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

[traxila]

A decent article for a change.

Titans of car-making are plotting the overthrow of Elon Musk

The world’s two biggest automakers — each sold roughly 10 or 11 cars for every one Elon Musk did last year

www.livemint.com

 

[heltok]

And to those wondering how interest rate hikes affects Tesla even though Tesla has very low outstanding debt and are not looking to borrow lots of money in the future. This has to do with a shrinking of money supply. As rate goes up, those who are buying Teslas will see higher monthly payments because they are the ones borrowing money. This can results in less demand as all their other potential debt payments (like mortgage interest/CC interest) can go up, especially for those who are not on fixed interest rates but variable rates.

There are many effects going on here. If you wonder how many eggs/milk/Model Ys you will be able to buy for one TSLA stock in the future given different interest rates from the FED the direct effect is that it’s unchanged.

OT:
There are secondary effects but they are really complex. I recommend Ray Dalios latest book if you want to go down that rabbit hole. Basically when you borrow to spend you will have to cut down spendings in the future to pay it back. When you have a reserve currency other people will do that for you until they won’t, which can encourage you to spend a lot while you can. Cutting down on spending is not politically popular and can lead to tumultuous times. Some people believe that MMT fixes the problem, Elon disagrees.

 

[ZachF]

There are many effects going on here. If you wonder how many eggs/milk/Model Ys you will be able to buy for one TSLA stock in the future given different interest rates from the FED the direct effect is that it’s unchanged.

OT:
There are secondary effects but they are really complex. I recommend Ray Dalios latest book if you want to go down that rabbit hole. Basically when you borrow to spend you will have to cut down spendings in the future to pay it back. When you have a reserve currency other people will do that for you until they won’t, which can encourage you to spend a lot while you can. Cutting down on spending is not politically popular and can lead to tumultuous times. Some people believe that MMT fixes the problem, Elon disagrees.

Being the reserve currency is actually a net negative for the US economy.

It requires us mathematically to run a trade deficit and shrink the size of the tradable goods sector, because net trade must always equal zero… the more dollars we supply to the rest of the world, the deeper the deficit must be.

Check out the book “Trade Wars are Class Wars” or “The Great Rebalancing” by Michael Pettis. Here is a primer:

Are We Starting to See Why It’s Really the Exorbitant “Burden”

carnegieendowment.org

So much of modern economics is garbage dogmatism and goal-seeking… Pettis is probably the person closest to looking at trade from a first principles stance.

 

[Dig deeper]

I'll be pleasantly surprised if Tesla delivers more than 1 car from Giga Berlin in the first half of this year. It's at the point where the red tape is like a Monty Python sketch.

https://twitter.com/x/status/1478852971689611266

Gives Tesla time to debug/troubleshoot the production lines. They will hit the ground running.

They are placing sticks in front of a tsunami.

 

[Phobi]

https://twitter.com/x/status/1478864302799048704

In Elon we trust.

 

[Krugerrand]

https://twitter.com/x/status/1478864302799048704

In Elon we trust.

It’s all about the attitude, huh?

 

[Dig deeper]

You’ll eat those words in time. It was a brilliant choice and it’s going to pay off big time. Patience.

Its going to be a vacuum for European engineering talent.

 

[Bladerskb]

Yeah, Tesla should be worried:
The goal with EyeQ Ultra is to provide a package to be introduced into robo-taxis and regular consumer vehicles in the 2025 timeframe, at a total cost of $5000-$10000 for a system-level power of around 100 W.

Anyone should be worried including Tesla if a company can achieve L4 on a consumer level car with a bill of material under $4k. Especially if they can prove that their system is statistically and independently significantly safer than a human driver. So yes, a real car, with a real L4 software coming in 2024 is better than Elon's repeated promises and guarantees of L4 "next year" every single year since 2015. Not only that but buying a car with a L4 option that cost $10k MRP is better than paying $10k for a L2 car you have to watch like a hawk and instantly take over to prevent crashes.

Don’t doubt that their system that might go into production in 2025 will beat the 2019 Tesla HW3. But it’s not like Tesla are intending to stand still. Having a high quality large dataset of varied and tricky situations is probably a lot more important than raw compute.

Its not a TOPS race, if it was Nvidia, Qualcomm, Huawei already won with Orin, Atlan, Snapdragon Ride, etc.
Lastly, here goes the so-called "data advantage" again. Remind me after 7 years of touting "billions of miles of data".
Where is the L4 or L5 system to show for it? The optimal fleet if you wanted raw data would be around ~10k cars which Mobileye almost has with Zeekr.
They don't need more data, they are getting 25 million km of map data a day from OEM cars and unlimited raw data from Zeekr cars.
Anything more and its simply overload and simply tossed out. Yet it seems people fail to understand this basic fact. They think the more cars Tesla sales the more data and "FSD" becomes. Completely void of any real life logic. If all you needed was cars on the road, self driving would already had been solved.

These guys are still stuck with traditional CV, he’s talking about “algorithm” to label pixels of road surface, another “algorithm” to generate 3D point clouds, then yet another “algorithm” to identify objects in there…

If these “algorithms” are neuro nets, their chips are clearly under powered. If not, good luck with that dead end. Past decade had already taught us where CV is heading, anyone disagrees need to win back Image Net competition first.

To me it sounds like they are targeting L4 in clearly marked highway only, and was not even attempting to solve generalized L5 FSD.

tl;dr: Not a competition to Tesla, not even close.

This is a classic example of having no software engineering knowledge with the only exposure to engineering being Elon regurgitating "neural nets, neural nets, neural nets, neural nets, neural nets". Its like new startups and companies now a days who use to buzzword "AI". "We use AI". When they wouldn't know what AI was if it was a nuclear bomb ramming down their bunker.

I wish you would even attempt to research just a lil bit. You would discover that Mobileye's ODD (if you even know what that means) is focused on urban dense traffic and the most difficult cities to drive in like Jerusalem, Tel-aviv, Germany, New York, Detroit, Paris, Japan, China, etc.

Places that are order of magnitude harder to drive in than the majority of cities FSD Beta is running around in. Yet FSD Beta as you all know still fails every acouple miles in city traffic. But yes no competition to FSD beta.

Bwhahaha ...

Elon said: If Radar & Vision disagree, who do you trust?
Mobileye: Well if ONE system says there is X, we act

Tesla so: well.. we removed radar because it caused too many false positives & phantom braking.

Mobileye will find that out eventually as well


Also: The calculation is ONLY correct if those two things are INDEPENDENT (also highlighted on the slides). But .. they capture the same scenery with the same things at the same time. And for Vision + Lidar even with the same physical method (Light) with all effects that corrolate with that

I guess Airlines are next to find out right? Next time you take a flight make sure to yell at the pilot that you won't let the plane take off until they ripped out all the redundant stupid and useless sensors and hardware. That they are unsafe and put the hundreds of passengers in danger! When they ask you why tell them elon told you so!

"Boeing has long embraced the power of redundancy to protect its jets and their passengers from a range of potential disruptions, from electrical faults to lightning strikes.​

The company typically uses two or even three separate components as fail-safes for crucial tasks to reduce the possibility of a disastrous failure. Its most advanced planes, for instance, have three flight computers that function independently, with each computer containing three different processors manufactured by different companies.​

So even some of the people who have worked on Boeing’s new 737 MAX airplane were baffled to learn that the company had designed an automated safety system that abandoned the principles of component redundancy, ultimately entrusting the automated decision-making to just one sensor — a type of sensor that was known to fail. Boeing’s rival, Airbus, has typically depended on three such sensors.​

“A single point of failure is an absolute no-no,” said one former Boeing engineer who worked on the MAX, who requested anonymity to speak frankly about the program in an interview with The Seattle Times. “That is just a huge system engineering oversight. To just have missed it, I can’t imagine how.”​

Boeing’s design made the flight crew the fail-safe backup to the safety system known as the Maneuvering Characteristics Augmentation System, or MCAS.​

A faulty reading from an angle-of-attack sensor (AOA) — used to assess whether the plane is angled up so much that it is at risk of stalling — is now suspected in the October crash of a 737 MAX in Indonesia, with data suggesting that MCAS pushed the aircraft’s nose toward Earth to avoid a stall that wasn’t happening. Investigators have said another crash in Ethiopia this month has parallels to the first.

Boeing has been working to rejigger its MAX software in recent months, and that includes a plan to have MCAS consider input from both of the plane’s angle-of-attack sensors, according to officials familiar with the new design. The MAX cockpit will now include a warning light that will illuminate when the two angle-of-attack sensors disagree. "​

Here is the latest:

I am not impressed. In the drive they did the car was not very consistently centered and to be frank(pun intendend) they showed very easy scenarios. And the way they talk is so lacking anything that would make talent want to join them compared to Tesla AI Day.

You are confusing VW's 1 forward camera ADAS using 2017's EyeQ4 with the current gen 360 cameras supervision.
Second of all, there are tons of video of supervision in areas that are orders of magnitude more harder to drive in than typical US cities.
secondly, maybe Amnon isn't in the game of repeatedly lying every year for the past 7 years that he guarantees he will have L5 "next year" and chanting "neural nets, neural nets, neural nets" without having a clue.

Amnon is an actual computer scientist, Ph.D from MIT and is still a professor of AI/ML at the top #1 Israel University.
Lastly what makes supervision any different from FSD Beta? Oh wait, its actually fully hands free door to door and works in china which is orders of magnitude harder to drive in than the US.

 

[Singuy]

Anyone should be worried including Tesla if a company can achieve L4 on a consumer level car with a bill of material under $4k. Especially if they can prove that their system is statistically and independently significantly safer than a human driver. So yes, a real car, with a real L4 software coming in 2024 is better than Elon's repeated promises and guarantees of L4 "next year" every single year since 2015. Not only that but buying a car with a L4 option that cost $10k MRP is better than paying $10k for a L2 car you have to watch like a hawk and instantly take over to prevent crashes.



Its not a TOPS race, if it was Nvidia, Qualcomm, Huawei already won with Orin, Atlan, Snapdragon Ride, etc.
Lastly, here goes the so-called "data advantage" again. Remind me after 7 years of touting "billions of miles of data".
Where is the L4 or L5 system to show for it? The optimal fleet if you wanted raw data would be around ~10k cars which Mobileye almost has with Zeekr.
They don't need more data, they are getting 25 million km of map data a day from OEM cars and unlimited raw data from Zeekr cars.
Anything more and its simply overload and simply tossed out. Yet it seems people fail to understand this basic fact. They think the more cars Tesla sales the more data and "FSD" becomes. Completely void of any real life logic. If all you needed was cars on the road, self driving would already had been solved.


This is a classic example of having no software engineering knowledge with the only exposure to engineering being Elon regurgitating "neural nets, neural nets, neural nets, neural nets, neural nets". Its like new startups and companies now a days who use to buzzword "AI". "We use AI". When they wouldn't know what AI was if it was a nuclear bomb ramming down their bunker.

I wish you would even attempt to research just a lil bit. You would discover that Mobileye's ODD (if you even know what that means) is focused on urban dense traffic and the most difficult cities to drive in like Jerusalem, Tel-aviv, Germany, New York, Detroit, Paris, Japan, China, etc.

Places that are order of magnitude harder to drive in than the majority of cities FSD Beta is running around in. Yet FSD Beta as you all know still fails every acouple miles in city traffic. But yes no competition to FSD beta.



I guess Airlines are next to find out right? Next time you take a flight make sure to yell at the pilot that you won't let the plane take off until they ripped out all the redundant stupid and useless sensors and hardware. That they are unsafe and put the hundreds of passengers in danger! When they ask you why tell them elon told you so!

"Boeing has long embraced the power of redundancy to protect its jets and their passengers from a range of potential disruptions, from electrical faults to lightning strikes.​

The company typically uses two or even three separate components as fail-safes for crucial tasks to reduce the possibility of a disastrous failure. Its most advanced planes, for instance, have three flight computers that function independently, with each computer containing three different processors manufactured by different companies.​

So even some of the people who have worked on Boeing’s new 737 MAX airplane were baffled to learn that the company had designed an automated safety system that abandoned the principles of component redundancy, ultimately entrusting the automated decision-making to just one sensor — a type of sensor that was known to fail. Boeing’s rival, Airbus, has typically depended on three such sensors.​

“A single point of failure is an absolute no-no,” said one former Boeing engineer who worked on the MAX, who requested anonymity to speak frankly about the program in an interview with The Seattle Times. “That is just a huge system engineering oversight. To just have missed it, I can’t imagine how.”​

Boeing’s design made the flight crew the fail-safe backup to the safety system known as the Maneuvering Characteristics Augmentation System, or MCAS.​

A faulty reading from an angle-of-attack sensor (AOA) — used to assess whether the plane is angled up so much that it is at risk of stalling — is now suspected in the October crash of a 737 MAX in Indonesia, with data suggesting that MCAS pushed the aircraft’s nose toward Earth to avoid a stall that wasn’t happening. Investigators have said another crash in Ethiopia this month has parallels to the first.​

​

Boeing has been working to rejigger its MAX software in recent months, and that includes a plan to have MCAS consider input from both of the plane’s angle-of-attack sensors, according to officials familiar with the new design. The MAX cockpit will now include a warning light that will illuminate when the two angle-of-attack sensors disagree. "​

You are confusing VW's 1 forward camera ADAS using 2017's EyeQ4 with the current gen 360 cameras supervision.
Second of all, there are tons of video of supervision in areas that are orders of magnitude more harder to drive in than typical US cities.
secondly, maybe Amnon isn't in the game of repeatedly lying every year for the past 7 years that he guarantees he will have L5 "next year" and chanting "neural nets, neural nets, neural nets" without having a clue.

Amnon is an actual computer scientist, Ph.D from MIT and is still a professor of AI/ML at the top #1 Israel University.
Lastly what makes supervision any different from FSD Beta? Oh wait, its actually fully hands free door to door and works in china which is orders of magnitude harder to drive in than the US.

Meh, looks like they have pre-mapped data using their REM map data collection system. So everything is pre-determined and doesn't require AI to make a lot of real time decisions.

The hardware might be..."affordable"...(cough probably will cost 30k for the customer), but there's a very expensive operating expenditure behind the scene maintaining the maps.

Also mobile eye supervision is just enhanced autopilot..don't see how that's any special.

Tesla is still working on a much harder problem while Mobile eye just needs to focus on not hitting things, which is pretty easy using lidar. Tesla needs to figure out the road it has never seen before AND not hit anything.

 

[Bladerskb]

Meh, looks like they have pre-mapped data using their REM map data collection system. So everything is pre-determined and doesn't require AI to make a lot of real time decisions.

The hardware might be..."affordable"...(cough probably will cost 30k for the customer), but there's a very expensive operating expenditure behind the scene maintaining the maps.

Also mobile eye supervision is just enhanced autopilot..don't see how that's any special.

Tesla is still working on a much harder problem while Mobile eye just needs to focus on not hitting things, which is pretty easy using lidar. Tesla needs to figure out the road it has never seen before AND not hit anything.

You love spreading misinformation and FUD don't you?

1. You clearly have no clue what AI is.​

​

2. Supervision's BOM cost is less than $1k and the option is sold for $5k on the 2021 Zeekr for the full door to door feature. But keep making up fud.​

​

3. Supervision IS NOT enhanced autopilot. Its a door to door system. How many times does that have to be repeated? I guess you don't care to have even watched the video i linked explaining supervision? Probably not, only interested in spreading FUD.​

​

4. This is supervision and according to you, all the pedestrians and cars in these videos that supervision is maneuvering and negotiating around/with ARE clearly fake and pre-determined. If a Tesla FSD Beta stops for pedestrians and maneuvers around them. THE GREATEST AI EVER. ARE WE SURE TESLA's FSD BETA ISN'T SENTIENT? If a non-tesla system stops for a pedestrians and maneuvers around them. GLORIFIED AEB SYSTEM! THERES NO AI HERE, ITS JUST DOING EMERGENCY BRAKE CHECK.​

Camera-only Supervision in New York

Camera-only Supervision in Isreal

Camera-only Supervision in Munich

 

[Knightshade]

Take with appropriate sized grain of salt (last sentence specifically)

https://twitter.com/x/status/1478600217930907648

 

[Singuy]

You love spreading misinformation and FUD don't you?

1. You clearly have no clue what AI is.​

​

2. Supervision's BOM cost is less than $1k and the option is sold for $5k on the 2021 Zeekr for the full door to door feature. But keep making up fud.​

​

3. Supervision IS NOT enhanced autopilot. Its a door to door system. How many times does that have to be repeated? I guess you don't care to have even watched the video i linked explaining supervision? Probably not, only interested in spreading FUD.​

​

4. This is supervision and according to you, all the pedestrians and cars in these videos that supervision is maneuvering and negotiating around/with ARE clearly fake and pre-determined. If a Tesla FSD Beta stops for pedestrians and maneuvers around them. THE GREATEST AI EVER. ARE WE SURE TESLA's FSD BETA ISN'T SENTIENT? If a non-tesla system stops for a pedestrians and maneuvers around them. GLORIFIED AEB SYSTEM! THERES NO AI HERE, ITS JUST DOING EMERGENCY BRAKE CHECK.​

Camera-only Supervision in New York

Camera-only Supervision in Isreal

Camera-only Supervision in Munich

Dude, I said the roads are predetermined via HD maps, not negotiating around pedestrians and cars. In fact I said that's the only thing the car needs to deal with.

Do you have FSD Beta? You can tell some of the decision making the car is doing and sometimes you wonder why it decide to do the wrong thing vs the right thing. What makes FSD more AI like is it's unpredictable nature because it's really hard to even reproduce a problem (like why did it decide to drive 20 mph under your set speed behind the car in front for 0.76 miles before wanting to switch lanes vs another car it wants to switch lane instantly?).

Lastly can you please stop with your condescending BS? Be helpful and make your case. No need for personal attacks.

 

[MXWing]

Take with appropriate sized grain of salt (last sentence specifically)

https://twitter.com/x/status/1478600217930907648

Seems more certain than if the tweet read as:

"Elon Musk visited the Longi factory and said it would be two weeks before their new US plant will start operation".

So on Giga Texas, is there anything administrative at all that is outside of Tesla's control to commence operations?

 

[Skipdd]

Dude, I said the roads are predetermined via HD maps, not negotiating around pedestrians and cars. In fact I said that's the only thing the car needs to deal with.

Do you have FSD Beta? You can tell some of the decision making the car is doing and sometimes you wonder why it decide to do the wrong thing vs the right thing. What makes FSD more AI like is it's unpredictable nature because it's really hard to even reproduce a problem (like why did it decide to drive 20 mph under your set speed behind the car in front for 0.76 miles before wanting to switch lanes vs another car it wants to switch lane instantly?).

Lastly can you please stop with your condescending BS? Be helpful and make your case. No need for personal attacks.

There are not only many threads on this topic, but an entire forum AI, Autopilot, & Autonomous/FSD

I know why this came up here, but the religious war of Lidar vs Pure Vision, vs a hybrid, and the companies that are doing each is well documented there. I will say this debate caused me to go there and look at some more current threads seeking an updated comparison - that others well acquainted with the various approaches are posting in (such as @diplomat33). Suggest that the level of detail @Bladerskb is providing go back there.

Agree with @Singuy, there is no call for the tone.

 

[Accident]

Futures, Asia, bitcoin, bonds… pointing to another red day tomorrow.

 

[TheTalkingMule]

Interesting....

The 10 Best Performing Commodities Of 2021 | OilPrice.com

Crude, Coffee And Copper rallied in 2021, but the real star in commodities posted triple-digit price gains

oilprice-com.cdn.ampproject.org

#1 Lithium
52-Week Change (2021) : 477%

Price (Dec 31): $42,256 per tonne

 

[PeterJA]

Titans of car-making are plotting the overthrow of Elon Musk

The world’s two biggest automakers — each sold roughly 10 or 11 cars for every one Elon Musk did last year

www.livemint.com

An unusually good article at Sucking Alpha provides a rebuttal to "automajors outsell Tesla" and "Tesla is losing marketshare." The author points out that Tesla is outselling everyone in the market segments in which it participates, in the country where it is least production-constrained. Guess what will happen when Tesla expands to all segments and countries.

Tesla Stock: Expanding Auto Segments, Margins, And Multiples (TSLA)

Many investors think Tesla is a frothy meme stock that can't be taken seriously. But its high multiple can be justified. Here's why.

seekingalpha.com

 

[PeterJA]

Steven provides an amusing pep-talk about the likely stock movement in the next four weeks.

 

[Beltsbear]

The infamous Berlin Airport took 16 years to finish. Completed to that, Giga Berlin is moving at lightspeed. Have a little patience.

It’s just hard to watch a country like Germany who many thought of the best and the brightest making bad decisions. Shutting down three nukes in the prime of their low carbon life while leaving multiple coal plants running. Then giving Tesla a hard time with water while slathering 100x the volume of water for continuing coal in the same region. I know it will pay off in the end. But time now is precious, not just for Tesla but for the environment.

 

[Singuy]

Interesting....

The 10 Best Performing Commodities Of 2021 | OilPrice.com

Crude, Coffee And Copper rallied in 2021, but the real star in commodities posted triple-digit price gains

oilprice-com.cdn.ampproject.org

So according to this article, Tesla deployed 18,700 tonne by building 500k cars.

Tesla consumes more lithium than four closest rivals combined

Using that calculation, it now cost $1580 in lithium today vs $300 dollars end of 2020(however lithium prices were unusually low in 2020. This is actually pretty significant and may hit gross margins. Perhaps this is why we saw the constant price increases throughout 2021.

Another possibility is that Tesla is hoarding mass tonnage of Lithium, ready for 4680 mass production and led to massive spike in lithium prices.