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

[Carl Raymond]

A $255 Billion EV Debate Is Raging Among the World’s Biggest Automakers

This article might cheer up investors. Reading between the lines it says: EVs will sell, but only if made by Tesla. For other car makers, the batteries are too expensive.

It omits to say that only Tesla has a supercharger network, and that only Tesla does OTA updates, but you get the feeling they are thinking it.

 

[Vaamo]

Not sure how I feel about posting this given the illustrious author and periodical (LL & BI), and if it was previously mentioned, apologies, but not good if some of this is true (skepticism is welcome):

Insiders describe a world of chaos and waste at Panasonic's massive battery-making operation for Tesla

500k cells a day are a scrap? Isn't the daily capacity around 700k cells?

That is plain lie right there in the beginning of the article . I doubt it that scrap rate is over 5%, which is still a lot.
Considering the number of Model3's produced, it just can't be 500k cells a day as scrap

 

[Fact Checking]

In particular the below suggests that scrap rates are responsible for c.3gWh of the current production shortfall vs capacity:
"Panasonic sends about 3 million battery cells over to Tesla a day, and the pressure is always on to beat previous goals."
"The winding phase is where much of the factory's scrap comes from — amounting to a fairly consistent half a million battery cells a day, said one current and one former employee. "

How did you arrive at 3 GWh/year scrap rate?

The Linette leak says ~3m cells per day, which with 18 Wh/cell means ~54 MWh/day, or 19.7 GWh/year. So that leak was probably from summer last year?

500 thousand cells of 3,000 thousand means a scrap rate of 16.6%.

If we apply that to the 35 GWh/year claim of Panasonic, that's 5.8 GWh/year scrap rate.

Or if we consider Elon's 24 GWh/year figure as the 83.33% net cell production rate, that extrapolates to 28.8 GWh. (Pretty close to the Carsonight estimate of 27.5 GWh/year btw.) That's a scrap rate of ~4.8 GWh/year.

I cannot arrive at a scrap rate of 3 GWh/year with either the Linette, the Panasonic or the Elon numbers.

 

[Mike Smith]

Two (slight) possibilities for Monday:

1) Tesla announces they'll be licensing Autopilot to other car makers (especially in China).

2) Tesla announces they're spinning off Autopilot
-GM raised $5B from Cruise spin off, valuing it at $14.6B. Imagine what Autopilot would be valued at.

It makes no sense to have Autopilot only available in 1-2% of cars. If it makes driving twice as safe then it's a safety feature that should be available in all cars. And Tesla would make a fortune.

 

[Fact Checking]

This also increases the price of non FSD upgraded cars by a lesser amount. Tesla will honour the original advertised upgrade price to current owners. The lower price of FSD upgrade has an option value, and this value will increase when the price of FSD on a new car increases.

Note that Tesla will probably only honor that price to the original owner. So generic resale value won't go up - only "can you please upgrade FSD before we move the car to my account" would achieve that I think, which is possible with family and friends, but probably not with non-trusted strangers such as used car companies buying your used car.

 

[Fact Checking]

Anyone else think Tesla might release the Tesla Network app at the Autonomy event?
It shouldn't take a huge amount of resources to put together a beta app, and it would help to add some substance to Tesla's ambitions, similar to their move to not allow end of lease purchases of Model 3.
It is perfect timing to release a direct Uber/Lyft competitor app in between Lyft and Uber IPOs.

If released the app would most likely be just for show with no use until Tesla actually delivers on full autonomy, but there is also a possibility Tesla could announce imminent launch of an actual Uber competitor with human drivers for now.

Some of the leaks in the past ~6 months would point in this direction, but I'd still put the probability somewhere around 30%.

Lyft/Uber are true cash burning operations: they are only growing because they are loss-leading against taxi companies. The problem is that anyone who can write an app and has money to burn can compete with Lyft/Uber, including existing taxi companies. There's literally zero competitive advantage in their business models and both the Lyft and the Uber valuations are ridiculous. The usual economies of scale simply don't exist for taxi companies, because you cannot scale human drivers in the Lyft/Uber business model. This is why we haven't seen big international taxi companies for 100 years - unlike many other businesses there's no big advantage for taxi companies to go global.

So I fully agree with @neroden that both Lyft and Uber are broken companies as long as they don't have functional, driver-less FSD cars.

The true business model for Tesla to follow is Airbnb - but that requires true FSD. So I think Tesla will (or should) only start Tesla Network if they get permission to run an FSD based service in certain carefully selected areas with a real driver (the customer) picking up the car, but with FSD allowing the car to slowly go back to the owner or to the next pickup point, empty, without a "safety driver".

So I think the Tesla Network will be introduced the following way:

"50% FSD Tesla robo-taxis with no dedicated safety drivers": a completely empty, driver-less and customer-less Tesla car in the Tesla Network will be FSD driving to the next pick-up point. It will be driving very carefully and timidly like Advanced Summon and NoA does today, but in larger city environments. Once customers at the pickup point enter the Tesla and start using it, one of them is expected to be the "safety driver", and they are responsible for and liable for driving safely during the trip. Once they exit the car at the destination, the car enters 'empty car FSD mode' again, and drives to the next pickup point.

This is a lot less risky in the legal sense, and FSD driving timidly won't annoy customers either - and that might make Tesla Network a reality. If the car gets stuck somewhere and has to pull over, or runs out of charge or gets stuck due to weather, then the owner (or a Tesla Network service technician) can still pick it up. A bit of "remote driving" done by trained Tesla Network personnel might also be permitted in certain jurisdictions and geographical areas, to safely park the car until a human can pick it up.

Note what a genius of a business model this kind of "50% robo-taxis" approach would be:

• Most taxi customers, especially in the U.S., can actually drive cars, they just don't have their own cars around right then.
• Most would actually love and enjoy driving a Tesla.
• (Note the advertising effect: anyone can "test drive" a Tesla, simply by hailing a Tesla Network car... This puts the attempted elimination of the test drives at Tesla Stores in a different light.)
• I'd expect a combination of Chill Mode, Valet Mode and other safety measures to be present to avoid abuse, including full Sentry Mode while a customer is in the car to make sure any damage to the car is documented correctly.
• The Tesla Network might also offer a tutorial about how to drive a Tesla and limit acceleration during first recorded use of a Tesla by a new "driver", to make sure it's a pleasant experience for everyone.
• Customers who don't want to or cannot drive a Tesla (medical condition, alcohol consumption, etc.) might also request a dedicated safety driver to be present, with higher fares.
  
• Lyft/Uber wouldn't be able to compete with Tesla Network prices and economics, because the Tesla Network wouldn't be using any dedicated "safety drivers" or "driver", at all.
• Even if Waymo removed the safety driver they would find it hard to compete with Tesla, given how large the Tesla fleet already is, and how much cheaper Tesla's FSD solution is compared to the LIDAR solutions of the competitors.

And yes, if Tesla can pull off this FSD model then raising the price of "professional FSD" (which allows your car to drive in the Tesla Network) to $10k or more would be absolutely warranted, plus a 30% share of revenues generated, plus various services offered (such as cleaning/maintaining the car).

If Tesla can pull this off, it's a gold mine, and I think Wall Street, despite their dislike and ignorance of Tesla, is going to finally recognize it as well. (I'm wondering whether @KarenRei agrees with this approach: the safety requirements for an empty FSD car driving to the next pickup point "carefully and timidly" are obviously different and can be weaker.)

Do I think that this is possible today with their existing neural networks, using the HW3 FSD computer? Yes, if limited to certain geographical areas and routes which are carefully vetted by Tesla for the car to be able to FSD-driving around empty in a safe fashion without endangering others. Then they can grow the areas/routes they support.

True historic photo of the Tesla Network:

If Tesla refines this "50% FSD" approach then the ICE vs. Tesla race would be over within 1-3 years, even without "complete" FSD functionality.

Whether Tesla is going to introduce such a Tesla Network ride-sharing and taxi business on April 22, I have no idea - but if they do and if it's offering real rides and a timetable to introduce the service then it would certainly create a justified buzz and wouldn't just be seen as hype.

 

[Carl Raymond]

Two (slight) possibilities for Monday:

1) Tesla announces they'll be licensing Autopilot to other car makers (especially in China).

2) Tesla announces they're spinning off Autopilot
-GM raised $5B from Cruise spin off, valuing it at $14.6B. Imagine what Autopilot would be valued at.

It makes no sense to have Autopilot only available in 1-2% of cars. If it makes driving twice as safe then it's a safety feature that should be available in all cars. And Tesla would make a fortune.

To stay within mission, change “in all cars” to “in all BEVs”.

Making ICE cars cheaper to run (autonomously) makes it harder to retire them.

I’m assuming that if they go to the trouble of making an autopilot car they also make it a FSD car.

The Tesla brand is untainted by tailpipe association. It needs to stay that way.

 

[KSilver2000]

Once customers enter a Tesla Network car at a pickup point and start using it, they are expected to be the "safety driver", and they are responsible for and liable for driving safely

I agree with everything you said. It’s one of the better ideas I’ve read so far ok slowly rolling out utilization of FSD when time comes.

Just one question: How would the TN customer be held liable as the safety driver? Every person who signs onto the TN program needs to provide an active automobile insurance during sign-up? That would negate the objective as many people living in urban areas have no insurance as they don’t own a car, and thus a reason for needing TN service. Or do customers pay an extra premium for “per use insurance”? That sounds far too expensive to use TN over other ride services.

I’m imagining the annoyance I hear listening to car rental agents explain the insurance. I would hate to be presented with such jargon every time I get into a TN car.

 

[Fact Checking]

Just one question: How would the TN customer be held liable as the safety driver? Every person who signs onto the TN program needs to provide an active automobile insurance during sign-up? That would negate the objective as many people living in urban areas have no insurance as they don’t own a car, and thus a reason for needing TN service. Or do customers pay an extra premium for “per use insurance”? That sounds far too expensive to use TN over other ride services.

So in the initial form (but I'm really just guessing here), I'd imagine something like this:

• Tesla Network could self-insure and the customer EULA could act as a binding contract to make the driver liable for damages that they cause. They'd display it prominently, there might even be a short voice message before the driver can operate the car, and voice recognition would require the driver to answer "yes, I agree", and maybe sign their name on the screen or something. They'd be required to have a valid driver's license, and they have to be legally able to drive at that moment. (True for 90%+ Americans.)
  
• Tesla Network might also require binding arbitration process before having to sue a customer for material damages. This usually makes it much cheaper and more reliable to recover small claims (which most traffic accidents really are).
  
• "Always On Sentry Mode" recordings would offer robust evidence that regular insurance companies almost never have - so even during a major accident or major damage they'd have objective evidence from several cameras, plus telemetry and an internal camera recording. This is why I think they rolled out Sentry Mode so quickly.
• Owners could be offered various convenience features, such as a new "loaner" car the moment there's a customer-induced accident with the car. I.e. it won't stop generating revenue just because a customer crashed the car.

So I think they can reduce their actual liabilities and legal costs immensely via these technological measures, plus they'd observe it during the prototype phase to see how much of an actual problem it is. The key observation is that video and audio evidence is both a powerful deterrent against abuse, and excellent evidence in any dispute resolution process.

Tesla also being a car company gives them unparalleled degrees of freedom to make this both profitable and convenient.

 

[lascavarian]

This is a lot less risky in the legal sense, and FSD driving timidly won't annoy customers either - and that might make Tesla Network a reality.

Interesting, but I have some reservations. There could be an annoyance factor with timid FSD.

The part I do like about timid driving is that you are pounding on the safety angle.

I think a case can be made for a safety augmented mobility service with a driver. Shadow mode becomes Safety Augmented Mode

 

[Droulas]

Maybe on 22th they will start the coast to coast fsd trip which will end 48 hours later, right on time for the financial results of the first quarter ?

 

[Fact Checking]

Interesting, but I have some reservations. There could be an annoyance factor with timid FSD.

It's an annoyance only if the "empty cars" are driving during rush hour (holding up traffic due to their timidity) - but note that most traffic during rush hour would be customers - most of whom would want to travel in the same direction as rush hour traffic.

The return trip of the 'empty car' would often be counter-cyclical, i.e. not in the "rush hour direction".

At least in the prototype phase Tesla could also use smart routing and avoid rush-hour traffic. This would be done in the Tesla Network data center, where I assume they'll have access to high grade real-time traffic metrics. Tesla would also offer their own cars into the Tesla Network fleet, so by over-allocating cars to a given route they can increase coverage and 'buffer' cars to meet demand, without having to drive empty cars there in high traffic.

Note that the larger Tesla fleet of 600,000+ cars will also offer real-time traffic metrics, which the Tesla Network could utilize ...

 

[WarpedOne]

Stop, just stop.
This is just to good to not be disappointed on 22nd.

 

[lascavarian]

Just rolling out ideas here...

Safety augmented mode implies a premium service while putting other ride services at a disadvantage if they don’t have it. We have 8 cameras and a NN so how does Uber etc compete?

The passenger/guest decides who has driving priority according to their comfort level. Augmented mode with the human as priority control or only as backup as needed.

Network includes remote monitoring as an option for a fee. Maximize guest comfort level according to their preference and wallet.

 

[ReflexFunds]

How did you arrive at 3 GWh/year scrap rate?

The Linette leak says ~3m cells per day, which with 18 Wh/cell means ~54 MWh/day, or 19.7 GWh/year. So that leak was probably from summer last year?

500 thousand cells of 3,000 thousand means a scrap rate of 16.6%.

If we apply that to the 35 GWh/year claim of Panasonic, that's 5.8 GWh/year scrap rate.

Or if we consider Elon's 24 GWh/year figure as the 83.33% net cell production rate, that extrapolates to 28.8 GWh. (Pretty close to the Carsonight estimate of 27.5 GWh/year btw.) That's a scrap rate of ~4.8 GWh/year.

I cannot arrive at a scrap rate of 3 GWh/year with either the Linette, the Panasonic or the Elon numbers.

It is just 18Wh/cell * 0.5 * 365 = 3.3 GWh. Although it is not clear exactly how Tesla defines the cell's capacity, most likely between 17 and 18Wh i think.
Knowing Linette, I would guess 3 million good cells per day is likely rounded down from anywhere between 3 and 4 million. 3.6 million per day would match Elon's 24GWh comments.
So perhaps 4.1 million cells produced per day of which 0.5 million are rejected, so a 12.1% scrap rate. This is also close to Carsonight's estimate of 4.2 million cells per day.

So reducing this scrap rate could maybe take Panasonic to 27GWh. I'd guess the rest of the shortfall vs 35GWh capacity is ramping up cathode production/supply, reducing line downtime and ramping up staffing.

 

[Dan Detweiler]

It's an annoyance only if the "empty cars" are driving during rush hour (holding up traffic due to their timidity) - but note that most traffic during rush hour would be customers - most of whom would want to travel in the same direction as rush hour traffic.

The return trip of the 'empty car' would often be counter-cyclical, i.e. not in the "rush hour direction".

At least in the prototype phase Tesla could also use smart routing and avoid rush-hour traffic. This would be done in the Tesla Network data center, where I assume they'll have access to high grade real-time traffic metrics. Tesla would also offer their own cars into the Tesla Network fleet, so by over-allocating cars to a given route they can increase coverage and 'buffer' cars to meet demand, without having to drive empty cars there in high traffic.

Note that the larger Tesla fleet of 600,000+ cars will also offer real-time traffic metrics, which the Tesla Network could utilize ...

Boy this all sounds great, but...

Can you imagine the FUD storm this would generate? Don't get me wrong, ANYTHING Tesla does will create a FUD storm. It's just what they do. Tesla needs to just do what is right for the company and its customers but this would make me swear off YouTube and all other social media as I know the internet would explode with crap! Let's just make sure we don't set ourselves up for disappointment. Monday can't come soon enough.

Dan

 

[lascavarian]

The return trip of the 'empty car' would often be counter-cyclical, i.e. not in the "rush hour direction".

Maybe returning vehicles could use certain routes or lanes. It might have to be customized based on the local conditions. I’m thinking about it but still worried about acceptance if FSD is not better in every way.

Appreciate the thinking!

 

[j6Lpi429@3j]

i don't believe any promises or timelines will be given for the tesla network on the 22nd. They will show off some improved capabilities of HW3, and much more importantly, they will stuff a hundred+ fund managers into those cars with or without a safety driver and go on some pre-arranged trip.
All they really need is for people who make investment decisions to be whisked around in a car with the steering wheel spinning and it doing the driving. Thats enough.

You need to overcome the 'tesla are not leading in self driving' bullshit before you even THINK about actively pushing the uber-competitor narrative with any seriousness.

 

[schonelucht]

Shorts dug up a Form 144 filed April 15th by Gracias that he is going to sell another 50k shares over the next weeks.

 

[Earlian]

Elon really tried to change from ‚overpromise’ to ‚underpromise’ and ‚deliver‘, but that also didn’t help.

Why?

Because now we are changing to overexpect upfront every upcoming event and leave no space for deliver a surprise. Everything possible, impossible and alienlike feature is discussed and expected to come true, so there’s only more disappointment to unfold.

For the 22nd, I expect showing some FSD features like Navigat on Autopilot or Advanced Summon to be presented to investors, who also may test the system and ‚maybe‘ some teaching about the new HW3 with a bit of a timeline. Nothing more!

This leaves room for getting satisfied even if there is no suprise!

Please calm down everyone.