been wanting to say the same thing all morning, thanks!

And in Battle Mode because we should all be fully aware of the likely increase in TSLA wild swings, BS stories from everywhere, sky is falling on social media, even this thread at some risk for some strange reason that's beyond me.

Obvious why I'm bullish where others think Elon's reckless. This is how warriors act when they just figured out how to work some new alien weapons and war is imminent because the attacks on Tesla are relentless with history in the making on a planetary scale. If he's anything like me, he's also enjoying this and dreamt it already happened.

I have a funny feeling Tesla already ripped off the bandaide, we just don't know it fully yet and it might even bleed some for a bit. It's gonna potentially shake some more "HODLers", but I have plans to go anywhere. I have a front row seat to this show.

And who knows, there could have been valuable resources in Buffalo needed for FSD along the lines of special skills or experience. Again, we don't know. But there is solid evidence that drives me around everyday now. It's got the human thing going on... and we're gonna talk to it soon? Just wow.

We're gonna laugh looking back at this whole Beta experiment by end of this year or next - that we even survived! Maybe someday I'll even share some stories being out there in the wild from day one. I actually have a recording of my first drive on FSD. Clueless how to engage it and thought I was at first, until...

 

[MP3Mike]

I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

That would still be a huge amount of work for them. They would have to create every edge case possible and have their Model 3 drive through it to capture how it responds. It would probably just be easier to hire a team of professional drivers to create, and drive through, every possible case/edge situation.

So no, I don't think anybody is going to try to use a Tesla's driving behavior to create their own system.

 

[Zero CO2]

The "dominance" part I legitimately don't understand. Self-driving is hard right now because no one has done it. Waymo sort of has done it, but they control all the vehicles and the vehicles only work in specific areas.

But if Tesla comes up with a general solution and then releases it out to the world (through software updates to the cars), I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

I think OpenAI even at one point had some clause about not allowing ChatGPT to be used to train other models. And they control the whole stack -- they can actually rate limit / ban people for bad behavior. But the FSD computer is just sitting in all of our cars.

what other company has the inference compute and required cameras installed in their vehicles ??? i have not tracked other EV makers in this regard ...
so even if they could replicate it on one car ... they are literally 6 years worth of volume behind ...Tesla has produced over 6 million vehicles and the majority are FSD capable ... as Elon mentioned numerous times competition will need a telescope ... it is not about a prototype is is about scaled volume and the per unit economics

 

[2daMoon]

Except for laying off the SC team, I still can't get on board with that one.

This is the problem with generalizing things. Doing so can lead to misinterpretation of the known facts. Try, instead, to be more specific.

i.e. :

Tesla has laid off a portion of the SC team. The part of the team that was laid off en masse appears to be associated with new site planning and installation, as per Elon's response indicating how SC sites under construction should continue, and those that haven't broken ground should pause, until they are contacted.​

​

There is no indication the portions of the team providing Supercharger manufacturing, site repair and updates, and network payment and availability services were as heavily impacted.​

To interpret this in any other way is a disservice to providing an accurate evaluation of Tesla to the investment thesis.

 

[SOULPEDL]

The "dominance" part I legitimately don't understand. Self-driving is hard right now because no one has done it. Waymo sort of has done it, but they control all the vehicles and the vehicles only work in specific areas.

But if Tesla comes up with a general solution and then releases it out to the world (through software updates to the cars), I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

I think OpenAI even at one point had some clause about not allowing ChatGPT to be used to train other models. And they control the whole stack -- they can actually rate limit / ban people for bad behavior. But the FSD computer is just sitting in all of our cars.

China would have done this long ago, and Taiwan (TSMC) holds all the worlds chips. Funny thing, this conversation is actually about 5 yrs old, on this thread.

The copycat method chases a moving target for yesterday's tech that's not as safe. Needs the same hardware? What a selling point! /s

Sure, there's a market for it, but it would likely take an army of engineers to do the equivalent of brain surgery for even a snapshot in time, several iterations too late, and no hope of ever catching up. There's no business model there. Funding declined... again.

 

[elasalle]

Elon says:


I believe this is the first time Elon has mentioned that they are working on robotaxi-specific behavior for FSD.

This is just what I was hoping to see, assuming they are indeed working on fleet-wide communication.

Isn’t closure fleet update just Waze like update? Nothing to do with FSD?

 

[SOULPEDL]

Oh, and it was along these lines, (this one's more funny though).

 

[alexxs88]

It's really puzzling to me how being well below the bear case (based on the Ark projections of 2019) in terms of car volume and margins is seen as an achievement and proof of good execution by some.

 

[Usain]

Isn’t closure fleet update just Waze like update? Nothing to do with FSD?

The update part is like Waze. But recognizing the situation and notifying the mother ship would be part of FSD.

With Waze, the recognition and reporting is done by humans. What Elon said is that FSD would do recognition and reporting automatically.

I think even Waymo is doing manual (human-driven) recognition and reporting for road updates.

So if Tesla is indeed working on this, then they are already working on stuff that robotaxis need and nobody else has.

 

[capster]

The "dominance" part I legitimately don't understand. Self-driving is hard right now because no one has done it. Waymo sort of has done it, but they control all the vehicles and the vehicles only work in specific areas.

But if Tesla comes up with a general solution and then releases it out to the world (through software updates to the cars), I don't understand what's stopping a competitor from just training their self-driving program off of "behave the way this Model 3 would in this situation". It feels like tesla's unique access to data + capital won't be that much of an advantage once everyone has the distilled result of it in their car.

I think OpenAI even at one point had some clause about not allowing ChatGPT to be used to train other models. And they control the whole stack -- they can actually rate limit / ban people for bad behavior. But the FSD computer is just sitting in all of our cars.

Technical, cost, and legal barriers prevent reverse engineering Tesla’s FSD and deploying it.

A thief would need billions of miles of driving encompassing all the meaningful situations Tesla has found and placed in its training data to extract the behaviors embodied in Tesla’s FSD. This requires a fleet and expensive compute.

They’d still probably wind up with a shadow of FSD that would be difficult to validate and have unknown gaps.

They’d be legally liable for those gaps, not to mention Tesla could sue them.

Who would trust your system or the thief given its lower safety and the fact they’re a thief?

edit: Also hurdles from regulators: “How does your system work and how does it stay current?" Thief: "Umm, we just steal from Tesla…"

 

[Usain]

It's really puzzling to me how being well below the bear case (based on the Ark projections of 2019) in terms of car volume and margins is seen as an achievement and proof of good execution by some.

Based on Ark's projections? I'm surprised you didn't trot out the projections from Warren Redlich.

 

[winfield100]

When was the last time an insider / executive bought Tesla stock on the open market? They're always just selling.

may i suggest you use SEC Edgar

SEC.gov | EDGAR—Search and Access

www.sec.gov

 

[Usain]

Technical, cost, and legal barriers prevent reverse engineering Tesla’s FSD and deploying it.

A thief would need billions of miles of driving encompassing all the meaningful situations Tesla has found and placed in its training data to extract the behaviors embodied in Tesla’s FSD. This requires a fleet and expensive compute.

They’d still probably wind up with a shadow of FSD that would be difficult to validate and have unknown gaps.

They’d be legally liable for those gaps, not to mention Tesla could sue them.

Who would trust your system or the thief given its lower safety and the fact they’re a thief?

edit: Also hurdles from regulators: “How does your system work and how does it stay current?" Thief: "Umm, we just steal from Tesla…"

It wouldn't be illegal. See Clean room design - Wikipedia

But everything else you said it true, which is why it would probably never happen. And I agree that even though it's not illegal, the term "thief" would still apply.