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

[Krugerrand]

We heard Tesla operated about ~100 units (incl. with their Pepsi partner). How many more could they need to properly test the semi in real-word conditions?

They need as many as they need. It’s just as likely they’ve been making changes during this trial period, and then testing the changes.

If they can speed up construction for the semi plant, why wouldn't they not accelerate today so they can produce it in volume before 2025? They may have reasons, but if volume production isn't reached by 2026, this is not up to Tesla's speed I got used to back in the early days of the company.

It’s not just factory build time. It’s production line build time. While they may be able to overlap those two a bit at the ends, they’re mostly separate tasks one after another. And they can’t start die build until there are 100% with the physical aspect of the vehicle. Die build is a year long process - give or take depending on several factors.

Tesla has a lot more on their plate today than early Tesla. Manpower and resources are divided amongst a lot more projects. Even earlier Tesla pulled employees and other resources from projects to ie., ‘get Model 3 ramped’. I’d confidently bet that that still happens. That projects get put on hold because something else has become more important in the moment. WAG, but not really - FSD just got put to the head of the line.

 

[SOULPEDL]

All this with only a half-pack of batteries? Whoa!
Tuesdays tend to be better than Monday's action. Imagination?

 

[Krugerrand]

An example of the Stab and Twist method of journalism.

Take a stab at using some facts, then twist the information to support a narrative that isn't actually supported by the facts.

But it’s not lying.

 

[Musskiah]

More complete list of all cities listed with openings here-View attachment 1036939

That's a whole lotta nothing!

 

[scaesare]

Context? When and where was this photo taken?

It was after the Isaacson book, in an interview with Franz... I think there were a couple of news outlets that carried the story at the time but here's the one that came up in a quick search when I was looking for the pic:

First Glimpse of Tesla's Mythical $25,000 Car? [Video]

Get a sneak peek at Tesla's $25,000 vehicle through Walter Isaacson's lens. Discover what lies ahead in the automotive landscape.

www.notateslaapp.com

Note: The article calls it a 3-wheeler, and we debated that a bit here, and the general consensus was that was wrong, and if you look closely the rear wheel is in line with the front, so really a 4-wheeler...

Apple's famous 2007 unveiling presentation for the original iPhone was like this. Steve Jobs' demo of the prototype phone was carefully scripted to ensure that the bugs didn't cause a failure.

https://www.nytimes.com/2013/10/06/magazine/and-then-steve-said-let-there-be-an-iphone.html?smid=tw-share&_r=0&pagewanted=all

During the Microsoft presentation of Windows 98, it crashed with the Blue Screen of Death during the live demo.

Boeing rolled out the 787 Dreamliner in 2007 to great fanfare, but was hiding the fact that it was a basically a half-built shell of an aircraft with no systems installed and some of the structure held together by fasteners bought from Home Depot.

The ‘fake’ Boeing 787 rollout

As Boeing rolled out its ‘Dreamliner’ to experts and journalists, under the surface not all was as it seemed.

www.aljazeera.com

All of these products ended up being extremely popular with customers in the long run.

Indeed. Lots of concept cars, electronic gadgets, software products, are mock-up that either don't work and just have pretty lights, or only have a subset work... like your iPhone example. Nobody things that what's unveiled on stage buy not likely to ship for months or years later is complete and fully functional.

Heck, I remember the Model 3 at the gigafactory prior to it's release, somebody snapped a pic of the side mirror starting to droop like it wanted to fall off. It turns out that the car was a clay model over a wooden frame with some sort of metallic skin on it...lol. But Tesla never claimed otherwise... and nobody reasonable expected a fully completed car at that point...

 

[UkNorthampton]

Given this and the fact Tesla saw a YoY decline in auto sales while the majority of the ICE auto industry saw YoY increases, I think one would have to conclude EVs are experiencing some demand softening. Clearly this isn’t just a Tesla thing; BYD posted poor numbers as well. However, since legacy auto has posted a string of solid quarters, it doesn’t appear to be channel stuffing.

So…when in doubt, zoom out. 2023 was a really good year for EVs; growing something like 47% over 2022 numbers. That kind of growth is tough to maintain, especially when the other brands are having issues, our President won’t recognize Tesla, and a popular presidential candidate consistently talks nonsense about electric cars. I’m confident eventually this negative perception will give way to the reality that EVs, particularly Tesla EVs, are just better vehicles, but at this point I’m just hoping the Model Y will hold onto the crown of being the world’s best selling car again in 2024.

So far I’ve been super impressed with FSD 12.3.3, and while I can’t say it’s passed the “wife” test yet, it certainly has impressed her. This and the 8/8 reveal of the RoboTaxi are excellent news! I don’t concern myself with the manual drive version of this car. This “manual drive” option will make the car more complicated and more expensive than the basic RT, so people will pay more to drive the car themselves. For this reason it makes sense the basic RT will be first to market, but it’s yet to be seen if Tesla actually sells them.

Anyway, it’s never a dull moment with Tesla. While I wish I was smarter and could have times the ups and downs in TSLA over the last few years to accumulate more shares, I’m content to just sit with what I have and watch all this unfold. That contentment is rooted in the fact that Tesla has the best and brightest in their field, and they have a track record of delivering the impossible…late.

Apologies if others have made my point, but I'm still wading through many pages.

USA ICE cars "sales" are best described as "Channel Stuffing" (via dealers - whose stock has ballooned to unsustainable levels). Even USA Toyotas & Hondas are having buildups in many models. Tesla's "inventory" (including cars on ships, rail & road transport) is way below other OEM's inventories in dealers.

What Is Channel Stuffing? How It Works, Purpose, and Legality

Channel stuffing is an unethical method of deceptively inflating sales figures immediately prior to a reporting period by forcing an oversupply of product onto retail and distribution channels.

www.investopedia.com

 

[AndrewZ]

That's a whole lotta nothing!

Tesla has had similar job postings for years

 

[unk45]

If they were stuffing the dealers inventory would go up or doesn't that count in that statistic ? Inventory/sales would only drop if sales go up ?

That depends on the source. Some report only dealer inventories, some only OEM, some both. Most OEMS do aggregate both for some reporting, but there are numerous 'workarounds'. Nearly all OEMs report sales when vehicle is assigned to a dealer (that means the vehicles may be in transit to a dealer). Those factors are why the Tesla inventory number does NOT survive comparison with the other OEMs. Tesla records a sale only on transfer of a vehicle and payment in full (leases are sales if sold to an external lessor, are recorded in the life of the contract if Tesla owns the lease.

Dealer 'stuffing' usually relates to slow selling models. Normally in cases of such 'stuffing' there are dealer finance incentives, various disclosed and/or non-disclosed dealer sales incentives. In common cases 'hot' models have selling prices above MSRP and are not subject to incentives.

For North America the best regular source is Automotive News, although dealer incentives disclosures vary tremendously. Automotive News usually finds them.

 

[UkNorthampton]

Honestly what's the benefit to a government. From the govt POV

You kill your national car manufacturers. HORRIBLE
You kill road tax, fuel tax, registration, etc etc...HORRIBLE
Cause congestion in urban areas...HORRIBLE
Kill taxi driver employement/gig workers etc....HORRIBLE

Only positives is a slight decrease in mortality & injury.

I actually don't see any benefit if I am the govt of Indonesia or Thailand or India or Korea or Germany etc etc.

Energy Subsidies – Topics - IEA

Explore analysis, reports, news and events about Energy Subsidies

www.iea.org

 

[scaesare]

I was by no means making an argument in Reuters's favor. I was more commenting on how journalists and their publishers keep getting away with falsehoods about Tesla, as long as they claim there are "sources." ...."We were only reporting what our precious sources told us...and no, you don't get to know who the sources are."

My use of the word "technically" was a half-effort reference to the old "technically correct" meme from Futurama, and I didn't mean it as a compliment or defense of Reuters:

View attachment 1036828

Gotcha... yup, I agree with that sentiment. Chances are Reuters is smart enough to technically claim they aren't lying, even when they are being deliberately misleading...


Sam: "Ralph, send me a report on Tesla dropping excess inventory in a volcano..."

Ralph tosses 1/2 page report over the cubicle wall 5 mins later.

Sam (writing): "According to a report seen by Reuters..."


So yeah... we are on same page.

 

[unk45]

Apologies if others have made my point, but I'm still wading through many pages.

USA ICE cars "sales" are best described as "Channel Stuffing" (via dealers - whose stock has ballooned to unsustainable levels). Even USA Toyotas & Hondas are having buildups in many models. Tesla's "inventory" (including cars on ships, rail & road transport) is way below other OEM's inventories in dealers.

What Is Channel Stuffing? How It Works, Purpose, and Legality

Channel stuffing is an unethical method of deceptively inflating sales figures immediately prior to a reporting period by forcing an oversupply of product onto retail and distribution channels.

www.investopedia.com

You're making the excellent point that bears repeating regularly. New people who come to this Forum and/or to TSLA rarely understand the vast different in Accounting and distribution between Tesla and virtually all others. Tesla not only has much less inventory, but it also includes categories others deem sold, so the differences are even more stark.

 

[cliff harris]

Good news on cybertruck charge rates:

https://twitter.com/x/status/1777689351327219726

from Drew Baglino:

Cybertruck charge curve improvements are coming OTA later this quarter to unlock up to 154 miles recovered in 15 minutes.

Which is awesome, as the one genuine complaint I kept seeing about the CT was the charge curve being a disappointment. Its not like Tesla was having trouble finding CT buyers (even pure foundation series ones) anyway, but ever bit helps.
Its funny how all the awesome stuff Tesla does gets forgoteen because we always chase the latest news. Robotaxi reveal date has everyone so excited we have forgotten that the cybertruck ramp is ongoing. Remember when all we had to be excited about was a rumored model 3 highland .
Lets not even mention that its time for a bot update.

 

[FEERSUMENDJIN]

Still, the opportunity to retrofit existing models should also offer cost-effective expansion.
The more Robotaxi types available the better the odds of scale efficiency.
The first ones will ideally be a small car and a van, as many observe. The form factors are an interesting question.
How about replicating compact exterior with expansive interior such as these:

LEVC | Across The UK | London Electric Vehicle Company

Visit London Electric Vehicle Company in Across The UK, established dealership and servicing specialist. Explore the website and browse through our full range of offers.

www.levc.com

With Tesla design addressing these interior capabilities the prices will be much lower and market adaptability probably higher. It would be surprising were Tesla not to have considered these models while designing robotaxi.

FWIW, Waze and Google Maps differ by origin (Waze started in Israel) but have modest differences in practice:

Waze Data Feed specifications - Waze Partners Help

Revision History The following revision table presents the changes made for each version of this document. Revision Date

support.google.com

Combining and Visualizing Multiple Data Sources | Maps JavaScript API | Google for Developers

developers.google.com

They are NOT identical, although why, since they are in the same company, is odd.
In actual routing, the primary issue here Waze seems to have slightly more rapid routing data while Google Maps, in some places, tends to have slightly faster traffic updates.
Waze also has pride of place in many corporate-level routing applications.

Tesla is perfectly fine in areas with sufficiently dense Tesla concentrations but less so as Tesla density decreases.

This perspective is not mine but comes from a friend who has worked on proprietary mapping solutions for vehicle fleets for some years. He also has used and tested numerous other proprietary mapping software and hardware choices, but that is off-topic here.

As we consider the prospects for Tesla Robotaxi and the 'last mile/km problem' my friend informs me that such problems are routinely solved for local delivery and service vehicles almost globally. That, according to him, includes generally unmapped but documented places mentioned earlier in one of my posts. His claim is that for things such as Robotic delivery each location is explicitly mapped and is updated with each use of the location, giving nearly 100% coverage of actual usage patterns. Hence, for hospital supply, for example, each specific delivery point in a large medical complex is individually mapped with access rotes and operating times/access data.

After looking at that and thinking, it seems to me that since Robotaxi will always tend to operate in specific geographic areas, it is with 'present data availability' to precisely map nearly all pickup/dropoff points. The only constraint is that such data tends to reside with express delivery companies and, in many cities, local taxi/rideshare data that is guarded diligently.

it seems to me that the FSD advances are crucial, but the localized data bases are almost equally important.

Localised pickup/drop off location data bases may be the magic sauce for a ride share aquisition by Tesla ? Nah. I would think this information could be provided to the training model using retrospective/prospective data from the fleet pretty quickly, especially as many Teslas are already being used in rideshare.

 

[GhostSkater]

All hail the 4680s

Average charging power of 236 kW in 15 minutes, right now it maintains 250 kW up to 29%

All the following are rough guesses since I don't have MATLAB with my charge curve code right now, but assuming that is from 10%, 154 miles of range added on Cybertruck AWD A/T is 48% added, meaning 10-58% in 15 minutes, and it should maintain peak power up to 40%, maybe more

On (true) V4 Superchargers it should be able to maintain 350 kW (if that is the V4 power) up to 25%~30%

In pool blue the current charging curve, purple a possible future curve Drew is talking about and olive green that extrapolated to V4

Will be nice to see if there is a improvement on the upper end also, because it's not good right now, and all this will make it so much better long trips and towing wise compared to the competition

As a reminder, blue is the curve needed at 350 kW to reach the 15-85% in 18 minutes Lars talked about, and green that curve limited to a V3 Supercharger

https://twitter.com/x/status/1777689351327219726

 

[MartinAustin]

Good news on cybertruck charge rates:

https://twitter.com/x/status/1777689351327219726

THIS is why I'm happy my Cybertruck reservation number is like #1,300,000

Meanwhile TSLA is singing "knock, knock, knockin' on $180's door... hey... hey... hey"

 

[jschwefel]

Whole heartedly agree with most of your points. Just want to add one thing that people may underestimate how long the traditional OEM's pipeline is. The fed actually keeps track of a statistics that might reflect, the auto inventory/sales ratio, while it's recovered some from the sharp drop in 2021, it's still way below pre-pandemic level. OEMs are still stuffing the dealers.

Auto Inventory/Sales Ratio

Graph and download economic data for Auto Inventory/Sales Ratio (AISRSA) from Jan 1993 to Feb 2024 about ratio, inventories, vehicles, new, sales, and USA.

fred.stlouisfed.org

This was super helpful. Thank you!

 

[scaesare]

It's like you don't understand how a conversation works. Think of a party

Person A: I like food, this party has OK food.
Person B: I make food, I can make better food.
Person C: Person A didn't say anything about making food you fool.

Don't be person C.

I did explain that I understood the point the person I riffed off of was making. I also veered away from it because I thought it was a bad idea, with a kernel of a good concept that could be expanded on.

It wasn't as much as taking issue with your premise of "reducing weight by using less massive components when not needed for the application", which I agree with and obviously makes sense. Don't need an off-road bumper either.

My point was with the original poster's claim that in that application you could get rid of them. No harm intended.

 

[hobbes]

At worst, how expensive would it be to hire a person to manage plugging in at existing Superchargers or dedicated robo-taxi yards?

I figure a single person could manage the daily charging needs of 100+ robotaxis over an 8 hour shift if there are enough chargers.

A while ago Tesla acquired the small German company Wiferion and kept the part of it that is about inductive charging. Then there was a picture in some Tesla presentation that showed an inductive charger in a garage. Wouldn´t that be a good match with robotaxi?!

You searched for inductive - Drive Tesla

driveteslacanada.ca

 

[hobbes]

Whoever came up with such a silly idea as a two seater?
For certain it was nobody at Tesla!

Looks like Tesla did think about a two seater at some point, this picture was in the Isaacson biography:

Tesla robotaxi to be unveiled August 8, 2024

Tesla CEO Elon Musk has announced the company will unveil its robotaxi design, based on their next-generation vehicle platform, on August 8, 2024. Earlier today we told you about a report from Reuters that said […]

driveteslacanada.ca

 

[unk45]

Localised pickup/drop off location data bases may be the magic sauce for a ride share aquisition by Tesla ? Nah. I would think this information could be provided to the training model using retrospective/prospective data from the fleet pretty quickly, especially as many Teslas are already being used in rideshare.

Most people who've used any rideshare or use any present auto nav in large complexes knows that they cannot solve septic sub-complex issues. There is an industry built around specialized solutions for urban delivery and service companies that customize data to specific needs. Zero chance that the training process can provide ability to navigate things that have no clearly found directions. People cannot do many of these things without asking directions. Can a Robotaxi ask a passing local where a given sub-building entry is?