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

[KBF]

This sums up the stupidity of modern western NIMBYism right here:

https://twitter.com/x/status/1575868620118188033

God forbid a barren field of mud have zero emission solar panels placed on it!

On my road trip I passed some awesome solar installations in either Saskatchewan or Alberta with sheep grazing underneath. Very efficient use of space, and it's a mutually beneficial relationship (better grass, extra shade for the sheep, sheep keep the weeds down for easier solar access, etc).

Solar is a tech that can actually enhance existing agriculture (and I agree - monoculture muddy fields are not beautiful... Or natural. This from a Prairie boy... Tilled soil is terrible for ecosystems - healthy soil requires diverse top growth to maintain vital microorganisms - one reason our food is far less nutritious than it used to be...)

To connect this to TSLA, with billions in cash, why isn't it being used to add solar roofs to Superchargers yet? Why else merge with Solar city? I guess red tape is involved - in Canada most/(all?) are built on parking lots owned by separate entities, and energy is provincial and Tesla still isn't allowed to sell electricity (only time while plugged in) - it's utterly ridiculous. But man, I'd sure love a few billion to be put to use in more infrastructure (and more service centres).

As much as a decentralized/flatter business structure is great for innovation, I feel like it hurts their ability to focus on service/maintenance areas.

 

[ZeApelido]

Making something understandable means it's good. There's a ton of "influencers" who pedal complete *sugar* or exaggerate the hell out of things.

I simply cannot believe this opinion is coming from the person who pimped Celsius over and over in an epic fashion. RIP Celsius

If y'all haven't heard about this yet, I just wanted to introduce you to the greatest free money printer I've seen in my life.

It's called Celsius.

If you store Bitcoin on Celsius you get 6% compound interest!

Screw crypto though, if you store CASH on Celsius you earn 10.5% compound interest!!! What!?!?!


~~~IN-POST MODE EDIT:

This post has been sent to Moderators’ attention for “Sounds scammy — ToS "advertisements, chain letters, pyramid schemes, and solicitations"? Although I think the user's probably well-intentioned”
Without further comment on its merits, weaknesses or anything else, a suggestion that ALL do all conceivable due diligence whenever there is a free lunch being offered.~~~

Lol Celsius sounds too good that it was labeled as scammy! Hah! That's crypto disruption for ya. I'll take pride in that mod edit.

I've gotten so many of my co-workers on Celsius. Seriously, 10.5% interest rate on cash like it's a total no-brainer.

But why wouldn't you hold it on Celsius instead? If you buy/hold crypto on Celsius you earn 5-6% interest. That's insane.

I actually think that these interest programs are LESS risky than exchanges.

 

[Prunesquallor]

This sums up the stupidity of modern western NIMBYism right here:

https://twitter.com/x/status/1575868620118188033

God forbid a barren field of mud have zero emission solar panels placed on it!

Those communities should then be blacklisted from receiving cheap solar/battery power for the indefinite future.

 

[Ipcress]

OK, here's more wild speculation on my part. This is very unlikely, but let's do a thought experiment - Optimus Pre-orders.

It's been a long time waiting for Cybertruck but it didn't matter much, we're still here. Same with Semi... how long ago did Pepsi open contracts on those units?

Who would not want one in their business ASAP?
Who would not drop $20K today for a spot in line?
Who would benefit most from Optimus data in the wild (not just Tesla factories)?
What would drive Dojo Commercial expansion faster than any ramp we've seen?

I give it about a 1% chance so soon because they've made it clear it's for the Factories initially. But someday, pre-orders in the 10s of Millions, easily. Having a Smart Robot could be required for many businesses within 10 yrs or sooner. This will be a race if/when they give the green light.

But what if Mr Evil wants one? Does Tesla sell to just anyone, or can they discriminate here?

That Dr. Evil, if you please.

 

[Gigapress]

Well yeah after the steep part of the S curve, growth slows down dramatically until phase 2 is introduced to Berlin and Texas. Phase 1 is 500k annualized production capacity per factory. It can take up to 2 years to maximize production capacity. Shanghai is just done with phase 2 now can hit 1M run rate, so lets expect Berlin/Texas to go slower.

Berlin's nominal Phase 1 capacity target is 10k Ys per week, but I don't believe that because Tesla has been understating their true factory capacity for Shanghai for years and is doing it again with the ">250,000" capacities listed on the Q2 earnings deck.

We also had the sign at the body-in-white line at the opening party saying that the takt time is 45 seconds, which implies 700k per year theoretical maximum capacity with 24/7 uptime, and probably more like >600k after accounting for downtime. This 45-second cycle was also recently confirmed by a UBS analyst report in August. We also don't know if Tesla is sandbagging with that claim of 45 seconds per car.

So if Texas/Berlin can hit 5k/week which is guided according to this slide, and hit 7k/week exiting Q4 2023 which looks to be guided according to that chart. This is a pretty fast ramp.

It took Fremont 1.5 years to go from 5k/week to 7k/week of model 3s btw.

That was with 2018-era Model 3 Production Hell, plus all the inefficiencies of the retrofitted Fremont facility, and that was happening while the manufacturing development focus was mainly directed to Model Y and building Gigafactory Shanghai. The 2018 Model 3 ramp was extremely difficult and riddled with mistakes, lost sleep and bad decisions.

Shanghai, a closer comparison, hit 5k per week in October 2020 and then reached 7k per week just 5 months later in March 2021, and after another 5 months hit 10k per week in August. So that's 10 months, or about 3 quarters, to go from 5k per week to 10k per week in Shanghai historically. This was also with the challenges of the new Model Y line going live at Giga Shanghai in December 2020.

For Berlin and Texas, we are talking about Model Y with the new and improved production methods, the latest factory operating software, and a lot more experience for the team, which should make it majorly easier to build and ramp than the 2018 Model 3 was. The big question is 4680 ramp, really.

If B&T merely ramp at the same speed Shanghai did, then if 5k per week is achieved by around December 2022, the 10-month lag would make 10k per week arrive in October 2023. This is a substantially faster ramp than the Reuters report suggests, as they have Tesla targeting 100k for Q3 '23 which is only 7.7k per week. Thus, Reuters is indirectly asserting that Tesla is internally targeting a slower ramp for B&T than the ramp for Shanghai was at the same stage. That's total nonsense and suggests that they didn't do their homework or think about this before publishing, or worse that they may be doing this for malicious reasons. Tesla explicitly said on the Q1 2022 earnings call that B&T should ramp substantially faster than Shanghai did and explained why that's the expectation.

Martin Viecha: (28:58)
Thank you. At what rate do you expect Berlin and Austin to ramp relative to Shanghai? Are you able to leverage learnings from Shanghai or are the processes substantially different in the new factories?

Elon Musk: (29:10)
Ramp production are faster than Shanghai because we have learned a lot and we have basically veteran teams that have seen the 3/Y ramp, the Y ramp especially, in multiple locations and we’re obviously sharing what we’ve learned. And so we don’t want to get complacent or entitled, but this should be a faster ramp because we have learned more and we have done a lot to simplify the production process of Model Y. That should lead us to a faster ramp in Texas and Berlin.

Speaker 1: (29:54)
We also have, because it’s structural and casting, about 30% less robots. We expect to almost double the capacity for body, for example, reducing the number of robots, but doubling our capacity in a lot of areas.

Elon Musk: (30:07)
Yeah, right. The body line for the structural pack, and if you got a structural pack, and front and rear castings, the body shop size drops by over 60% relative to the standard way of making a car.

Lars: (30:33)
That tacks into general assembly and everything else because we have the structural battery, the floor is the battery. We put the seats on the battery and then we put that in their car, so it’s actually between 10 and 15% less stations in GA because of the general assembly start as well. So really I think about this in the way that we think about cars. If you’re waiting for the best Tesla, you’re going to be waiting forever. If you’re waiting for our best factory, you’re also going to be waiting forever, because every new factory is better than the last one because we take all that learnings and we throw it into the new one.

Elon Musk: (31:03)
Yeah.

On the Q2 call, Elon reiterated that he's confident B&T will do 10k per week by the end of the year, but the supposed plan from Reuters doesn't seem to indicate there being the kind of safety margin in the plan necessary to make such a statement.

Elon Musk, Q2 Earnings Call: "And so that's why I'm confident we'll get to 5,000 cars a week at -- in Austin and Berlin by the end of this year or early next year and probably but not certainly, 10,000 cars a week at both locations by the end of next year."

 

[aubreymcfato]

I simply cannot believe this opinion is coming from the person who pimped Celsius over and over in an epic fashion. RIP Celsius

Ok, HERE is WHERE I heard about Celsius for the first time! I couldn't remember it, now I can blame Xepa777 for my 60 bucks lost...

 

[willow_hiller]

Who is this and what has he done with Elon?

https://twitter.com/x/status/1575970770584379392

 

[Singuy]

Berlin's nominal Phase 1 capacity target is 10k Ys per week, but I don't believe that because Tesla has been understating their true factory capacity for Shanghai for years and is doing it again with the ">250,000" capacities listed on the Q2 earnings deck.

We also had the sign at the body-in-white line at the opening party saying that the takt time is 45 seconds, which implies 700k per year theoretical maximum capacity with 24/7 uptime, and probably more like >600k after accounting for downtime. This 45-second cycle was also recently confirmed by a UBS analyst report in August. We also don't know if Tesla is sandbagging with that claim of 45 seconds per car.


View attachment 858609



That was with 2018-era Model 3 Production Hell, plus all the inefficiencies of the retrofitted Fremont facility, and that was happening while the manufacturing development focus was mainly directed to Model Y and building Gigafactory Shanghai. The 2018 Model 3 ramp was extremely difficult and riddled with mistakes, lost sleep and bad decisions.

Shanghai, a closer comparison, hit 5k per week in October 2020 and then reached 7k per week just 5 months later in March 2021, and after another 5 months hit 10k per week in August. So that's 10 months, or about 3 quarters, to go from 5k per week to 10k per week in Shanghai historically. This was also with the challenges of the new Model Y line going live at Giga Shanghai in December 2020.

For Berlin and Texas, we are talking about Model Y with the new and improved production methods, the latest factory operating software, and a lot more experience for the team, which should make it majorly easier to build and ramp than the 2018 Model 3 was. The big question is 4680 ramp, really.

If B&T merely ramp at the same speed Shanghai did, then if 5k per week is achieved by around December 2022, the 10-month lag would make 10k per week arrive in October 2023. This is a substantially faster ramp than the Reuters report suggests, as they have Tesla targeting 100k for Q3 '23 which is only 7.7k per week. Thus, Reuters is indirectly asserting that Tesla is internally targeting a slower ramp for B&T than the ramp for Shanghai was at the same stage. That's total nonsense and suggests that they didn't do their homework or think about this before publishing, or worse that they may be doing this for malicious reasons. Tesla explicitly said on the Q1 2022 earnings call that B&T should ramp substantially faster than Shanghai did and explained why that's the expectation.



On the Q2 call, Elon reiterated that he's confident B&T will do 10k per week by the end of the year, but the supposed plan from Reuters doesn't seem to indicate there being the kind of safety margin in the plan necessary to make such a statement.

I honestly don't think we can compare anything to Shanghai. Production ramp correlates with working hours. China is full of people ready to work long hours, Texas and Berlin not so much. It's a tougher time to get the crew assembled elsewhere vs China which is an abundance of people. They pump out 11M college grads a year with a near 20% unemployment rate.

 

[WattsHappening]

https://twitter.com/x/status/1575884161440776195

Wait times back to 4-8 weeks in China.

 

[WolfHero]

OT, first 10 min of next week's CNBC show Chasing Tesla - American Greed

 

[Gigapress]

I honestly don't think we can compare anything to Shanghai. Production ramp correlates with working hours. China is full of people ready to work long hours, Texas and Berlin not so much. It's a tougher time to get the crew assembled elsewhere vs China which is an abundance of people. They pump out 11M college grads a year with a near 20% unemployment rate.

Berlin and Texas are fundamentally more efficient factories than 2020 Shanghai.

1. They don't have to deal with the hassles of Model 3, which is inferior to Model Y in terms of manufacturability. Shanghai in the Oct 2020-Aug 2021 period was mainly producing Model 3s, which handicapped their ramp compared to the much faster and smoother Model Y ramp which was mainly inhibited by supply chain issues and COVID lockdowns instead of problems within Tesla.
2. The new 4680 battery assembly, front castings, structural battery and other improvements drastically reduce the amount of labor required in the first place.
3. All internal software development that has been implemented since 2020 can be downloaded as-is at the new factories, so they have better digital infrastructure than Shanghai had. Likewise, Berlin and Texas are making the latest and greatest version of Model Y that's at the cutting edge of Tesla's relentless design-for-manufacturability improvements. We can observe this progress in the Austin Model Y teardown on Munro Live.

At the end of the day, earlier this year Elon, Drew and Lars said in response to a direct question about this that Berlin and Texas should ramp materially faster than Shanghai did, so I think that's what's most likely to happen.

 

[printf42]

China order queue returned to 4-8 weeks.

设计您的 Model 3 | Tesla

Model 3 现已开放选配。了解 Model 3 配置和价格，下单订购您的 Model 3。

www.tesla.cn

I think the reason for order rate drop last month were:

• Rumors of “price cut imminent”, from articles like this. In the article they even included Tesla spokesperson debunking the rumor, yet they still run with the title.
• Elon himself claims raw material prices dropping and at sometime prices will need to drop from “embarrassingly high” to “comfortably high”, this kills orders from professional “car flippers”, which probably is the reason Elon said that in the first place.
• Many customers might be expecting HW4 to start being installed in new cars right after AI day, and would wait for clarity, remember China FSD take rates are only single digit so they don’t get free upgrades if take delivery before the switch, and Chinese hate to miss out on free stuff.(before I got roasted, I am Chinese)

All these will become clear over this weekend either ways. So I expect order rate to return to normal pretty quickly.

 

[Xepa777]

I simply cannot believe this opinion is coming from the person who pimped Celsius over and over in an epic fashion. RIP Celsius

1) I’m not trying to be an authoritative influencer personally benefitting my brand from my public contributions. I’m just one person who has some right opinions and many wrong opinions. If I opened up your investment (or life) portfolio there would be plenty of embarrassing things to look at too.

2) My personal opinions have nothing to do with what Douma does. Going to this level of character assassination personal attack is as immature as politics as saying “wow Obama smoked weed in the past let’s discredit his renewable energy platform.”

3) You can choose to listen to him or not whatever it’s a free world. I personally have FSD on my Tesla so I’m not anti-FSD. Rob has had bears on his channel and Gali has gone debates. I know for a fact Dave and Duoma have turned down FSD debates from people who actually have ML jobs. They have the right to do that, and I have the right to call out propaganda bullshit when I see it. And you have the right to throw personal attacks at me when you read that. God bless the internet.

4) I don’t have a bookmarked list of everything he’s said that’s been wrong (that’s obsessive) I just know from my limited background and ml circles I’m in that a lot of the stuff he said in the past has been technically flawed/controversial in regards to architecture and capability progression. And when I saw this example it was an easy reference.

I wish TMC had that “remind me in 2 years” feature that Reddit does. It comes in quite handy, at least for basketball debates I’ve been in and taking multiple years for opinions to play out

 

[Big Time]

OT, first 10 min of next week's CNBC show Chasing Tesla - American Greed

T-MILT Lol

Put him in a room of 30 PhDs

 

[Singuy]

Berlin and Texas are fundamentally more efficient factories than 2020 Shanghai.

1. They don't have to deal with the hassles of Model 3, which is inferior to Model Y in terms of manufacturability. Shanghai in the Oct 2020-Aug 2021 period was mainly producing Model 3s, which handicapped their ramp compared to the much faster and smoother Model Y ramp which was mainly inhibited by supply chain issues and COVID lockdowns instead of problems within Tesla.
2. The new 4680 battery assembly, front castings, structural battery and other improvements drastically reduce the amount of labor required in the first place.
3. All internal software development that has been implemented since 2020 can be downloaded as-is at the new factories, so they have better digital infrastructure than Shanghai had. Likewise, Berlin and Texas are making the latest and greatest version of Model Y that's at the cutting edge of Tesla's relentless design-for-manufacturability improvements. We can observe this progress in the Austin Model Y teardown on Munro Live.

At the end of the day, earlier this year Elon, Drew and Lars said in response to a direct question about this that Berlin and Texas should ramp materially faster than Shanghai did, so I think that's what's most likely to happen.

Well maybe, it took a year from delivery to 5k/week for Shanghai. However this was during Covid it's probably not apples to apples. Tesla is guiding for 5k/week Q4 of 2022 which is ~4 months after first deliveries?

 

[TheTalkingMule]

T-MILT Lol

Put him in a room of 30 PhDs

I'm 47% sure I had a long LinkedIn argument with Trevor way back in the day. I gotta figure out how to search my activity log and frame that.

What a doofus.

 

[Captkerosene]

Those communities should then be blacklisted from receiving cheap solar/battery power for the indefinite future.

Kid: But I don't want to eat my broccoli.

Parent: Fine, your punishment is no delicious broccoli for ten years.

 

[insaneoctane]

same here .. submitted this a while ago I think there was an email from Telsa IR then 0 feedback

i am guessing they are only accepting Say - linking shares

are others comfortable with this ? started the process with Say and the want login credentials for Brokerage/IRA/401 K accounts... seems risky to me ... but what do i know ....
View attachment 858572

Yep. Asked for my login credentials...that's a hard NO. I'm sure there are many far more tech savvy that would tell me that's safe, but there's zero chance I'm providing login/password. Sorry.

 

[Discoducky]

Ok, ok, I'm a bit excited!

 

[BitJam]

Would like to point out that James Douma was confidently and almost condescendingly dead wrong when he said this the original picture of the hands was just a pretty render and had nothing to do with the robot.

Douma is nowhere near some like sandy Munroe and the only reason he has relevance is because Dave gave him a platform to speak.

IMO James provides many useful insights independent of the platform/creator. OTOH, he is human like the rest of us and makes mistakes. IMO his claim that FSD improvement is exponential because Tesla cites many improvements in percent was a stretch. In an interview with Tesla Owners of Silicon Valley, Elon said FSD has been improving logarithmically in a piecewise fashion (which is the diametric opposite of exponential growth). Unfortunately, these tenuous claims of exponentially fast FSD development reverberate in an undamped internet echo chamber.

In the software world we say that the last 10% of a project takes 90% of the work. This is usually because it's not clear exactly what the software needs to do until a lot of it has already been built. At the opposite extreme is AlphaGo that crossed a threshold so it can master well defined games, far better than humans, at breathtaking speed.

IMO FSD development is more like normal software development than it is like AlphaGo mastering well defined games. This is because the FSD problem is still not well defined in a software algorithmic way. The more you solve, the more you learn about new things that need to be solved or new approaches that you need to take.