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

[Zero CO2]

I don't think CT helps mission right now, these are trophy cars for almost all users. They have other cars and usually an EV. We need 10 million more EVs a year. Each CT is going to eat cell production for I guess 4 25k cars. The 25k cars should take ICE off the road.

anecdotally not my experience reading the Cybertruck owners club forum ... many are new to Tesla and are previous truck owners per their bios

and i have seen @anthonyj venting about his CT over there also ... patience grasshopper

 

[FreqFlyer]

James has modest expectations for Q1 earnings:-

There are a number of macro and one-off factors impacting on Q1 results.

But from my point of view anything Tesla was doing wrong in Q1 2024, they were also doing wrong in Q4 2023.

Some issues have been exposed, and areas where Tesla needs to do better are now exposed with more clarity.

All the "one-off" factors impacting on Q1 should not be a factor in future quarters.

The real world is more complex than many of the solutions posted here imply.

Advertising is one example, it hasn't been a silver bullet, in part because everyone does it, and a vastly larger sum of money is focused on spreading the opposite message on EVs.

So it isn't just advertising that is required, but the right kind of advertising, and a small advertising budget has to punch punch above it weight. IMO the most likely path to doing that is to leverage the fact that advertising is a "creative industry", many people who are good at painting and writing are also understandably good at advertising. And a good idea can cut-through and punch above it weight. So the right team of people need to be given the right brief. My suggestion is that humour is entertaining and memorable, but all really know is, Tesla is still searching for the right formula.

The other consistent theme here is that people seem to think that Tesla should invest a whole bunch of capex and build some new factories to build some new models. What would that do to the financials and sales of existing models in a challenging macro environment? The short term way forward is to balance supply and demand for their existing production capabilities.

Yet another theme is the race for EV market share isn't a race and that many people will foolish enough to buy an ICE when EVs reach price parity. Every sustainable price reduction unlocks some additional market share, there is sometimes a lag, and addressing that lag brings us back to advertising. But keep working on cost reduction and product improvements in parallel with attempting to sort out advertising.

But that is the thing, no new factories are needed. The factories now are not at full utilization. There is slack in the system that could be used.

 

[Mengy]

Lol. Quite the bear den in here

It's understandable, there has been a lot of frenzy and disappointing news for Tesla lately.

Q1 P&D numbers are way lower than expected. The consumer $25K car might be getting pushed back, brining the auto growth story to a grinding halt. More price cuts to possibly decimate margins even further. EPS estimates have been cut substantially. 20,000 Tesla employees laid off. Growth and marketing team given the boot. Elon had his compensation package taken away from him by a judge and a guy with nine shares. TSLA is at $140 and most likely about to fall down even further after the ER tomorrow.

I'm honestly amazed its as positive as it is in here!

 

[SPadival]

FSD is not just a "good L2 system". There are no other systems like it. Suggesting that it's a L2 system of comparison tells me there are better or worst FSD like products out there which doesn't exist.

Yes. How hard is it for Tesla to roll out an L3 system under these ridiculous conditions!! (copy paste from mbusa website)

• Clear lane markings on approved freeways
• Moderate to heavy traffic with speeds under 40 MPH
• Daytime lighting and clear weather
• Driver visible by camera located above driver's display
• There is no construction zone present.
  

Meanwhile, ZackRigsEverything previews some spon con from mbusa & gets community noted

https://twitter.com/x/status/1781801228068102163

 

[Usain]

FSD is not just a "good L2 system". There are no other systems like it. Suggesting that it's a L2 system of comparison tells me there are better or worst FSD like products out there which doesn't exist.

Correct. This is why the levels don't matter in the real world.

For example, Tesla FSD is far, far superior to the Mercedes system that is classified as L3.

 

[Artful Dodger]

Why 2170s and not 4680s?

You kidding, right? LG doesn't have any 4680s to sell in those sorts of quantities. It will be years before they're ready to sell large volumes of the new size cells. Meanwhile, Tesla semi is being built with the old size cells.

 

[Singuy]

Yes. How hard is it for Tesla to roll out an L3 system under these ridiculous conditions!! (copy paste from mbusa website)

• Clear lane markings on approved freeways
• Moderate to heavy traffic with speeds under 40 MPH
• Daytime lighting and clear weather
• Driver visible by camera located above driver's display
• There is no construction zone present.

Meanwhile, ZackRigsEverything previews some spon con from mbusa & gets community noted

https://twitter.com/x/status/1781801228068102163

Yeah that guy ragged on Tesla for praising a guy who reached the hospital on FSD as dangerous and irresponsible. He then went ahead and called Mercedes the first L3 autonomous car which misleads people thinking it can fully drive itself and then ends up getting T-boned from failing to stop at a red light.

 

[Knightshade]

FSD is not just a "good L2 system".

That is literally what it is.

There are no other systems like it.

Define "like it"

There's systems in China that are quite a lot like it (some might suggest TOO like it if China had stronger IP laws). Are those "like it"? If not, why?

There's also other systems operating above L2, both in the US and in China (including active, passenger-carrying, robotaixs). Those are harder to judge if they're "like it" or not since usually they're limited in some ways FSD is not (for example they're usually geofenced in various ways), just as FSD is limited in some ways those systems are not (mainly the ability to operate >L2).


But even if there weren't ANY like it it wouldn't magically make it not be L2 all of a sudden, nor make it not be a good system-- so you appear to be trying to create an argument where no basis for one exists with the actual words I wrote

 

[dhanson865]

This is a superb example of absence of critical thinking. Reducing price on FSD is really a waste. It could and should be a Premium feature. Reducing the price of a fundamentally revolutionary product actually diminishes the perceived value. The idiotic idea that a revolutionary product is linearly price sensitive is simply a foolish waste of revenue. I am appalled! Allowing transferability was a great idea, incentivizing without diminishing value. Flat out price reductions? In a word: WRONG!

Tesla could have taken advice from Bernard Arnault but instead is turning to Juan Ricardo Luciano or, even worse Crazy Eddie.
This one is the most value-destructive decision yet. Made because idiots equated FSD, the Premium Producr of decades, with a commodity.

With this latest move I am shocked! There is no coming back from this one…

There was a time when basic Auto Pilot cost extra. For example in 2019 it was $3,000 but now it is free, that isn't because FSD price is going to zero, that's because FSD is building on top of the existing layers, to new capabilities.

EAP was until very recently a layer between AP and FSD with prices of $6,000 and $12,000

At some point down the road features that are in EAP might be the new base with what we now call FSD being the middle layer and what is build on top of that being the new top layer.

As long as that newest top most layer is more expensive than the old top layer they can absolutely drop the price of an existing layer without cheapening the perception of the software stack

Figure out the naming but you had

Autopilot -> Enhanced Autopilot -> Full Self Driving

what if some day the break down is


Autopilot+EAP = Enhanced Autopilot (built in to the cost of the car)
FSD = Supervised Full Self Driving in the middle slot as a paid upgrade from the first layer
Some new name = The top layer with more capabilities than supervised FSD comes along, even more expensive because it adds L3 or L4 modes?

not to mention network fees on top of that if you want to use it commercially.

I think your worry is because you don't see the new top layer yet.

My expectation is that FSD or some other name that builds on FSD software will be more expensive later this year. I don't see $8000 as the top coming down to $0, I see $8000 as the old FSD becoming the middle choice in the stack making room for a new top layer. Maybe it's $8,000 for this middle layer and $16,000 for a new top layer. Maybe it keeps going down and becomes the new $6,000 middle to some new $12,000 top.

Whatever the pricing mix becomes I don't see the current drop as the end, it's just a transition.

 

[AndrewZ]

Correct. This is why the levels don't matter in the real world.

For example, Tesla FSD is far, far superior to the Mercedes system that is classified as L3.

The Level 3 module is just the Traffic Jam Assist where they actually take liability under the specified conditions.

People act like that’s all it does — that’s not all it does. When not in a traffic jam, it’s a Level 2 ADAS like Autopilot and it has the ability to flip in and out of Level 3 depending on the conditions.

It’s an implementation of NVIDIA’s Drive Orin platform. I have NVIDIA’s system in my 2021 Genesis and can go hours on the highway without needing to steer, just some torque on the wheel and eyes pointed through the windshield.

 

[Singuy]

That is literally what it is.





Define "like it"

There's systems in China that are quite a lot like it (some might suggest TOO like it if China had stronger IP laws). Are those "like it"? If not, why?

There's also other systems operating above L2, both in the US and in China (including active, passenger-carrying, robotaixs). Those are harder to judge if they're "like it" or not since usually they're limited in some ways FSD is not (for example they're usually geofenced in various ways), just as FSD is limited in some ways those systems are not (mainly the ability to operate >L2).


But even if there weren't ANY like it it wouldn't magically make it not be L2 all of a sudden, nor make it not be a good system-- so you appear to be trying to create an argument where no basis for one exists with the actual words I wrote

The words "not just" in front of "good L2 system" suggest I am saying it is more than a good L2 system. Not saying that it's NOT a L2 system. Maybe it's my weak English because I am foreign.

 

[⚡️ELECTROMAN⚡️]

The words "not just" in front of "good L2 system" suggest I am saying it is more than a good L2 system. Not saying that it's NOT a L2 system. Maybe it's my weak English because I am foreign.

He's saying you're wrong. He's the self appointed autonomous driving expert. He thinks mercedes, and certain Chinese companies have superior, or at least near, what tesla's autonomous driving capabilities are. Prove he's wrong.

 

[Knightshade]

The words "not just" in front of "good L2 system" suggest I am saying it is more than a good L2 system.

I suppose there's two ways to read "more than a good L2 system"

One is it's more than L2 in some way-- which it objectively is not.

The other way would be you're arguing a greater adjective than good should be in that description.... which is obviously subjective and I don't think worth debating much, but seems to be what you were going for there.... But that gets back to the "systems like it" debate- because there's other L2 systems- including ones that also do city streets/pedestrians, etc... XNGP for example works in every road across the entirety of china, local and highway.... Now- since that one ONLY works in China, and FSD ONLY works in NA, it's difficult to do a really objective comparison to see in what regards one is better than the other at specific things... but it's certainly like FSD in most respects so your suggestion there's "nothing like it" doesn't seem to hold up unless you're defining "it" VERY narrowly.

He's saying your wrong. He's the self appointed autonomous driving expert. He thinks the mercedes, and certain Chinese companies, have superior, or at least near, what tesla's autonomous driving capabilities are. Prove he's wrong.

I'd appreciate if you didn't outright make stuff up about what I've actually written.

Because I never said anything about Mercedes being "superior" for example. They do offer something that's L3, but it's hilariously narrow in ODD... and their L2 system is much less capable than FSD (comparisons to AP would work out better for Mercedes though).

Thanks in advance.

That said, Chinese companies DO have systems "at least near" what tesla's capabilities are in terms of L2.

They ALSO have actual L4 robotaxis on the roads- though like Waymo they're geofenced more narrowly than FSD is so it's harder to discuss what's "superior" and what isn't in comparing "L4 some places" to "L2 all places"

 

[⚡️ELECTROMAN⚡️]

Group 3: People who want a Tesla to try but doesn't need a Tesla. I feel like this group is pretty much everyone besides die hard fanatics. People who needs a car ASAP due to them missing work or etc would op for the safer option which is an ICE car. People want certainty and ICE provides them with such comfort. Because EVs has many new caveats with them, people who ends up looking for one are the curious ones and require months of preparation (mentally, as in talking themselves into one and becoming excited for one). The curious ones who has this as a WANT most likely wants to wait for the next best thing. The Model 2 could potentially be better than the Model 3(because it's more forward thinking, who knows) and will be 30% cheaper.

But it's just because they don't know better. how do we enlighten them?

 

[⚡️ELECTROMAN⚡️]

I suppose there's two ways to read "more than a good L2 system"

One is it's more than L2 in some way-- which it objectively is not.

The other way would be you're arguing a greater adjective than good should be in that description.... which is obviously subjective and I don't think worth debating much, but seems to be what you were going for there.... But that gets back to the "systems like it" debate- because there's other L2 systems- including ones that also do city streets/pedestrians, etc... XNGP for example works in every road across the entirety of china, local and highway.... Now- since that one ONLY works in China, and FSD ONLY works in NA, it's difficult to do a really objective comparison to see in what regards one is better than the other at specific things... but it's certainly like FSD in most respects so your suggestion there's "nothing like it" doesn't seem to hold up unless you're defining "it" VERY narrowly.






I'd appreciate if you didn't outright make stuff up about what I've actually written.

Because I never said anything about Mercedes being "superior" for example. They do offer something that's L3, but it's hilariously narrow in ODD... and their L2 system is much less capable than FSD (comparisons to AP would work out better for Mercedes though).

Thanks in advance.

That said, Chinese companies DO have systems "at least near" what tesla's capabilities are in terms of L2.

They ALSO have actual L4 robotaxis on the roads- though like Waymo they're geofenced more narrowly than FSD is so it's harder to discuss what's "superior" and what isn't in comparing "L4 some places" to "L2 all places"

Thanks for clarifying. I wasn't too worried about my accuracy since obviously you would correct any inaccuracies. Thank you.

Did you imply earlier that Chinese companies might be stealing Tesla's tech?

 

[⚡️ELECTROMAN⚡️]

They ALSO have actual L4 robotaxis on the roads- though like Waymo they're geofenced more narrowly than FSD is so it's harder to discuss what's "superior" and what isn't in comparing "L4 some places" to "L2 all places"

How well do you think Tesla good do, with their technology, if they concentrated on a small geofenced area?

 

[Singuy]

I suppose there's two ways to read "more than a good L2 system"

One is it's more than L2 in some way-- which it objectively is not.

The other way would be you're arguing a greater adjective than good should be in that description.... which is obviously subjective and I don't think worth debating much, but seems to be what you were going for there.... But that gets back to the "systems like it" debate- because there's other L2 systems- including ones that also do city streets/pedestrians, etc... XNGP for example works in every road across the entirety of china, local and highway.... Now- since that one ONLY works in China, and FSD ONLY works in NA, it's difficult to do a really objective comparison to see in what regards one is better than the other at specific things... but it's certainly like FSD in most respects so your suggestion there's "nothing like it" doesn't seem to hold up unless you're defining "it" VERY narrowly.






I'd appreciate if you didn't outright make stuff up about what I've actually written.

Because I never said anything about Mercedes being "superior" for example. They do offer something that's L3, but it's hilariously narrow in ODD... and their L2 system is much less capable than FSD (comparisons to AP would work out better for Mercedes though).

Thanks in advance.

That said, Chinese companies DO have systems "at least near" what tesla's capabilities are in terms of L2.

They ALSO have actual L4 robotaxis on the roads- though like Waymo they're geofenced more narrowly than FSD is so it's harder to discuss what's "superior" and what isn't in comparing "L4 some places" to "L2 all places"

You sure? They said it's HD mapped. Also just from watching XNGP videos, the car is programmed to just go...it does not wait for pedestrians. I saw a bicyclists had to slam on their brakes for the car. The car will also just go straight for a bus because it has one mode, which is just continue to go. This demonstration did not have a lot of full 90 degree turns and from the 40 mins I've watch, it acts like it's the only car on the road. It is how the Chinese drive however so maybe that's the right thing to do. China FSD is pretty easy, just go straight and cares about no one.

 

[Knightshade]

Thanks for clarifying. I wasn't too worried about my accuracy since obviously you would correct any inaccuracies. Thank you.

Did you imply earlier that Chinese companies might be stealing Tesla's tech?

No need to imply, Tesla has outright stated this and filed lawsuits over it.

Tesla settles lawsuit against engineer who it claims stole Autopilot source code for Chinese competitor

Tesla has settled a lawsuit against a former engineer who the company claims stole its Autopilot source code before going...

electrek.co

See also:

Tesla doubles down on claim Chinese EV startup stole its Autopilot source code

Tesla is doubling down on its claim that Xpeng, a Chinese electric vehicle startup, has stolen and is using its...

electrek.co

How well do you think Tesla good do, with their technology, if they concentrated on a small geofenced area?

If they want to also do very detailed HD maps for the geofenced area? Probably quite good after investing enough time/effort to build those maps and then keep them up to date. But they'd still have some sensor issues like the blind spots around the front of the car very close to it--- and the side-view intersection issues Chuck Cook has done a number of videos on being unable to see around obstructions without, at times, too-far creeping (the very issue that seems to have led to an accident in this very thread earlier today--- which is TOTALLY the fault of the driver who rear ended the creeping Tesla, but obviously a less desirable behavior than if the car could've properly seen cross traffic and not needed to do that stop-start dance.

 

[Knightshade]

You sure? They said it's HD mapped. Also just from watching XNGP videos, the car is programmed to just go...it does not wait for pedestrians. I saw a bicyclists had to slam on their brakes for the car. The car will also just go straight for a bus because it has one mode, which is just continue to go. This demonstration did not have a lot of full 90 degree turns and from the 40 mins I've watch, it acts like it's the only car on the road. It is how the Chinese drive however so maybe that's the right thing to do. China FSD is pretty easy, just go straight and cares about no one.

You posted a 3 month old driving vid then pointed out the flaws.

Exactly the stuff folks here get furious at Dan O'Dowd for doing to FSD.

The works-everywhere-in-china rollout was only 6 weeks ago.

But this again gets back to my earlier, unanswered, question of what you mean by "nothing like it"

That system is obviously "like" FSD in that it's intended to be an L2 system that works everywhere in the country on all types of roads, and is available to consumers today.

Do you accept that means there IS something "like" FSD or do you plan to No True Scottsman this into the ground?

 

[Singuy]

You posted a 3 month old driving vid then pointed out the flaws.

Exactly the stuff folks here get furious at Dan O'Dowd for doing to FSD.

The works-everywhere-in-china rollout was only 6 weeks ago.

Dan stage his videos to specially cause fsd to fail. Also I would imagine none HD map would have worst performance vs what is shown here with HD maps. Honestly if Tesla gets mapping right fsd is pretty much solved. Mapping is pretty much everyone's issue right now cause traffic negotiations are mostly on point.