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

[2daMoon]

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…

From a first principles perspective it seems to me they want more training data, and offering free FSD trials and reduced purchase and subscriptions to FSD is about increasing the number of disengagement data points they can harvest for solving edge cases.

It has not seemed to me this is something being done to increase sales so much as to increase data collection rates. Tesla folk have alluded to how they can comfortably process more data now than ever before. Price drop followed this announcement.

The sale pricing may very well be about getting to full autonomy sooner. Once the goals are achieved they can price FSD at what the market will bear.

 

[Krugerrand]

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…

I understand your POV. I think you’re wrong. On the marketing aspect and on this thought line. Not wrong about marketing , wrong about what needs to be done on the marketing front in the moment.

I’ve seen what’s currently going on happen dozens of times. Elon is streamlining everything, ripping processes apart, pearing everything down, increasing efficiencies, etc…. It’s exactly first principles. He’s learning right now.

He’ll make changes just as quickly if there’s an error he’s just made. Remember, you only have some of the information. He might not have it all in the moment, but he’s got way more than you or I combined.

You got caught up in something you believe you know more about than Elon. And you might be right. I’ve thought the same a time or two. The difference; you ran out of patience and convinced yourself you’re the smarter person in the room on the topic. Sorry to say, but that’s kind of just like Gary et al…. I decided I might not have the whole picture and decided to wait it out - again.

You’re not the first to think there’s ‘no coming back’. Heard that hundreds of times. I believe you just don’t *see* because you’re convinced the company has to move in a certain way. Interesting, though, you’re already looking for things to be a certain way so you can invest again. Have to say that made me ‍.

 

[2daMoon]

An update on Waymo disengagements in California

Each year California requires companies with AV testing permits in the state to report on their disengagements. These reports assist the state in its understanding of how well a self-driving car is doing its job.

waymo.com

https://www.eetimes.com/waymo-cruise-dominate-av-testing/

Looks like Waymo is 1 every 17k miles now.

...and this compares to what?

EDIT:
I did download the 2023 CSV file and the description for the "disengagements" in the column labeled "AV System, Test Driver, Remote Operator, or Passenger" were all noted as "In-field retrieval"

It would be helpful to know how Cali DMV and Waymo define what gets recorded as a disengagement. If it is only those instances where the vehicle must be retrieved in-field (sounds like going there and getting the vehicle to me) then, it may not be counting as "disengagement" those times the operators are intervening during normal use to make corrections unless it results in a retrieval of the vehicle.

Tesla FSD disengagements are mostly intervene and resume events.

 

[uscbucsfan]

...and this compares to what?

FSD comparison...

Home

FSD Community Tracker

www.teslafsdtracker.com

They are close, but Waymo is using HD maps and locked into an area. It's certainly not apples to apples.

 

[Skryll]

Reducing price on FSD is really a waste. It could and should be a Premium feature.

My image of you is that you are highly skilled in an area and have good contributions to make, while maybe missing the point in some other areas:

Now that compute is not the bottleneck, training data from real world driving is.

The more people use it the faster FSD value will rise. Unlike some people experience, 12.3.4 is still far from usable unsupervised.

In the light of how many more cars Tesla will build, the existing fleet are all early adopters today and are laying the groundwork for future mass production and mass adoption.

$99/month for FSD will turn out to have been a brilliant move. The next massive up price opportunity comes when unsupervised is achieved.

Also about car prices - the mission is not maximize profits.

Profits are a means to the mission by being able to be independent of banks and able to take massive action, self sustained by free cash flow.

Thankfully for us investors massive profits and more free cash flow a nice side effect of building company value.

The actual mission translates into getting more people to ditch their gas car for electric and migrate to the new world forever. Lowering the orices as well as informing about value advances the mission.

Just look to First Príncipes to see where the problems are: Placement and Positioning.

Lack of distribution and lack of support: those are the issues.

Production and Product itself is evolving nicely, although with near histrionic convolutions in the process.

Placement is really, really deficient.

I would love to hear more about placement strategies.

When I hear placement I think of movies like the one with those FSD cars in that apocalyptic movie recently and cybertrucks driven by celebrities.

Teach us.

 

[Singuy]

An update on Waymo disengagements in California

Each year California requires companies with AV testing permits in the state to report on their disengagements. These reports assist the state in its understanding of how well a self-driving car is doing its job.

waymo.com

https://www.eetimes.com/waymo-cruise-dominate-av-testing/

Looks like Waymo is 1 every 17k miles now.

I thought it used to be 1 every few hundred k?

 

[Snerruc]

Much though I enjoy, as an investor, Elon's attitude of focusing purely on the long term and ignoring bumps along the way, I do think in this position that I would put a team on the boring company vegas loop to get it to 100% driverless ASAP. I wonder if they are even running FSD12 on those cars now? It seems like it would be relatively trivial to do, and a good way to shut up some of the less-informed and whiny shortsellers.
But maybe its a deliberate decision NOT to do that? It would lead to press suggesting FSD only works in fixed tunnels? Still it seems like the boring company would love to reduce their labor bill by 90%?

The problem is that Vegas does not allow self driving in the system yet. I have no idea when or even if they will allow it.

 

[30seconds]

Not related to BYD specifically, and more of a personal desire than an actual evaluation of the market, but I'd be very happy with a $10K or even $15K EV in the US, even if the range was only 150 miles. With the $7500 tax credit, even around $20K might be worth it for many, I think. In a way, I think this would be very useful for people who normally buy decent used cars, just need a car for local trips, and don't want to be stuck buying another ICE -- a new EV at used ICE prices. I also view something like that as a great second car, or a great car for an older relative, or a first car for the kiddo, etc. Kids are, of course, creative...but at least the shorter range, and a low top speed, and low acceleration, and low cost, would help put a cap on the amount of trouble they could get into.

150 miles with an LFP battery would still be very useful since you wouldn't be limited by the recommended 80% charge for daily use on other chemistries. And, of course, I'd want modern safety features -- that's the main advantage over finding a similarly priced used car.

And, honestly...even better if it's a Tesla and includes the autopilot safety features and an options for FSD .

A used Model 3 would get you all that plus 200+ mile range around $15m right now.

 

[FastEddieB]

The truth is, nothing has changed, everything is moving forward according to plan.

We can only hope!

There’s an expression that I may not fully understand or agree with: “The Perception is the Reality”.

My “social media” is comprised mostly of Threads and Facebook, having abandoned Twitter some time ago. On both of these, the majority perception of Tesla, the CyberTruck and Musk is one of derision, mockery and overall negativity. This is in spite of algorithms that I think are pushing posts to me based on my interests. Whatever the “plan” is, it’s certainly not resonating with the American public at large. And I think that perception, legitimate or not, may be a large part in why the stock is floundering.

 

[Musskiah]

Waymo is currently the yardstick for autonomy. Can their approach scale ? No. Commercially it’s probably a dead end. But the fact still remains that Tesla would have to reduce disengagements by an order of magnitude to match waymo.

The FSD Tracker site filters by states and disengagement frequency in states with less complex roadways (like AZ for example) is much closer to zero than the average. Tesla could geofence many areas here in Ohio where I drive daily and the frequency is near zero. No doubt Tesla sees these locations and may decide to fence them off for Robotaxis when it is time; they may simultaneously continue to operate supervised FSD in places where FSD needs further training.

Comparing Waymo to Tesla's FSD is like comparing bumper bowling to normal bowling. When the possibility of rolling gutter balls is removed entirely, ability is misrepresented. The handicap must be acknowledged for what it is. Furthermore, if the system is not allowed to make mistakes to find methods of failure, the system cannot learn. Tesla's approach maximizes failure rates (which are dropping) in order to maximize learning. The best approach is to fail fast and fail often to evolve quickly. Leaving the gutters exposed removes the false sense of security which prevents improvement.

 

[Singuy]

Other manufacturers have reported good YoY values for Q1 2024. It's blatantly obvious it's not (just) the interest rates. If BMW can increase their sales 2.5% YoY for Q1, there's no reason to not expect Tesla, who's expecting double digit average annual growth, to do the same.

We have discussed this already. BMW and all car manufacturers had supply issue due to chip shortages, so yes as they go back to normal, they may see a transitory increase in sales before normalize again.

 

[Quesder]

The LT framework should be RoboTaxi business model rather than subscription revenue. Work out revenue and cost per mile and multiply by # miles driven.

Revenue: ~$1-$2 / mile (uber benchmark)
All in operational cost: $0.15-$0.30 / mile (depending on your assumptions around energy price, asset life, etc.)

Early years = Tesla charge $1/mile and pay $0.3 due to low scale until human operated taxis are outcompeted - say (1m vehicles * 200mi/day * 365 days/yr * $0.7 net margin) = $51b in sweet sweet moolah. No matter which way assumptions are stressed in the early years there is a lot left over for us shareholders.

Longer term, per mile costs tend towards all in operational cost while # miles driven goes way up.

I had a hard time to totally understand the math, but since I see $0.7 “margin”, I suppose $51b is gross profit a year.
I am curious and think the number is too big, because current UBER’s market cap is $144B only. So you mean RoboTaxi business can earn a whole UBER market cap in 3 years? That doesn’t sound right.

 

[Krugerrand]

Elon has said that FSD’s price increases as it gets better.

And the prices did increase. Or did you miss that part.

Hilarious the quotes from years ago, ignoring what did happen, and ignoring that nothing is static in life, circumstances change, sometimes overnight.

Such disingenuousness. It’s nauseating.

 

[Musskiah]

An update on Waymo disengagements in California

Each year California requires companies with AV testing permits in the state to report on their disengagements. These reports assist the state in its understanding of how well a self-driving car is doing its job.

waymo.com

https://www.eetimes.com/waymo-cruise-dominate-av-testing/

Looks like Waymo is 1 every 17k miles now.

That is impressive, to be sure.

 

[2daMoon]

That is impressive, to be sure.

A closer look turned up some questions about that report. See Edit in the earlier reply

 

[Singuy]

We can only hope!

There’s an expression that I may not fully understand or agree with: “The Perception is the Reality”.

My “social media” is comprised mostly of Threads and Facebook, having abandoned Twitter some time ago. On both of these, the majority perception of Tesla, the CyberTruck and Musk is one of derision, mockery and overall negativity. This is in spite of algorithms that I think are pushing posts to me based on my interests. Whatever the “plan” is, it’s certainly not resonating with the American public at large. And I think that perception, legitimate or not, may be a large part in why the stock is floundering.

Yes, hence why Musk is trying to get fsd in the hands of skeptics. This is not Teslas first rodeo. Musk has obvious failed himself to an orbital starship, world wide internet with speeds rivaling fiber, fsd that works, the most robust charging infrastructure, and the best selling EV in the world.

 

[ZeApelido]

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…

You can’t look at the FSD price cut in a “for profit” vacuum.

As discussed in this thread by some genius, Tesla’s data needs are growing rapidly as the FSD gets better and as compute Increases.

The price cut may be one way they are using to increase FSD usage and therefore usable data.

https://twitter.com/x/status/1781044658673717248

 

[unk45]

I’m about to ship you my 3 just so you’ll stop opining about this.

Excellent! Please send me the shipping details!

 

[unk45]

From a first principles perspective it seems to me they want more training data, and offering free FSD trials and reduced purchase and subscriptions to FSD is about increasing the number of disengagement data points they can harvest for solving edge cases.

It has not seemed to me this is something being done to increase sales so much as to increase data collection rates. Tesla folk have alluded to how they can comfortably process more data now than ever before. Price drop followed this announcement.

The sale pricing may very well be about getting to full autonomy sooner. Once the goals are achieved they can price FSD at what the market will bear.

The last I heard they can collect data from vehicles equipped with the hardware and software even if they're not subscribers to FSD. Is that not true? Since hearing that I wonder how they evaluate the relevant cases for data use I have no idea. I don't remember where I understood that so I question my logic anyway on that point. I do recall they do use all drivers data in assessing accident risk for underwriting insurance where that is legal to do. e.g. NOT California.

 

[AndrewZ]

You can’t look at the FSD price cut in a “for profit” vacuum.

As discussed in this thread by some genius, Tesla’s data needs are growing rapidly as the FSD gets better and as compute Increases.

The price cut may be one way they are using to increase FSD usage and therefore usable data.

https://twitter.com/x/status/1781044658673717248

Data are always being collected in shadow mode including running FSD in the background and seeing how the driver’s maneuvers etc deviate from what FSD would have done.

FSD purchase/subscription is not required.