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

[Gigapress]

The Cyberlandr camper designed to mount on Cybertruck is still happening and they have a physical prototype now. They’ve been working with Munro & Associates for designing and industrializing this. Sneak peek in this video.

The North American recreational vehicle and camper market is tens of billions of dollars annually today. I expect this to be a significantly popular use case for Cybertruck because it’ll be such a big improvement on the entire experience for towing and driving dynamics, climate control, electric power for appliances, fuel savings, and the option for solar panels to recharge the battery while camping.

 

[Krugerrand]

No, it is fine if all of this escapes them. let the discussion elevate, not dumb down.

Least Common Denominator LCD discussion degrade to gordo and memes.

There is nothing dumb about pizza, visual learners, ADHD afflicted individuals, or not having had enough coffee.

It might, however, be considered a bit thick to not have grasped the blatant humor.

 

[Ogre]

This is why I have a tough time modeling for future products not yet available.

I don’t think you should or need to.

Tesla is a solid investment based on auto + energy. Maybe there are better investments out there but it’s quite good.

The Robotaxi/ Optimus wildcards are lottery tickets that make it irresistible.

 

[TN Mtn Man]

Ok, you are wrong

(well, the dilution is kind of a thing- though a bit more complex since it's mostly diluting the shares the hostile takeover person owns.... as everyone ELSE gets the chance to buy new shares at a discount but that person does not- but I mean wrong on the vote part).

As a few folks have mentioned, twitters articles of incorporation include surprisingly broad powers for the board to fight takeover attempts [B}without[/B] any shareholder approval needed.... one of the reasons it's called out in the articles filed with the SEC is so a potential investor is aware the board has such powers available and can act without their (the shareholders) approval on these matters in certain ways and that can inform their decision to invest in that company or not.

As another example- the board can (without approval from anyone) issue a bunch of preferred shares, with whatever voting rights they specify, as a measure to block a takeover. (this is not unique to twitter and is sometimes called a Blank Check Preferred Stock defense) where they say issue X shares of this stock and each share gets 1000 votes instead of the 1 vote a common share gets.





The idea there's a political bias has been pretty thoroughly debunked- not just by the reddit CEO but lots of examples of folks of both leanings making censorship claims... and not just on this specific platform. It's not "political speech" that's banned- it's wildly out of bounds behavior that happens to include political speech.






They do have a fiduciary responsibility, but it'd be up to courts, usually after the fact, to determine if the measures they took, within the allowed bounds of their granted powers, were in the financial interests of shareholders or not.

It's not nearly as simple as "Elon offers more than the current price, so if they don't take it the shareholders got shafted"-- if it was that simple every takeover bid above current price would have to get approved and that's not remotely the case.

So any anti-takeover measure they take- from rejecting the offer, to poison pills, to the blank check thing I mention, and more- is a balancing act on their part regarding what they CAN legally do up front to block the takeover (which is a lot) versus what they believe they could successfully defend in court after the fact as not violating their fiduciary responsibilities.

Good to know, thank you for the explanation. Surprising to me that actions like this that are narrowly targeted as a specific owner (or owners if a specific block of shareholders) would be considered ethical, let alone legal.

 

[Knightshade]

Good to know, thank you for the explanation. Surprising to me that actions like this that are narrowly targeted as a specific owner (or owners if a specific block of shareholders) would be considered ethical, let alone legal.

So I'm not a specialist in securities law or anything but my understanding in general on this type of stuff is it's legal because it's not targeted at a person, but a class of shareholder.

Like they couldn't say "If Elon Musk specifically buys more than X % of the shares, this triggers some stuff"

But they can say 'If anybody buys more than X % of shares, this triggers some stuff" even when only Elon Musk has suggested being interested in doing that.

And they can try and justify that in court later if it comes to it with some argument around they felt it was in the best interests of the company to discourage any single shareholder from accumulating that much power or whatever their argument might be.


The other thing here is- if the rule itself is sufficient to discourage the takeover bid- then the rule never triggers, and there's no damages or dilution for any shareholder to sue over around the rule. Lots of past examples where just adopting a rule like this killed the attempt and never triggered. That's likely their current hope.

 

[Rarity]

O'Dowd is a moron.

Californians love there Teslas, and they love Elon. This political stunt might work someone else, but not in California.

But, free advertising for Tesla, which will get Elon more Twitter followers, which will sell more cars.

We are going to see a demonstration of the comparative inefficiency of advertising and the power of word-of-mouth marketing. We have seen this dynamic for the last decade with Tesla and why would it change?

For some reason, politicians love advertising. Embedded in that article is the fact that you have to spend a ton of money on advertising to move the needle in California. To translate that into everyday speech: advertising is inefficient. Dowd is throwing his money down the rathole, even if he's getting a candidate's discount for it.

 

[Accident]

We know Tesla has similar contracts. I wonder if we’ll see anything similar to this with the Q1 results:

Volkswagen reaps unexpected $3.8 billion windfall on nickel price spike

The carmaker, hit by a range of component shortages led by semiconductors, also warned of renewed pressure on already fragile supply lines.

www.mining.com

 

[TheTalkingMule]

Heated discussion on the nature of human existence. Quite the bullish TMC indicator! Is it Monday yet?

 

[EVDRVN]

We are going to see a demonstration of the comparative inefficiency of advertising and the power of word-of-mouth marketing. We have seen this dynamic for the last decade with Tesla and why would it change?

For some reason, politicians love advertising. Embedded in that article is the fact that you have to spend a ton of money on advertising to move the needle in California. To translate that into everyday speech: advertising is inefficient. Dowd is throwing his money down the rathole, even if he's getting a candidate's discount for it.

Well he is charismatic and a likable person

 

[jhm]

It's probably too early to be mathing, but independent odds do multiply, and dependant add. So, in that, @MABMAB is correct.
Odds of rolling 2 sixes is 1/6 * 1/6, each due to each roll being independent. Odds of both die being 5 or 6: (1/6 * 1/6) + (1/6 * 1/6) since it must not be 5 in order to be six.

Odds of a truck with FSD: .5*.5=25%
Odds of a truck and no FSD: .5*(1-.5)=25%
Odds of no truck, but we get FSD: (1-.5)*.5=25%
Odds of at least one of truck and FSD happening=25%+25%+25%=75%
More simply: 1-((1-.5)*(1-.5))=75% i.e. 100% minus odds of both failing.

If we ignore the FSD/taxi linkage, the odds of none of those things happening is (1-.25)*(1-.5)*(1-.5)*(1-.25)=0.14 meaning the odds of at least one highly profitable product being produced is is 1-.14 or 86%.

Let's all be very careful about multiply probabilities. That works only if the event are truly independent. But to the contrary the dependencies here are quit substantial. To understand that we need to think in terms of conditional probabilities.

Much depends on whether FSD is a a success. If FSD is a failure, there is practically no chance that robotaxis will succeed. But if FDS is a success, then there is a higher chance that robotaxi will succeed.

Accounting for conditional probabilities looks like this, and I'll illustrate with made up numbers.

P(Robotaxi) = P(Robotaxi & FSD) + P(Robotaxi & not FSD)
= P(Robotaxi | FSD) P(FSD) + P(Robotaxi | not FSD) P( not FSD)
= 50% × 50% + 0% × 50%
= 25%

So note that the conditional probability of Robotaxi given FSD, P(Robotaxi | FSD) = 50%, is not the sames as marginal probability, P(Robotaxi) = 25%. Anyone multiplying two marginal probabilities to get a joint probability is wrong.

Similarly, we need to consider the conditional probabilities of Bot or Truck conditioning on FSD. In my view if FSD is successful this improves the probability that Bot will succeed (because Tesla would have made more progress with AI) and decreases the probability of success with Trucks (because the opportunity to make money with FDS, Robotaxis and Bots would be more attractive than trucks and compete for priorities). Expressed mathematically my belief is

P(Bots | FSD) > P(Bots | no FSD)
P(Trucks | FSD) < P(Trucks | no FSD)

My basic thinking is that pursuing Trucks aggressively (both Cybertruck and Semi) is a really important fallback if Tesla fails to make serious progress with AI. But in the more optimistic scenario where Tesla makes serious AI progress, this is first of all monetized through FSD on private vehicles. The next more immediate, but riskier way to monetize this is through Robotaxis. The big risk here is getting regulatory approval to deploy Robotaxis. So as a fallback to political barriers, Bots can monetize Tesla's most advanced AI capabilities.

So it may be helpful to model this more like a decision tree for Elon to navigate. Elon is masterful at creating multiple paths to nearterm and long-term success. He likes to have lots of options to choose from. For example, stationary storage and Tesla Energy are great options for dealing with any short-term surplus of cell capacity and for long-term value creation. They've largely been on the back burner because cells are in short supply and the Models Y and 3 are in the near term more attractive areas to grow the enterprise. So the idea that Tesla needs to hit all of FSD, Trucks, Robotaxi and Bots in some window of time misses just how strategic and adaptive Elon is. Elon will seize upon whatever the best opportunities are at the time while continuing to nurture bigger opportunities for the future.

In this regard, it is telling that Elon is pressing for massive scale up of Models Y and 3 this year. So FSD is not quite ready to be the focal point of growth. Trucks are pushed off to next year. Would they get push off even later is FSD advanced to Robotaxi level this year? When Elon says that Bots are the most important product to develop right now, I'm inclined to think that Robotaxi AI may require advances that must be learned via the Bot form factor or that the politics with Robotaxis are just too formidable in the near term. Either way Bots may become commercially viable before Robotaxis. Regardless how advanced Tesla AI gets monetized, Tesla does need to keep driving hard on the battery supply chain to get to 3TWh ASAP. Batteries and AI are parallel tracks. AI basically enables Tesla to get the highest profit margin possible for whatever battery supply they can muster.

BTW, a certain social media platform may be yet another form factor for advanced AI, batteries not required.

 

[wtlloyd]

I agree with you completely about Elon accomplishing the "impossible" over and over again - well said.

With regards to how Elon manages to create and sustain such high performing teams, I also agree we don't know much... besides inuendo about how severe (but fair?) he is in reacting to staff dumb ideas (couldn't be more cruel than Jobs though!)

Eric Schmidt & Jonathan Rosenberg wrote the book (literally) about how Google works. I'm reading it now and actually a bit disappointed because it's more about management principles than the history of Google. I picked the wrong book for what I wanted to learn about Google, but the book is insightful IMO.

Dave Lee started down the path of asking about Elon's management processes, but Elon didn't like (or understand) what he was asking and we didn't learn anything:

https://twitter.com/x/status/1488050840095047688

So this critical insight about 'how Elon works' will stay publicly unknown for now. Hopefully one day there will be a bio where Elon looks back and shares his management principles and insights. I loved Ashlee Vance's Musk bio (after which I was convinced to buy TSLA in 2018), but it was more a history of than how Elon manages.

I believe the book you are referring to was already written some years ago - I had the good fortune to build an extensive home library for the couple who conceptualized and authored the concept "Hot Groups". Reading just a bit of a synopsis should cause a shiver of premonition and recognition.
Read this page below. It's all there, back in 1995.

How to Cultivate a "Hot Group" to Ignite An Organization

www.gsb.stanford.edu

 

[Hock1]

Ok, you are wrong

(well, the dilution is kind of a thing- though a bit more complex since it's mostly diluting the shares the hostile takeover person owns.... as everyone ELSE gets the chance to buy new shares at a discount but that person does not- but I mean wrong on the vote part).

As a few folks have mentioned, twitters articles of incorporation include surprisingly broad powers for the board to fight takeover attempts [B}without[/B] any shareholder approval needed.... one of the reasons it's called out in the articles filed with the SEC is so a potential investor is aware the board has such powers available and can act without their (the shareholders) approval on these matters in certain ways and that can inform their decision to invest in that company or not.

As another example- the board can (without approval from anyone) issue a bunch of preferred shares, with whatever voting rights they specify, as a measure to block a takeover. (this is not unique to twitter and is sometimes called a Blank Check Preferred Stock defense) where they say issue X shares of this stock and each share gets 1000 votes instead of the 1 vote a common share gets.





The idea there's a political bias has been pretty thoroughly debunked- not just by the reddit CEO but lots of examples of folks of both leanings making censorship claims... and not just on this specific platform. It's not "political speech" that's banned- it's wildly out of bounds behavior that happens to include political speech.






They do have a fiduciary responsibility, but it'd be up to courts, usually after the fact, to determine if the measures they took, within the allowed bounds of their granted powers, were in the financial interests of shareholders or not.

It's not nearly as simple as "Elon offers more than the current price, so if they don't take it the shareholders got shafted"-- if it was that simple every takeover bid above current price would have to get approved and that's not remotely the case.

So any anti-takeover measure they take- from rejecting the offer, to poison pills, to the blank check thing I mention, and more- is a balancing act on their part regarding what they CAN legally do up front to block the takeover (which is a lot) versus what they believe they could successfully defend in court after the fact as not violating their fiduciary responsibilities.

Exactly. In terms of the potential of success in thwarting (in this case) a reasonable offer in court, there is a big difference between having a poison pill in place and cooking one up in reaction to a current offer.

 

[Buckminster]

Too much negativity in this thread today. FSD (and therefore robotaxis) is a solved problem - they will keep knocking down the remaining challenges as they have done so successfully up until now. Many of you have become rich beyond your dreams and think it impossible that TSLA will keep going up because you have never been this lucky before. It is more logical to believe that the trend will continue. Tesla will execute and TSLA will go up more than you think. The sky is not the limit.

Stop it or I will start posting super bull poems to drown out the rot.

 

[strago13]

Couple thoughts on our last two road trips to and from Florida (MD start). 2021 Model Y LR w/ FSD (no, not beta, they wont let me in!). First trip was in January for a wedding.

Traveling on cold days is terrible, we started the trip with the temps in the 20's and it really is not good on the range, even in the Model Y w/ the heat pump. It just saps energy where I need to buffer 80-100 miles of extra range just to make my next charging stop. FSD in January was wretched and almost unusable because of all the phantom breaking issues. Since summer of 21 there are more super chargers on the way down, but from Halifax NC-> Santee, South Carolina is a big hole of 250kwh chargers and you are going to get stuck with a crappy 150kwh one that is probably only 8 stalls, which means you are probably sharing at 75 kwh an hour.

On our trip last week, FSD was MUCH improved to the point where it did the majority of the driving for us. Charging is still an issue. Looking at the map, it looks like 3 new ones should pop up in 2022 to help ease the issue, but its still not ideal. The other problem with some of these chargers is they are not in places where there are 24/7 facilities (one is at a grocery store). One other nuance I noticed is the car either kept trying to reroute me to other stops or say I couldnt make it where I knew I could. This is quite frustrating as I know that if I stop at supercharger XXX I get 250kwh chargers and food/rest stops that I want. Not just whatever the car chooses for us. It seems like they added some logic about how busy a supercharger is in to its dynamic routing.

To me it also felt like after leaving home on full, charging at a 250kwh charger when I was near empty went pretty quickly. However, the next charge felt slower and not as ideal I think largely because of the 250kwh charger gaps on the I95 corridor. In my dream world, the Cyber Truck will get ~500 miles of range and I only need to stop once to charge where the other stops are likely just food/bio breaks. Sometimes I just want to power through but cant due to that issue. One gentleman I was talking to had a Plaid Model S and he raved about how much easier it was because he had ~450 miles of range.

Anyways, in general, I think things are better now than they were on the January trip, fingers crossed it gets even better on our next trip.

 

[Usain]

We are going to see a demonstration of the comparative inefficiency of advertising and the power of word-of-mouth marketing. We have seen this dynamic for the last decade with Tesla and why would it change?

For some reason, politicians love advertising. Embedded in that article is the fact that you have to spend a ton of money on advertising to move the needle in California. To translate that into everyday speech: advertising is inefficient. Dowd is throwing his money down the rathole, even if he's getting a candidate's discount for it.

I do have to admit, it's a creative (ab)use of election law. He gets preferential time slots and preferential advertising rates. So just by making a bogus run for office he can spread his self-serving message more efficiently.

It's sleazy, but it's kind of brilliant.

(And I agree with Rarity that ultimately, it won't work.)

 

[samn]

I don’t understand why you can’t or haven’t connected the dots.

Why do we need to transition to sustainability in the first place? Hint: because humans are cesspools and are happy to behave badly when they aren’t held accountable for their actions. That’s exactly what’s going on on social media.

Why did Tesla not just make EVs, but instead also went down the energy creation and energy storage path as well? Why did Elon not just start Tesla but also SpaceX to take humans off the planet? Why is Elon working on a method to make human consciousness permanent via an AI uploading platform? Why is he doing Optimus with a view to make everything in life bountiful for all humans (per his recent TED talk)? Hint: for the same reason he’s now looking at Twitter; it’s all the connected human condition.

You can’t fix one element, but not the rest. Won’t work. If people won’t willingly change, then you either force them by taking away all options, and/or you create a world and a life that naturally changes the human condition. What is there to be angry about if every human has everything in life they need without having to fight and claw and struggle?

Maybe that’s naive of Elon. Maybe the human condition will always be a cesspool. But he’s going to die trying and I’m okay with that.

Nominate for quotes of unusual merit (or whatever it is called). Well said and insightful

 

[Gigapress]

Let's all be very careful about multiply probabilities. That works only if the event are truly independent. But to the contrary the dependencies here are quit substantial. To understand that we need to think in terms of conditional probabilities.

Much depends on whether FSD is a a success. If FSD is a failure, there is practically no chance that robotaxis will succeed. But if FDS is a success, then there is a higher chance that robotaxi will succeed.

Accounting for conditional probabilities looks like this, and I'll illustrate with made up numbers.

P(Robotaxi) = P(Robotaxi & FSD) + P(Robotaxi & not FSD)
= P(Robotaxi | FSD) P(FSD) + P(Robotaxi | not FSD) P( not FSD)
= 50% × 50% + 0% × 50%
= 25%

So note that the conditional probability of Robotaxi given FSD, P(Robotaxi | FSD) = 50%, is not the sames as marginal probability, P(Robotaxi) = 25%. Anyone multiplying two marginal probabilities to get a joint probability is wrong.

Similarly, we need to consider the conditional probabilities of Bot or Truck conditioning on FSD. In my view if FSD is successful this improves the probability that Bot will succeed (because Tesla would have made more progress with AI) and decreases the probability of success with Trucks (because the opportunity to make money with FDS, Robotaxis and Bots would be more attractive than trucks and compete for priorities). Expressed mathematically my belief is

P(Bots | FSD) > P(Bots | no FSD)
P(Trucks | FSD) < P(Trucks | no FSD)

My basic thinking is that pursuing Trucks aggressively (both Cybertruck and Semi) is a really important fallback if Tesla fails to make serious progress with AI. But in the more optimistic scenario where Tesla makes serious AI progress, this is first of all monetized through FSD on private vehicles. The next more immediate, but riskier way to monetize this is through Robotaxis. The big risk here is getting regulatory approval to deploy Robotaxis. So as a fallback to political barriers, Bots can monetize Tesla's most advanced AI capabilities.

So it may be helpful to model this more like a decision tree for Elon to navigate. Elon is masterful at creating multiple paths to nearterm and long-term success. He likes to have lots of options to choose from. For example, stationary storage and Tesla Energy are great options for dealing with any short-term surplus of cell capacity and for long-term value creation. They've largely been on the back burner because cells are in short supply and the Models Y and 3 are in the near term more attractive areas to grow the enterprise. So the idea that Tesla needs to hit all of FSD, Trucks, Robotaxi and Bots in some window of time misses just how strategic and adaptive Elon is. Elon will seize upon whatever the best opportunities are at the time while continuing to nurture bigger opportunities for the future.

In this regard, it is telling that Elon is pressing for massive scale up of Models Y and 3 this year. So FSD is not quite ready to be the focal point of growth. Trucks are pushed off to next year. Would they get push off even later is FSD advanced to Robotaxi level this year? When Elon says that Bots are the most important product to develop right now, I'm inclined to think that Robotaxi AI may require advances that must be learned via the Bot form factor or that the politics with Robotaxis are just too formidable in the near term. Either way Bots may become commercially viable before Robotaxis. Regardless how advanced Tesla AI gets monetized, Tesla does need to keep driving hard on the battery supply chain to get to 3TWh ASAP. Batteries and AI are parallel tracks. AI basically enables Tesla to get the highest profit margin possible for whatever battery supply they can muster.

BTW, a certain social media platform may be yet another form factor for advanced AI, batteries not required.

Great summary and proper usage of Bayes’ theorem for conditional probabilities!

I think this inequality should be flipped though:

P(Trucks | FSD) < P(Trucks | no FSD)​

Likewise I think this is the opposite. “it is telling that Elon is pressing for massive scale up of Models Y and 3 this year. So FSD is not quite ready to be the focal point of growth.”

If autonomy is really coming this year as Tesla leaders ostensibly believe, then the optimal strategy is to grow the fleet as fast as possible while FSD development matures, so that when the software is ready more vehicles can run it from day one.

The Tesla Semi becomes an order of magnitude more valuable with autonomy, and Elon said so at the Semi reveal presentation in 2017. If they have this working they’ll move heaven and earth to scale manufacturing because materials and labor cost barely even matter if they have a product so valuable.

Autonomy Advantages:

• Delete direct driver-related expenses incl wages, benefits, etc.
• Delete mandatory driver rest periods —> Faster deliveries & Better ROI for semi purchase
• Convoys can draft each other’s airflow optimally like race cars or birds
• Acceleration, braking and turning optimized by AI and physics simulator to maximize energy efficiency and speed objectives
• Cabin can be deleted
  • No windshield, seat, doors, mirrors, frunk, climate control, dashboard, sound system, floor, airbags, nor steering column
  • Tons of weight removed —> More cargo capacity
  • More freedom to optimize front end of vehicle for low aerodynamic drag coefficient
  • More free space for batteries —> Longer range & Faster charging possible
  • Front end could be shorter enabling tighter turn radius for e.g. urban deliveries
• Reduced collision insurance costs
• More reliable delivery schedules
  • No stoppages from medical emergencies
  • Far lower frequencies of collisions, tipovers, and jackknifing

In the event that Tesla has a legitimately functional Optimus robot and the limiting factor on having more of them is simply battery supply, then the overwhelming economic value of Optimus would surely lead to Tesla being willing to pay extremely high prices for battery minerals and factories and in a matter of a couple of years will be able to accelerate supply. They would outbid basically everyone in the whole market. And they’d begin this process before Optimus is ready for prime time but after progress is convincing enough that they are confident enough to bet big on scaling.

 

[MikeAtkinson]

Similarly, we need to consider the conditional probabilities of Bot or Truck conditioning on FSD. In my view if FSD is successful this improves the probability that Bot will succeed (because Tesla would have made more progress with AI) and decreases the probability of success with Trucks (because the opportunity to make money with FDS, Robotaxis and Bots would be more attractive than trucks and compete for priorities). Expressed mathematically my belief is

P(Bots | FSD) > P(Bots | no FSD)
P(Trucks | FSD) < P(Trucks | no FSD)

I disagree about that. Success in FSD adds a capability to Truck and Semi that will increase their competitiveness. Even though Robotaxi and Bot will be more attractive products, that does not mean that Tesla will not make a success out of Truck or Semi, Tesla will be able to do many things at the same time.

Also I think P(Trucks) = 1, that is there will be a cybertruck. Your argument above uses point probabilities, which I think is wrong. Instead there is a probability density function (PDF) of number of cybertrucks sold per year. The mean value of that PDF is higher if FSD works (reaches level 5), or to put it another way the entire PDF is shifted to higher numbers of trucks sold for each year if FSD is successful.

 

[friit]

I got my first Tesla a couple of weeks ago (Model 3). It's the first electric vehicle in my family and circle of friends here in Estonia, so lots of explaining is to be expected. But I did encounter a new one I hadn't seen on here - my very wealthy older acquaintance was surprised I got the car so fast (6 long long months from order to delivery), and thus surmised that it must not be a very popular car after all. My attempts to assure them that it's actually quite wildly popular was unconvincing to them.

I realised that I actually have no idea what the wait times for cars are. Usually people buy what is available at the dealer, but perhaps if you are after a particular build it takes a really long time? Anyway, that caught me off guard a bit.

Other most common misconceptions are that it's a 100k car and the quality is bad. My car was half that and absolutely bloody perfect.

 

[Surfer of Life]

The Cyberlandr camper designed to mount on Cybertruck is still happening and they have a physical prototype now. They’ve been working with Munro & Associates for designing and industrializing this. Sneak peek in this video.

The North American recreational vehicle and camper market is tens of billions of dollars annually today. I expect this to be a significantly popular use case for Cybertruck because it’ll be such a big improvement on the entire experience for towing and driving dynamics, climate control, electric power for appliances, fuel savings, and the option for solar panels to recharge the battery while camping.

Nice! But what a tedious long-winded video. 2x speed necessary.