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

[StealthP3D]

Funnily enough a member of my family has a Hyundai/Kia BEV, and yes it has also been hit by the LG pouch problems. So (according to them) the software now has a smaller usable SoC (so room reserved both at top and bottom of the SoC), throttled charging rate, and no parking allowed inside. That doesn't just sound just like machine alignment issues affecting only tabs and suchlike. Can you say dendrites ?

More generally the problem with Hyundai/Kia in the BEV-space is their access to batteries. They appear to be getting approx 75% of their batteries from LG with lesser amounts primarily from SKI and Panasonic. They will need to resolve this urgently if they are to recover the ground they are losing. A problem they share with every other LG pouch user.

View attachment 772394

While I wish other BEV manufacturers could pull off this move to electrification more gracefully, incidents like this show they haven't done their battery homework. Because these problems could mostly be avoided by:

1) using suitable cells for the job

2) developing a battery management system (software and hardware) that properly manages voltages, amperages and temperatures of every cell.

Even the most consistent and best battery will perform terribly if it's not managed within the proper parameters all the time. Tesla knows how to do this and that is a big advantage when it comes to sourcing the cells in the highest demand at the most favorable prices. Because battery manufacturers obviously want their cells to perform well. It's a huge liability to hand cells over to a manufacturer that doesn't fully know what they are doing.

 

[Cosmacelf]

This is so incredibly true and relevant. We have FSD beta on both our vehicles. Overwhelming reason I disengage or do something different from FSD is to avoid road rage from aggressive drivers. It’s comical that human drivers brought the stop sign roll to the attention of nhtsa when humans do the roll every damn time. It’s even to the point where I don’t feel safe not rolling through with another vehicle behind me.

The NHTSA was being vindictive on the rolling stop issue. Or a bunch of Karens. When a safety agency doesn't know how humans drive, it is pretty bad.

 

[AndreP]

From a long term investment standpoint, it doesn't really matter when Robotaxis happen as long as Tesla is the first one to get to a large scale deployment. Think about the Model 3 ramp, they were a couple months behind production targets, but it didn't matter in the long run because no one else was even ramping at all. As far as current scale, it seems as if Tesla is in the lead by ~2 orders of magnitude 60,000 Beta vehicles Vs 700 Waymo cars (In California). Granted this really isn't an apples to apples comparison (Tesla is geoscaled, Waymo has higher localized ability but is geofenced), however it does lend credence to the thought that Tesla is currently in the lead in getting to a scaled robotaxi service.

That is completely fair. When this technology is ready, the change will likely overhaul the entire automotive landscape.

When that will happen is heavily debated even among the relatively few people who are experts in the technical nitty gritty (AI, sensors, etc) underlying this stuff

The NHTSA was being vindictive on the rolling stop issue. Or a bunch of Karens. When a safety agency doesn't know how humans drive, it is pretty bad.

The NHTSA was assuredly acting on the complaints of FSD Beta users who don't want their vehicles rolling through stop signs when they are liable for the car's actions

 

[StealthP3D]

To assume that elemental properties cannot be engineered to optimize material use is to deny the most fundamental of human abilities.

Respectfully, you have quite completely missed the point. Tesla, SpaceX and Elon Musk are all about engineering. Physics has laws. Engineers, at their best, work with the laws to do impossible things.

I think you have missed my point.

4680 does not refer to engineering or a specific technology. It is a form factor, that is all. A cylindrical cell 46mm in diameter and 80mm in length. There will be advancements in chemistry and there will be other form factors to fit various applications but, undoubtedly, the 4680 form factor will be useful in multiple applications for decades to come. Not multiple centuries durable, but 40-80 years durable, IMO.

 

[Mengy]

The NHTSA was assuredly acting on the complaints of FSD Beta users who don't want their vehicles rolling through stop signs when they are liable for the car's actions

You don't truly believe that, do you?

 

[Right_Said_Fred]

The Autonomous Vehicle subforum is probably one of the best places on the Internet to discuss this new technology and most people in there are very reasonable about it.

I enjoy taking in the other perspective from people here and sometimes offering my own for the sake of understanding. If you’re only interested in one side of the coin, I think you’re doing yourself a disservice.

You are right that the Autonomous Vehicle subforum is one of the best places to discuss this new technology. In fact, it's the only place to discuss it. We don't need a repeat of that discussion in the investment thread (which happens anyway every few weeks, despite of constant efforts by moderators to confine it to the subforum where it belongs).

We appreciate you want to show us the other side of the coin on autonomous driving, but thank you, we don't need it. If we need it, we can find plenty of it in the autonomous subforum.

A few years ago we had someone here who wanted to show us the other side of the coin on Tesla's service and communications. He warned us repeatedly about the future, sold his shares and lost millions in potential profit. The rest of us made milllions.

In short: we don't need saving.

 

[Cosmacelf]

I would guess that opinions may vary depending on your particular area. A few particular routes I happen to drive regularly, the beta has a very difficult time with and has not improved as the versions have been released. While I do see improvements in some areas of driving with new versions, the way it reacts to the situations which cause problems for me is not very inspiring. To me it just serves to highlight how many edge cases exist, and how far away we are from it working well broadly. I'm bullish long term, but I am of the opinion that it will take quite a bit longer than recently estimated. I'd be super happy to be proven wrong though!

While I agree that the current FSD beta has a long way to go, be aware that improvements won't only be addressing edge cases (which would imply a very slow rate of improvement). They keep working on the underlying perception system which will improve all edge cases at the same time. But those major underlying improvements only happen every 3-6 months or so since they are so major.

BTW, I think it was during the latest Starship update that Elon said he spends most of his time on Starship engine development and FSD. Considering how much he must know of FSD development, he seems to be falling into the engineer's trap of knowing about the next big re-write and thinking, "This time, it'll fix everything!". But no really knows until they build the software/neural net and try it out.

Personally, I think Elon has a big giant blind spot when it comes to AI and the limits of current neural net perception. I think he was blinded by how well the hack of lane detection worked and extrapolated way way too much from there about how easy FSD would have been. From following how the system works during the past several years, I personally wouldn't have felt the system at any point would have been able to handle FSD. They are only now building a proper 3D vector space in the neural net. And only now attempting to properly integrate time based remembering of that vector space, but even that is still a hack and isn't being done right.

 

[Big Time]

Im sure Nikola will do well so theres that.
Waiting to see how "Swift and powerful" our response will be.

My guess is with the way things have been going, that were going to get knocked down again into the 700s.

 

[JRP3]

Everything taking a dip, some news?

 

[StealthP3D]

I'm just saying it's weird to see someone post exclusively anti-Tesla sentiments in the Tesla crash threads when this is a fan site.

Even crashes that clearly had nothing to do with autopilot and we're just the FUD-of-the-week such as the Houston incident.

He's here to save us from our own optimism, just like the thousands of other 'helpful' posters since TSLA stock went public at a split-adjusted $3.40 per share.

 

[Cosmacelf]

While I wish other BEV manufacturers could pull off this move to electrification more gracefully, incidents like this show they haven't done their battery homework. Because these problems could mostly be avoided by:

1) using suitable cells for the job

2) developing a battery management system (software and hardware) that properly manages voltages, amperages and temperatures of every cell.

Even the most consistent and best battery will perform terribly if it's not managed within the proper parameters all the time. Tesla knows how to do this and that is a big advantage when it comes to sourcing the cells in the highest demand at the most favorable prices. Because battery manufacturers obviously want their cells to perform well. It's a huge liability to hand cells over to a manufacturer that doesn't fully know what they are doing.

The fundamental problem that the other EV manufacturers didn't take into account is cell manufacturing defects. Large cells means when a manufacturing defect rears its head (and over millions of cars they will!), you've got to cut off and isolate that cell, and not only electrically but mechanically as well or else you'll have contagion. Bigger cells makes that much much harder. You cannot design an EV with the expectation that no battery cell won't go into thermal runaway sometime in its life across the fleet. You have to assume that this will happen.

So the fundamental problem isn't with cells, it is with the battery design. Can the battery handle a single cell blowing up?

I don't know if safety was the major considered engineering reason when JB was tinkering with different cell types in the early Roadster days, or whether they chose the 18650 cells due to overall performance/cost reasons. Either way, Tesla dodged a huge bullet.

At any rate, GM and Kia and others that use pouch cells did not design the battery properly to deal with individual cell thermal runaway, whereas Tesla certainly did with their 18650 cells (the small fusible link was only one part of a complex system of mitigation measures). I believe LG is the one that actually designed the battery pack as well - they sold the complete battery as a unit to at least GM so the liability lies squarely with them on the engineering side. But GM is the one that believed that a battery cell manufacturer could build a robust EV battery pack when they had no demonstrated expertise to do so. Congrats GM, you saved yourself the cost liability of a bad battery design, yet your brand is trashed and you now can't sell EVs. Genius.

Interestingly, BYD seems to have learned this. Their LFP blade battery is supposedly very robust to cell defects. Different chemistry, lower energy density, but the point is that cell defects shouldn't cause the car to catch fire.

 

[strago13]

IMHO, TSLA is exaggerating the macros by x10 today

 

[M|S|M MYP]

Shook the couch for 8 more shares at $820....this will soon pass and we will all be happy again

 

[Big Time]

Sold the couch, I'm dependent on cub scouts to feed me now

 

[Beltsbear]

Im sure Nikola will do well so theres that.
Waiting to see how "Swift and powerful" our response will be.

My guess is with the way things have been going, that were going to get knocked down again into the 700s.

Tesla may be considered a value stock if the 800’s continues into q3!

 

[elasalle]

looks like they shooting for 800 and all the stop losses they can trigger ...

(when will lunch break be over ...)

 

[TheTalkingMule]

He's here to save us from our own optimism, just like the thousands of other 'helpful' posters since TSLA stock went public at a split-adjusted $3.40 per share.

I just think it's weird someone would sign up for a Tesla fan site and immediately begin by trashing Tesla/FSD in crashes where clearly it was simple standard driver neglect. How does a fan never have anything good to say about the target of his fandom?

I hate an echo chamber as much as the next fellow, and I've gone so far as to PM you @StealthP3D back in the day when I thought you were shouting down dissent to aggressively. But I think mods are too nice here and there are some clear trolling accounts that should be axed. Or at least banned from the non-technical sections. All they do is bump conversations that have been done and dusted for months/years.

Bickering and unease on the rise. Suggestions of bannings and moderator complaints are flying. Hyper-bullish TMC indicators!

 

[mongo]

TSLA drop aligns in time (10:28 eastern) with article about UI/ driver run over (ended up fine) getting published. Based on this thread:
Why didn't the car go into Park Automatically here?
Driver error: didn't put it in park, opened door, car rolled, pressed accelerator, fell out, car rolled over them and stopped. No major injury.

 

[JRP3]

TSLA drop aligns in time (10:28 eastern) with article about UI/ driver run over (ended up fine) getting published. Based on this thread:
Why didn't the car go into Park Automatically here?
Driver error: didn't put it in park, opened door, car rolled, pressed accelerator, fell out, car rolled over them and stopped. No major injury.

Except everything started dropping at the same time.

 

[SwedishAdvocate]

Everything taking a dip, some news?

Could this be it? Gary Black has tweeted about it...

Elon Musk Laughed at the Idea of Tesla Using Too Much Water. Now It’s a Real Problem​

Scientists say the factory Tesla is building in Germany would exacerbate local climate change issues

Source: Bloomberg - Are you a robot?

The link works. I don't know why it changes the title...