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

[AndrewZ]

I thought they all had cabin cameras since the Model 3? Which were all upgraded to HW3 from HW2.5, ours included. So no cabin cameras are already a diff branch, no?

IR cams needed - Assumed bc others have it? Didn't Tesla go to pixel hardware level on all these cams with much higher resolution?

You folks seem to be more informed here on this. I'll just crawl back into bed and start over...

Yeah there probably already is a separate branch but would need to dig a bit to confirm that, others probably know off the top of their heads

I think IR cameras are better at seeing through stuff like sunglasses and in the dark etc, whether regular camera vision can be supplemented sufficiently to get the regulators on board is the question. I think some of the more recent vehicle builds are using IR cameras too, I see Reddit threads etc of people with late 2022 and now 2023 Model Ys that appear to have infrared cabin cameras.

So many different hardware configurations out there now. And then if you start allowing people with certain builds/vehicles to have better functionality than those with other configs, that could create some animosity if everyone paid for the same thing.

 

[SOULPEDL]

There is nothing that suggests Tesla will ever settle for less than complete autonomy.

Ark Invest used to tout that Tesla should have started the ride-hailing service with a driver and begin collecting revenue until full autonomy is possible. Tesla never did that either.

It isn't important to Elon who does it first, rather it is who does it right.

Both Geo-fencing (supervised), and Ride-hailing with a driver have a human element that can fail. Full autonomy that exceeds an appropriate number of 9s to the right of the decimal point without a human watching over its shoulder is the goal.

I'm glad you put (supervised), I'm talking about unsupervised geofencing, and I further suggested by States as well.
Now as to the fuss that would be created if I got it and you all didn't because I'm in Chandler (or Fla for example)... ya we'd hear that for sure. Doesn't mean it's not seen as a gradual roll-out, coming soon to your locality. Just like Superchargers. Ouch!

 

[Tony73]

I don't know how many times I have to say this but a geofenced ride hailing service is equivalent to going to the back yard and lighting up a bunch of cash to make a bonfire. Show me a profitable robotaxi service and I'll shut up. That business model will NEVER be profitable. The bigger the expansion, the more operating cost they accrue. It's not rocket science to do the calculation when you have more high paying geo fencing/trouble shooting/service engineers than passengers on a given day.

Do you know the number of high paid engineers are required for a regular taxi? ZERO. Tesla's approach is full autonomy with close to zero engineers behind the scene. It's either this or BUST. All of the paper math calculations come from the assumption that operating cost is close to zero besides a server to run the app or else a robotic taxi will always be more expensive than a manual driven taxi.

Full autonomy may be the goal but it’s not happening anytime soon. Even after the technical challenges are solved, the regulatory hurdles at the federal, state and local level will take years to clear. This is all uncharted territory.

Tesla has already won autonomy because it has already monetized it. The drivers willing to pay for supervised autonomy are overwhelmingly Tesla customers. No other company could have pulled off FSD beta on this scale. The immediate goal should be to get FSD to more general adoption; increase the take rate; get to millions of subscribers. Full autonomy will cost tens of billions to develop and deploy and he who generates the most cash along the way will win. Google has figured that out and cut Waymo loose.

 

[insaneoctane]

You could be right. Maybe they're that good at hiding it. This was the one, pretty hard to tell.
But why would it be recalled? Eyes OFF the road are assumed to cause a nag still.

Tesla has had multiple exchanges with the 2 federal agencies about driver monitoring frequently. This article mentions Tesla increasing the frequency in 2018. I just don't see the nag going away, unfortunately.

 

[Singuy]

I guess you'll have to hear it one more time. I'm referring to a goefenced Tesla ride-hail service. Their data must be showing risk levels by geo-location - at least this is what I'd expect. Again, Chandler please.

Cruise automation generated 53M dollars in revenue and lost GM 1.4 BILLION dollars. Ride hailing service is EXPENSIVE, especially when you add another layer of costly complexity. So I am thanking the heavens that Tesla is not doing some kind of small geo fencing just to compete in that space and lose all the sweet free cash flow we have been generating.

53M dollars of revenue @ 10% global FSD take rate will require Tesla to sell 35k cars to make it up. So as GM generates 53M per year, Tesla generates 53M per week.

 

[Dutchie]

Another hit story about FSD. However, this one has a lot of detail. Author has done his homework. This does not mean i agree with it.

https://www.washingtonpost.com/technology/2023/03/19/elon-musk-tesla-driving/

 

[Electroman]

See

The $100/KWh point for Tesla was reached some time ago, I believe. It wouldn't surprise me if their cost was well below that currently, which may change that $68K pack starting value significantly...

that number $100/kWh is at cell level or pack level?

 

[SOULPEDL]

Cruise automation generated 53M dollars in revenue and lost GM 1.4 BILLION dollars. Ride hailing service is EXPENSIVE, especially when you add another layer of costly complexity. So I am thanking the heavens that Tesla is not doing some kind of small geo fencing just to compete in that space and lose all the sweet free cash flow we have been generating.

53M dollars of revenue @ 10% global FSD take rate will require Tesla to sell 35k cars to make it up. So as GM generates 53M per year, Tesla generates 53M per week.

Agree, but I think you're missing some posts where I clarified.
Enough said, we shall see. Maybe May.

 

[loquitur]

[....]

So many different hardware configurations out there now. And then if you start allowing people with certain builds/vehicles to have better functionality than those with other configs, that could create some animosity if everyone paid for the same thing.

"Specifications are subject to change without notice."

 

[DFibRL8R]

Another hit story about FSD. However, this one has a lot of detail. Author has done his homework. This does not mean i agree with it.

https://www.washingtonpost.com/technology/2023/03/19/elon-musk-tesla-driving/

I don’t click on WaPo articles by Faiz Siddiqui, he has had an anti-Tesla/Elon bias for years.

 

[scaesare]

See

that number $100/kWh is at cell level or pack level?

Nearly 5 years ago at the 2018 Annual Shareholders Meeting Elon stated they were on track to reach $100/KWh at the cell level later that year and expected it at the pack level within 2 years. Of course, he did caveat that with raw materials prices needing to stay on their current trajectory, which probably got derailed by COVID somewhat.... but he also mentioned other advancements in energy/price density coming as well:

I'd be surprised if they weren't there (or lower) at this point. I can't find it, but there was an event some time ago where an audience member asked, and Elon's "No Comment" demeanor looked rather smug...

 

[FSDtester#1]

You could be right. Maybe they're that good at hiding it. This was the one, pretty hard to tell.
But why would it be recalled? Eyes OFF the road are assumed to cause a nag still.

I have watched him for years... he is excellent at hiding the left leg providing resistance. Last week he drove his dad's MS... He couldn't do that with a yoke so he had to consistently tug the yoke.

 

[nativewolf]

I think @nativewolf caught my oversight. Fixed it.

Yet the timing of this announcement seems leading... where's Tesla ride-hailing in China? Maybe not quite ready for China general case. The above companies are geofenced, big difference. I still think that Tesla's could take it cautiously by releasing in some cities only at first, States as well. Recommending Chandler Az however, it's been raining ALL winter.

California has a very well defined process as does Arizona and both Cruise and Waymo offer service and shortly others will be there. Waymo is leading...momentum with Waymo is clearly accelerating. By the end of the year they'll do robotaxi across san fran and much of LA would be my guess. We'll see in a few months.

Lots of discussion on a good thread over in the autonomy section. In the investment thread it gets culled so if you are curious head over to the autonomy section.

 

[Brn2Run]

Another hit story about FSD. However, this one has a lot of detail. Author has done his homework. This does not mean i agree with it.

https://www.washingtonpost.com/technology/2023/03/19/elon-musk-tesla-driving/

Not good when the CEO takes the top engineers out of the company you own shares in, to work in his own private company.. Only the cult cannot see that is troublesome..

 

[philw1776]

Not good when the CEO takes the top engineers out of the company you own shares in, to work in his own private company.. Only the cult cannot see that is troublesome..

Not a good thing for us shareholders as stated.

Yet the article contradicts itself. Example after example about how Elon is a bad manager, driving FSD engineers too hard without adequate release test safeguards, etc. Then concluding opinions cited saying Elon is needed back from Twitter ASAP to run TESLA.
Which is it? Author shows his bias.

 

[bkp_duke]

Not good when the CEO takes the top engineers out of the company you own shares in, to work in his own private company.. Only the cult cannot see that is troublesome..

Oh, so YOU know which engineers in Tesla are the best? And YOU know which ones were temp moved to Twitter? And YOU know that this temp move impacted FSD development?

 

[Singuy]

California has a very well defined process as does Arizona and both Cruise and Waymo offer service and shortly others will be there. Waymo is leading...momentum with Waymo is clearly accelerating. By the end of the year they'll do robotaxi across san fran and much of LA would be my guess. We'll see in a few months.

Lots of discussion on a good thread over in the autonomy section. In the investment thread it gets culled so if you are curious head over to the autonomy section.

Momentum of them losing more money is clearly accelerating.

Just like Lucid with their fancy drivetrain, every one of these companies razzle dazzle the public while driving themselves into bankruptcy. Tesla is sitting around watching them burn as they pocket more positive cash flow.

Waymo is L4
Boston Dynamics can parkore
Lucid can go 500 miles

These are all science projects until they are cash flow positive.

 

[Featsbeyond50]

Full autonomy may be the goal but it’s not happening anytime soon. Even after the technical challenges are solved, the regulatory hurdles at the federal, state and local level will take years to clear. This is all uncharted territory.

Tesla has already won autonomy because it has already monetized it. The drivers willing to pay for supervised autonomy are overwhelmingly Tesla customers. No other company could have pulled off FSD beta on this scale. The immediate goal should be to get FSD to more general adoption; increase the take rate; get to millions of subscribers. Full autonomy will cost tens of billions to develop and deploy and he who generates the most cash along the way will win. Google has figured that out and cut Waymo loose.

I don't know if this is right, but I heard autonomous driving is permissible in most of (or maybe 40% of) the United States right now. Probably only because the regulator types are behind the curve but, nevertheless, the regulatory hurdle might be less significant than one would think.

 

[Singuy]

So I guess this is about 1/13th of the participants that has the app? Which means we have to multiply 86 X 13= ~1000 people now has V11.

https://pircher.bz.it/teslafirmware/fsdbeta

 

[nativewolf]

Momentum of them losing more money is clearly accelerating.

Just like Lucid with their fancy drivetrain, every one of these companies razzle dazzle the public while driving themselves into bankruptcy. Tesla is sitting around watching them burn as they pocket more positive cash flow.

I mean seriously...Google running out of money? Next year you'll see Geely delivering robotaxis for them to sell service at scale across many metro regions. That's the window I see for Tesla because people will discover the robotaxi business sucks. Just my hunch. I think robotaxi will be grossly crowded and lack profits. I also think Tesla will be frozen out of China due to security concerns. But Waymo won't run out of money on the way to robotaxi service. Cruise..no idea. Mobileye no idea. Waymo is doing real robotaxi rides now. Speed limit is basically highway driving speds. Door to Door. Still working at things but they seem to finally be scaling. Until Geely delivers cars they only need to do so much so fast. Once Geely starts deliveries...poof..no profits. 2024 My prediction we discover robotaxis will produce no profit.