Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

This site may earn commission on affiliate links.
Is Buffet invested in car insurance business? That might explain the snarks from Buffet and Munger ;)

Btw, Buffet used political lobbying twice now against Tesla - once in Nevada against solar and now again in Texas in an attempt to close Tesla service centers (luckily the status quo remains for now)

Yes, Buffet rarely says something that isn't calculated to help his businesses. I remember him spreading destabilizing FUD in the wake of the 2008 financial meltdown that was designed to push policy makers to help his insurance businesses. I thought it the height of irresponsibility at the time. Buffet has carefully constructed a "golly gee aw shucks" persona while actually acting as a ruthless cold hearted business person. Do remember that Buffet (Berkshire) owns about 25% of BYD, the Chinese EV manufacturer. Maybe, just maybe, BYD is a little bit nervous about Tesla entering China in a big way.
 
"{...with a} large amount of data so that you can prove with high confidence, statistically speaking, the car is dramatically safer than a person. And that adding in the person monitoring does not materially affect the safety...Proved by incidents per mile...There are enough crashes...so you can assess what is the probability of a crash; then there is another step which is probability of injury, then probability of permanent injury, then probability of death. And all of those need to be much better than a person, by at least perhaps 200 percent....The system is improving so much, so fast, that this {vigilance} is going to be a moot point, very soon...I'll be shocked if it's not next year, at the latest, that having a human intervene will decrease safety."
Then Musk will be shocked, I guess. I am quite sure that having a highly skilled human intervene will increase safety for... well... forever, really.

"Statistically speaking", with a population of mostly-bad drivers, doesn't tell you what happens with highly skilled drivers.

But Audie is right about one thing: *actuarially*, most humans won't be particularly useful. Most humans are terrible drivers. If Tesla were to start an insurance division, they could take advantage of this. Tesla does not have the paperwork skill or actuarial skill to start an insurance division, however.
 
Lately comments about musk as ceo despite buffet running investment firm and only company he tried running, the original textile firm Berkshire went bankrupt with him running it. He also has investment in a different electric car company as well
To be fair BYD is not just 'a different electric car company'. They are having major success in both battery supplies and busses, to name two areas. Further BYD is exceedingly well situated for long term growth and decent stability. Berkshire clearly favors less dramatic but successful investments.
 
Musk 'swipes away' the arguments with just talking about the endpoint: saying that it doesn't matter anyway once the statistics prove that FSD is safer than humans and any human decision to intervene would just increase the chance of the car getting crashed.
This will never ever happen (do I have to bring up the dirt one-lane roads with blind corners again), so Musk's view here is irrelevant.

Lex showed with his study that AP users have a higher awareness of road conditions when using the system and have a better understanding what the limitations are. So the scenario of drivers falling a sleep behind their steering wheel because they wrongly rely on AP to get them to their destination is less likely.
This is important. :)
 
  • Helpful
Reactions: Mobius484
On a side note, the hundreds of thousands of vehicles with the required sensor suite sold prior to the point FSD becomes available should also increase dramatically in value as soon as the TN becomes available and would likely be purchased by enterprising funds.

A couple comments on this interesting thesis.

EM makes this comment in the MIT podcast interview related to an appreciating asset.


This appreciating asset is first owned by TSLA as the manufacturer. Manufacturing is funded by demand from consumers for the vehicle. TSLA makes a profit from sales but......

TSLA remains in control of the SW load through the life of the vehicle. TSLA can monetize the value or appreciate the value of an already manufactured and sold vehicle. This is not specified as to all the permutations yet.

Imagine that TSLA offers a subscription SW load for the Tesla (something) Network. This is the real cash stream going forward.

Customer pay for the build of the vehicle today and then in a second life, payments/revenue come from mobility services supported by the TSLA SW load and app. Value appreciation through extended utility.

Another way to look at it would be the the crossing of 2 curves.
1. The depreciation curve of all vehicles which goes down over time.
2. The newly discovered/recognized monetization value of mobility services which is just starting to rise.

At some point the lines cross and the value of the vehicle as a platform for mobility becomes accretive to the depreciation curve and the appreciation happens.

My second comment is related to the MIT study.

I think the MIT study confirmed one thing. We are nearly as vigilent as backseat drivers (what we become when in autopilot) than we are when actually driving.
 
Then Musk will be shocked, I guess. I am quite sure that having a highly skilled human intervene will increase safety for... well... forever, really.

"Statistically speaking", with a population of mostly-bad drivers, doesn't tell you what happens with highly skilled drivers.

But Audie is right about one thing: *actuarially*, most humans won't be particularly useful. Most humans are terrible drivers. If Tesla were to start an insurance division, they could take advantage of this. Tesla does not have the paperwork skill or actuarial skill to start an insurance division, however.
In my opinion both Neroden and Audie has excellent points. After all, insurance is by it's nature actuarial. Just as with aircraft design, it is quite possible to design for high competence, thus requiring special training to operate a machine safely. From early Learjets to A320NEO, from Cobra 427SC to SP100D it is very possible to design to permit high performance while mitigating almost all, but not all, accidents.

What we might learn from early AP-1 accidents and B737MAX accidents is that if the standard of performance by a very few operators is substantially below the design standard, death and other disastrous consequences are prone to happen. Tesla has learned valuable lessons, but still needs to understand that being better than the average human is not enough; being almost perfect is barely adequate.
 
Last edited by a moderator:
I think the world is over estimating how much use ride sharing services will get.

Example. When purchasing identical items at a local store 2 miles away vs a store that is 10 miles away.
I just can't get my wife to assign driving cost to the purchase. She sees the cost as only the gas or electricity used.

She would see a fee of 30 cents a mile as excessive.
1. Parents who need to shuttle their kids around. Many need to leave work early for that.

2. The service will allow people to get rid of at least one car, the calculation is completely different. Several years ago I had a colleague just left college. He leased a BMW 3. After half year, he found between his bicycle and uber, he is throwing money into water with his lease payment. Your wife consider .3 per mile excessive because she has to pay for the family car anyway. Between insurance, deprecation and maintenance, car is the second biggest expense in the family. People will come to their senses when it is time to pay car insurance, or fix the transmission, or buy a new car.
 
Digital Trends on Twitter
Thing of beauty. #shellingpeas
It's interesting to see which bits they failed to automate. Side curtain airbag installation! Obviously a lot of wiring attachment. Various bolt tightenings. What else can people get out of this video?

Edited to add: I see jump cuts at particular points which I'd be interested in like seatbelt installation -- I figure they're hiding trade-secret or patented automated processes there.
 
Last edited:
That's the old nomenclature, deprecated now via Elon's Mar 29, 2019 tweet:

“Anyone who purchased full self-driving will get FSD computer upgrade for free. This is the only change between Autopilot HW2.5 & HW3. Going forward “HW3” will just be called FSD Computer, which is accurate. No change to vehicle sensors or wire harness needed. This is very important.”​

So we don't call it HW3, per Elon. We call it FSD Computer, because that's the major difference.

Cheers!

Ya, so there is a HW3 (aka APH4, aka FSD computer)

I, for one, will keep using HW3 as it is a lot faster to type than FSD computer (since FSD alone is ambiguous). Then you have HW2 + FSD computer upgrade as another case where the sensor HW is different (2.5 also changed radar suppliers).
 
Inflection Point?

I am having a hard time reconciling the latest statements from Elon regarding FSD, autonomous cars and the SP. It is well known on this board that Waymo has a private market cap of twice or more compared to TSLA. I am still not sure what Waymo is selling right now outside of the future FSD that they promise. But they have been anointed the future market leader and as such it is posited that will have access to the dominant leader position in what has been advertised as perhaps a two trillion dollar market.

Even based on that, TSLA SP makes no sense (although is explained with all the mostly nefarious powers that we know are aligned against TSLA, with good reason as far as they are concerned). Maybe Tesla is behind Waymo... but so much so that their FSD program should be assigned a zero or perhaps even negative number?

But now we have Musk stating "game, set match" on FSD in the next 6 to 18 months. The demonstration of FSD capabilities that Elon claims are still improving exponentially (with a new chip board which has NOT been maxed out on its processing power even will full frame processing from all eight cameras) is to take place on the 22nd. What if Elon is on target (and he is a lot more frequently than the MM or plenty of people on this board give him credit for)?

If Tesla is suddenly and undeniably ahead of Waymo or any other autonomous driving company based on its demonstration, does WS come around and agree that maybe another 20B should be added to TSLAs market cap? 50B? How about equal footing with Waymo at 100B? That would triple the SP from here, not taking into account that the upward violent movement would definitely and finally trigger a short squeeze of epic proportions. And it would not reflect an actual market that exists, in the same way that Waymo's market cap is based on a market that does not yet exist.

What if the demonstration is really meh and Elon has truly lost it? Does this mean the SP should collapse? Its FSD program is already valued at zero. I would expect some downward movement, but so much negativity is priced into the SP right now, the only way I see it continuing to descend would be hard confirmations that demand has really dried up and that no one wants to buy Tesla cars.

In trying to assign probabilities to the possible outcomes, I am coming up with great risk / reward analyses that pound the table on a probable dramatic SP price rise, although tamped down as best as possible by all the usual FUD methods. It is not possible to time exactly when this would happen, although April 22nd might be the date. Probably not.

TSLA has been range bound for almost two years now. Its position is vastly improved compared to two years ago. If FSD is demonstrated at a high level in addition to the advancements of the past two years, we should see a SpaceX launch on the SP.

Elon, you still look stressed and tired, and no one has more reason to be than you, insofar as your workload and competition. Please tell me you are on the mark here. Yeah, you left yourself an out that maybe you are wrong. But you also said you do not see it, and once again, the 99% of the global data input Tesla receives on a daily basis to train its NN because of the FSD sensor ready fleet of 400K plus cars already on the road concludes with 'game, set, match' Tesla.

I know that the bears are engaging in their usual ridicule of the whole affair and will continue full throttle until they are all dead from old age, but if this FSD demo lives up to its hype, there are plenty of deep pocketed powers that will be willing to jump on the bandwagon, FUD be damned. Might even see Buffet, step up to the plate. How else would he hedge against the possible long term terminal decline of his car insurance profits?

Yeah, I have a lot of TSLA stock for me. But am thinking of increasing position dramatically.

Would love to get knocked down here with solid arguments. If all you got is that Musk has lost his marbles, please do not bother so much as I am taking that into account.

As always, I seem to be a contrarian indicator. Do not take my advice. There are plenty of financial advisors out there you can rely on. Do not listen to me. And don't buy any individual stock that you can not afford to mark down to zero the next day.
 
This will never ever happen (do I have to bring up the dirt one-lane roads with blind corners again), so Musk's view here is irrelevant.


This is important. :)

I found the interview a little disappointing, as Lex tried several times to get Elon to engage in more philosophical discussion about the design philosophy for Autopilot and the mating of human and machine but Elon was in sales mode for most of the interview.

At one point I felt bad for Lex when Elon basically discounted his whole area of study with his statement that it's irrelevant to monitor humans because the real end goal is to have the machine be safer than the human, at which point the human could be detrimental to safety. I can understand Elon's point, as it's the same focus on the end goals that leads him to say that FSD needs to be solved with vision but in this case I don't have much confidence in Elon's timelines and I think having a human watching the machine will be of benefit for a long time to come.

Keep up the good work @Lex_MIT
 
Post of the year?


Only thing you missed, people will actually pay more for autonomous on top of the novelty/safety/comfort that you described:

  1. More room with no driver
  2. No driver in your space - you can do what you want
    1. Party time - your music
    2. No talking to the driver
    3. No driver to listen to your private discussions
    4. Freedom to stop off anywhere
    5. No risk of being molested etc. by the driver (parents will use more readily for their kids)

Trying to imagine being a passenger who’s just hailed a driverless Model 3 in the Tesla network. The car has been running for hours, it’s done 200 miles already that day. The car arrives at your location and you get in and you immediately detect a horrible odor as the car pulls away and then you see there’s a used diaper on the backseat and there’s fast food trash and beer bottles all around the floor and the car reeks but none of the windows are open. There’s no driver and the car is moving and it’s at this point you whip out your smartphone and try to reach a human thru the T Network app but all you get is a “we’re sorry all our customer representatives are busy your estimated hold time is seven minutes” and you realize you’ll be 2/3 of the way to your destination by then so you put the phone down in a cupholder without hanging up and quickly forget about it as you make the mistake of inhaling through your nose and you try to open a window but for some reason the windows are locked so you try not to breathe as the robot car makes its way across town on a two-lane road. Up ahead a school bus is approaching in the opposite direction and you see its blinking yellow lights go on as it slows down and then you see the bus driver extending the little stop sign along the side of the bus and the blinking yellow lights become blinking red lights and now children are getting off the bus and you realize this robot car is not slowing down and you hold on tight as you shout “watch out” but there’s no driver next to you to hear you and suddenly you hear a tinny voice coming out of your smartphone in the cupholder and you realize it’s customer service as your robot car whizzes right by the stopped school bus. As it passes by at the road’s posted speed limit you catch a glimpse of the shocked bus driver glaring at your robot car, his mouth open, clearly yelling at you while his bus horn is honking. Five minutes later you arrive at your destination and gladly get out as fast as possible. The robot car silently rolls away to points unknown. You go to your meeting and try to forget the experience. Your ride will be the robot car’s last for the day and now the robot car goes home to its owner who sees it pull into the driveway and up to its designated spot next to a Tesla home charger. The owner walks out, fetches the charge plug, plugs it into the robot car, and charges it, then opens the door, notices the stench, fetches some cleaning spray and towels, wipes the interior down, then, amid loud gripes and groans and cursing, collects the beer bottles and fast food trash and used diaper and throws all that away, then goes inside and signs into the Tesla Network - Owner Dashboard and checks who was the last passenger and finds your name and then reports you for trashing the car with beer bottles and fast food trash and a used diaper and leaves a very negative passenger reputation rating on your record even though it wasn’t your fault at all, but you won’t find any of this out for ninety two more minutes at which point you are done with your meeting and you pull out your phone, fire up the Tesla Network app to hail another robot car to take you back home, only to discover your negative reputation report and an error message saying “we’re sorry, but you have been temporarily banned from riding in the Tesla Network” and a line below it saying “Reason:” and then some lowercase text saying “left car dirty with beer bottles, fast food trash, and a used diaper.”

The future’s so bright, I gotta wear shades.
 
This will never ever happen (do I have to bring up the dirt one-lane roads with blind corners again), so Musk's view here is irrelevant.

I know you’re a huge FSD sceptic and you’re just trying to come up with what you feel is an impossible edge case to make your point. But there are plenty of markets where 100% of journeys can be completed without using unmarked dirt roads. In most of the rest, I’d tend to think the percentage of journeys that can be timeously completed without using such roads tends very close to 100%. What’s wrong with reverse geolocking (i.e. platform works on all routes apart from those marked red)? Seems like such a “limited” solution would still be worth a lot more than $46.2bn.
 
Tl;dr Good thing there is an interior camera.

Trying to imagine being a passenger who’s just hailed a driverless Model 3 in the Tesla network. The car has been running for hours, it’s done 200 miles already that day. The car arrives at your location and you get in and you immediately detect a horrible odor as the car pulls away and then you see there’s a used diaper on the backseat and there’s fast food trash and beer bottles all around the floor and the car reeks but none of the windows are open. There’s no driver and the car is moving and it’s at this point you whip out your smartphone and try to reach a human thru the T Network app but all you get is a “we’re sorry all our customer representatives are busy your estimated hold time is seven minutes” and you realize you’ll be 2/3 of the way to your destination by then so you put the phone down in a cupholder without hanging up and quickly forget about it as you make the mistake of inhaling through your nose and you try to open a window but for some reason the windows are locked so you try not to breathe as the robot car makes its way across town on a two-lane road. Up ahead a school bus is approaching in the opposite direction and you see its blinking yellow lights go on as it slows down and then you see the bus driver extending the little stop sign along the side of the bus and the blinking yellow lights become blinking red lights and now children are getting off the bus and you realize this robot car is not slowing down and you hold on tight as you shout “watch out” but there’s no driver next to you to hear you and suddenly you hear a tinny voice coming out of your smartphone in the cupholder and you realize it’s customer service as your robot car whizzes right by the stopped school bus. As it passes by at the road’s posted speed limit you catch a glimpse of the shocked bus driver glaring at your robot car, his mouth open, clearly yelling at you while his bus horn is honking. Five minutes later you arrive at your destination and gladly get out as fast as possible. The robot car silently rolls away to points unknown. You go to your meeting and try to forget the experience. Your ride will be the robot car’s last for the day and now the robot car goes home to its owner who sees it pull into the driveway and up to its designated spot next to a Tesla home charger. The owner walks out, fetches the charge plug, plugs it into the robot car, and charges it, then opens the door, notices the stench, fetches some cleaning spray and towels, wipes the interior down, then, amid loud gripes and groans and cursing, collects the beer bottles and fast food trash and used diaper and throws all that away, then goes inside and signs into the Tesla Network - Owner Dashboard and checks who was the last passenger and finds your name and then reports you for trashing the car with beer bottles and fast food trash and a used diaper and leaves a very negative passenger reputation rating on your record even though it wasn’t your fault at all, but you won’t find any of this out for ninety two more minutes at which point you are done with your meeting and you pull out your phone, fire up the Tesla Network app to hail another robot car to take you back home, only to discover your negative reputation report and an error message saying “we’re sorry, but you have been temporarily banned from riding in the Tesla Network” and a line below it saying “Reason:” and then some lowercase text saying “left car dirty with beer bottles, fast food trash, and a used diaper.”

The future’s so bright, I gotta wear shades.
 
Out here, human supervision is necessary to get into people's driveways, for the "last 1000 feet". Frankly, taxi drivers usually need to be supervised by the homeowner.

I think there is a very bright future for "the best driver assistance on the market", and a very dim future for "sleep in your car". I wish Musk would focus on the former; he's not going to achieve the latter in a commercially useful timeframe.