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

[Electroman]

Reduce the speed by 35% and vehicle fatality may go down to near zero.

I noticed that, at the speed limit (or say 5 miles below limit) my AP is flawless in almost all conditions.

 

[Motor Mouth]

Meanwhile... 4 pages of vision and radar debates has me concluding FSD is progressing quite well. Strong correlation (hunch) between the chatter here and FSD progress. This is just another one of those moments. Don't feed the trolls.

In my investment thesis there are 2 primary tipping points. (1) ICE to EV adoption by the masses. IMO this has already occurred. Not to say that EV currently outsell ICE (which they clearly don't), but rather that the point at which ICE can recover has passed. I believe it will only accelerate from here. (2) Human driving to FSD (or any other autonomous driving system). This one's difficult to predict and I personally would hedge it to a decade plus. I'd be pleased to be wrong, but it seems to me that the point at which there are substantial autonomous and human drivers sharing the road will be a highly challenging one for us (the royal "us") to handle for a variety of reasons. I'm fine with time passing for (2) since (1) is pretty plainly happening after being such a pipe dream for the past several years.

 

[betstarship]

I went back to this article and feels so long ago, but also so very true as how things have played out so far:

Liebreich: Peak Emissions Are Closer Than You Think – and Here’s Why | BloombergNEF

So here we are, standing on the threshold of a new decade. It will be, to paraphrase Winston Churchill, a decade of consequences. Play it right, and we have a chance of avoiding the worst impacts of climate change. Waste it, and we are in uncharted territory.

about.bnef.com

 

[JRP3]

Watch a movie or sleep, having no car payments and near zero chance of death but your travel time to work increase by 35%?

So you have to wake up earlier for work because of the longer trip and then try to fall asleep again in the car? No thanks. Just because someone else is driving me doesn't mean I want the trip to take longer.

 

[Motor Mouth]

So you have to wake up earlier for work because of the longer trip and then try to fall asleep again in the car? No thanks. Just because someone else is driving me doesn't mean I want the trip to take longer.

Jello Biafra was right - "Give me convenience or give me death!"

EDIT: /s

 

[EVNow]

Meanwhile... 4 pages of vision and radar debates has me concluding FSD is progressing quite well. Strong correlation (hunch) between the chatter here and FSD progress.

Strong correlation between people who use FSDb daily and those who are skeptical about robotaxi.

Tipping point to me is when Tesla actually starts publishing disengagement rates

 

[Nocturnal]

So you have to wake up earlier for work because of the longer trip and then try to fall asleep again in the car? No thanks. Just because someone else is driving me doesn't mean I want the trip to take longer.

Fewer accidents will mean an improvement in travel times. Easier/safer travel will result in more cars on the road.

It will be interesting to how these two things net out.

 

[Dutchie]

FSD already is safer than humans according to Tesla’s stats

View attachment 918511
What doesn’t exist is a level of safety mitigating sufficient risk to allow assumption of liability for the dynamic driving task

But still a long way from the March of 9’s. If it is five times better than average it is only in about 1/6 from right hand side of Bell curve. (About 84% of population perform worse, 16% better)

 

[AndrewZ]

Hmmmm. Statistically, if Tesla can provide data that shows FSD drives at an accident rate that is one order of magnitude superior to humans, then that seems like the correct basis to a winning defense against a lawsuit that would infer negligence. Assuming the data is general, accurate and representative -not unique or cherry picked, then NOT driving with FSD becomes negligent. But, I don't think Tesla has this general, accurate and representative data yet. While FSD is currently, statistically, safer, I think the current data is not a good representation of all roads, conditions, or situations, because our FSD drivers are choosing which roads, conditions, and situations they will engage the system, and thus limiting the data collection. This choice, IMO, excludes many dangerous roads, conditions, or situations from the dataset.

Once you have the dataset that stands up in court against negligence, then Tesla insurance cleans up....at 1/10 the accident rate, they could offer insurance for 50% reduction of premiums and still have a 5x reduction in liability/payouts!

It makes me wonder if Tesla will eventually start building in some type of geofencing, because I think it could perform pretty well in good weather and certain road types but getting good stats and solid performance on other road types and everything else required for a true generalized robotaxi seems so far away.

I don't know if geofencing would deal with the other stuff like redundancies and backups required in the event of damage, system failures, the ability to pull over safety if something happens, etc etc. But it seems far closer than generalized Level 4-5.

 

[EVNow]

FSD already is safer than humans according to Tesla’s stats

Wrong. You are badly misinterpreting.

FSD + Humans is safer than just humans. FSD is atleast 100x less safe than humans currently.

ps : My last comment on this for today. Anyone who is wondering what the actual status of FSD currently is - pls visit and read those of us who are into the weeds in the autopilot sub-forum.

 

[Gigapress]

Highly unlikely Musk cares about buses. He sees a world full of driverless vehicles where buses are rarely the right size vehicle. How often do 56 people need to start and finish in the same places at the same time? Once the driver is removed, it‘s about cost per seat and cents per km. Mass production favours smaller vehicles. 5 x 12 seaters for a lower price than one bus. With platooning you get energy efficiency.
Where I live there’s 2 buses per day. You could run a driverless 12 seater service every 15mins or on demand for less cost.

Bigger question, does public transport make sense at all if robotaxi is cheaper than a bus ticket? It may be that the cost becomes irrelevant and the deciding factor is which vehicles best utilise available road lanes. Different models for big cities, lower densities and rural.

We need to see if the Gen3 van has a "Mini-bus" mode, and what the seating capacity is.

If the van can seat 10, or more buses could be replaced by multiple vans.

For really high urban transport demand, trams and subways are better than buses.

The is still a niche for buses, but it is a smaller niche.

A group of autonomous vans "loosely platooning" almost approximates to a trackless tram. Not al carriages (vans) need to stop at every stop, those behind need to stop if the one in front stops. On average fewer stops might mean higher throughput. or the worst case is the same performance as a tram.

I must disagree. Busses are needed and used in nearly every urban area in the world. In much if the world large busses are profitable. There are substantial fleets of electric busses also, which, among other obvious benefits, are much cheaper to operate and maintain.

Tesla does not need to enter that market because it is becoming well filled by BYD, MarcoPolo, even Thomas with more and better ones coming too. You’re missing THE most developed BEV market in the world:

What’s really driving the trend in e-vehicles? Your local electrical bus

The share of the global municipal bus fleet that consists of electric vehicles far surpasses that of electric passenger vehicles.

www.weforum.org

High-density passenger transport was discussed in Master Plan Part Deux. As far as we know this is still the plan.

MPP2: With the advent of autonomy, it will probably make sense to shrink the size of buses and transition the role of bus driver to that of fleet manager. Traffic congestion would improve due to increased passenger areal density by eliminating the center aisle and putting seats where there are currently entryways, and matching acceleration and braking to other vehicles, thus avoiding the inertial impedance to smooth traffic flow of traditional heavy buses. It would also take people all the way to their destination. Fixed summon buttons at existing bus stops would serve those who don't have a phone. Design accommodates wheelchairs, strollers and bikes.

Reasoning by analogy to assume that diesel to electric transition for busses will be a one-for-one substitution will lead to wrong conclusions. Unfortunately, this is exactly what most people are doing, including transit planners and policymakers.

The major portion of the cost structure of traditional busses is the driver compensation and the operation and maintenance costs of the diesel powertrain, both of which favor having a smaller number of larger busses in order to exploit economies of scale. Autonomous electric busses, with zero driver expenses and perhaps 75% lower O&M costs would change this calculus in favor of smaller minibusses/vans with:

• higher percentage of seats actually occupied
• no center aisle wasting space and materials
• greater number of stops/stations in the service network yet more express-like service
• coverage further out away from the urban core deeper into the suburbs

Boring company is a huge wildcard in predicting where this is headed. If you believe it will succeed and scale quickly, then “public mass transit” might actually change to being mostly single or dual passenger minicars with some mix of larger sizes up to perhaps 16 passengers per vehicle max. More than 16 is probably unnecessary for a major route served by one or more Loop trunk lines, because that’s already sufficient to beat every train line in existence, and that’s not hyperbole.

In 2019, Elon mentioned a 1-second headway as being feasible, which would imply 57k passengers per hour per lane with a 16-pax van. I think 1-second is plausible with autonomous operation and the ridiculously tight control and security of the tunnels. Humans today typically do about 2 or 2.5 seconds between vehicles on highways. Autonomy eliminates about 1 to 1.5 seconds of human reaction time. We can debate a few tenths of a second here and there, but even with 1.5 second headway the numbers still look great.

Here’s some rough estimates I’ve done of potential AEV throughput in a Loop compared with high-volume train or bus lines. To a close approximation, Passengers/hour = Passengers/vehicle * Vehicles/hour = Pax/veh * (3600 sec/hr) / (Sec of headway between vehicles)

Although a 12-16 passenger robovan would hold about 4x fewer people than a bus, the Loop system more than compensates for this by winning on headway by at least an order of magnitude—if not two orders of magnitude—relative to actual real-world bus lines than usually have at least a few minutes between busses.

 

[Dan Detweiler]

So you have to wake up earlier for work because of the longer trip and then try to fall asleep again in the car? No thanks. Just because someone else is driving me doesn't mean I want the trip to take longer.

It is truly amazing that we ALL so blatently and unapologetically chose to ignore the law in regards to speed limits. I'm no different, it is just amazing, when you think about it, that this has been allowed to become the accepted norm. Can you thimk of any oyher law that is so universally ignored?

Dan

 

[philw1776]

Jaywalking

 

[Webeevdrivers]

Wrong. You are badly misinterpreting.

FSD + Humans is safer than just humans. FSD is atleast 100x less safe than humans currently.

ps : My last comment on this for today. Anyone who is wondering what the actual status of FSD is currently - pls visit and read who are into the weeds in the autopilot sub-forum.

Absolutely. FSD is getting pretty good. But again yesterday after a pretty successful drive it failed to yield to a van in a traffic circle and without intervention would have resulted in a collision. A couple days ago It also entered an intersection after the light was red to make a left turn. That would have been an expensive ticket for sure.

I enjoy FSD but at the present rate if improvement I would think we are a decade away from anyone certifying it for autonomous use. I mean, it’s not even close.

Jmho.

 

[Krugerrand]

I have about 35,000 miles of road trips and I find it far better than dumb cruise control. It’s been a long time since I had phantom braking on the highway. YMMV
City driving is another story.

There’s definitely a wide range of experiences that people are having. And while the set-up of each car may slightly differ/different software versions installed thus causing some of the issues, I question the reality of *some* people’s experiences as factual or dramatized. Yeah, I do - given people’s propensities.

I’ve been a passenger in cars where the drivers are stomp the accelerator and stomp the brake, and brake late, and accelerator early, and are entirely unable to be consistent with the accelerator giving me motion sickness when I’m not prone to it etc….and they think they’re *good* drivers. My car on TACC or FSD Beta has never even come close to replicating those kinds of drivers at anytime. Ever.

Just used TACC on a new route last week - car has never driven on those roads that included some city, some single lane highway, some mountain pass. Car behaved very well. A few ‘soft’ brakings mostly heading into tight turns, which were warranted.

‍ Car works quite well for me, but then I’m not expecting it to make toast - yet.

 

[Singuy]

So you have to wake up earlier for work because of the longer trip and then try to fall asleep again in the car? No thanks. Just because someone else is driving me doesn't mean I want the trip to take longer.

This is just the beginning. As Tesla reiterates with future updates the speed at which you get to work gets faster and faster while keeping the same safety risk. Tho lots of people has zero problem spending 30 mins on the road reading a book vs 20 mins driving themselves.

I don't know why everyone feels like the car must be able to do some speed racer Kung fu super human type driving before its feasible. That's like saying electric cars shouldn't exist until a 450 mile car exist.

Yes, the 35% slower robotaxi is not for everyone, until one day it is for everyone, just like a 60 mile EV1 was not for everyone, but it have a demand from a bunch of people, but it has to start somewhere.

 

[Xepa777]

Absolutely. FSD is getting pretty good. But again yesterday after a pretty successful drive it failed to yield to a van in a traffic circle and without intervention would have resulted in a collision. A couple days ago It also entered an intersection after the light was red to make a left turn. That would have been an expensive ticket for sure.

I enjoy FSD but at the present rate if improvement I would think we are a decade away from anyone certifying it for autonomous use. I mean, it’s not even close.

Jmho.

Yah that’s why I hate comments like these from shameless social media influencer shills

https://twitter.com/x/status/1636393138066423808

https://twitter.com/x/status/1636134823143288832

The moment any tesla influencer actually drives FSD they quickly change their tune (Gali, Rob).

 

[Nocturnal]

This is just the beginning. As Tesla reiterates with future updates the speed at which you get to work gets faster and faster while keeping the same safety risk.

I don't know why everyone feels like the car must be able to do some speed racer Kung fu super human type driving before its feasible. That's like saying electric cars shouldn't exist until a 450 mile car exist.

Yes, the 35% slower robotaxi is not for everyone, until one day it is for everyone, just like a 60 mile EV1 was not for everyone, but it have a demand from a bunch of people, but it has to start somewhere.

Also worth pointing out that speeding only really improves ETA on long trips with open roads. Most of our driving is around town and in traffic so going 80 in a 65 for 1/10th of the trip isn't going to make much difference vs doing 65 the whole way.

 

[Rashomon]

High-density passenger transport was discussed in Master Plan Part Deux. As far as we know this is still the plan.



Reasoning by analogy to assume that diesel to electric transition for busses will be a one-for-one substitution will lead to wrong conclusions. Unfortunately, this is exactly what most people are doing, including transit planners and policymakers.

The major portion of the cost structure of traditional busses is the driver compensation and the operation and maintenance costs of the diesel powertrain, both of which favor having a smaller number of larger busses in order to exploit economies of scale. Autonomous electric busses, with zero driver expenses and perhaps 75% lower O&M costs would change this calculus in favor of smaller minibusses/vans with:

• higher percentage of seats actually occupied
• no center aisle wasting space and materials
• greater number of stops/stations in the service network yet more express-like service
• coverage further out away from the urban core deeper into the suburbs

Boring company is a huge wildcard in predicting where this is headed. If you believe it will succeed and scale quickly, then “public mass transit” might actually change to being mostly single or dual passenger minicars with some mix of larger sizes up to perhaps 16 passengers per vehicle max. More than 16 is probably unnecessary for a major route served by one or more Loop trunk lines, because that’s already sufficient to beat every train line in existence, and that’s not hyperbole.

In 2019, Elon mentioned a 1-second headway as being feasible, which would imply 57k passengers per hour per lane with a 16-pax van. I think 1-second is plausible with autonomous operation and the ridiculously tight control and security of the tunnels. Humans today typically do about 2 or 2.5 seconds between vehicles on highways. Autonomy eliminates about 1 to 1.5 seconds of human reaction time. We can debate a few tenths of a second here and there, but even with 1.5 second headway the numbers still look great.

View attachment 918536

Here’s some rough estimates I’ve done of potential AEV throughput in a Loop compared with high-volume train or bus lines. To a close approximation, Passengers/hour = Passengers/vehicle * Vehicles/hour = Pax/veh * (3600 sec/hr) / (Sec of headway between vehicles)

View attachment 918537

Although a 12-16 passenger robovan would hold about 4x fewer people than a bus, the Loop system more than compensates for this by winning on headway by at least an order of magnitude—if not two orders of magnitude—relative to actual real-world bus lines than usually have at least a few minutes between busses.

I've argued this thesis for the last 5 years: electrification and autonomy will most effect bus transit, pushing buses to smaller sizes, more frequency, and more use. There are a lot of hurdles for this happen, not the least being the mostly highly unionized and bureaucratic public bus operators. Additionally, there are issues of passenger safety with only remote supervision, and boarding and onboarding issues, particularly with handicapped passengers. These will almost all be overcome eventually, but almost certainly not quickly, and rapid bus transit with dedicated lanes will become much more common as many cities in Europe, and eventually the US, deprecate automobile transit in urban areas. The urban transit future is a combination of many modalities, with automobiles taking a smaller share as other modes increase, from pedestrian to micromobility to rail and bus transit to even eVTOLs. As Musk has noted, urban transport is a geometry problem, and automobiles are the least geometrically efficient alternative. (On throughput per equivalent lane, automobiles lose by factors of 3 to 20 to other modes). The Boring Company, of course, is trying to change the geometry problem, but that's a project of decades.

 

[Krugerrand]

That was rather more complicated and ended out putting Aaron Burr and Alexander Hamilton on opposite sides. Before Jefferson made his statement, in 1799 Aaron Burr founded the Manhattan Company, ostensibly to supply water to New-York as it was styled then. He cleverly specified in the chapter that the company could take deposits, thus forming The Bank of the Manhattan company, which after countless mergers led to JPMorganChase of today.
The famous duel between Burr and Hamilton was mostly about divergent views on banking, although few accounts today describe that role. Jefferson was almost immediately drawn into the battle, sided with Hamilton, and so it went... Burr charged with murder.

In comparison the battles regarding Tesla are quite mild. OTOH, stock manipulation was there even in 1799 and before. These issues seem lost in history but are still continuing today in different form. Specifically the conflict over short selling goes back centuries. Tesla, involved in revolutionary technology as it is, is the very symbol of disdain by the same forces that advocated for the Electoral College composition to protect slave states.

When we seem surprised about all the misrepresentation and lies we should recall that such acts were there in the formation of the US and directly part of the creation of the Constitution of 1789. I assert that nobody who actually studied US Constitutional history would be surprised by the present state of play, including protecting short sellers and option writers while defeating manufacturers right to protect their own products. Seriously, the roots of our present Tesla travails are all the way back to the 1790's. exacerbated by the Industrial Revolution leading to the Sherman Anti-Trust act of 1890, which led us sequentially to the current warranty and parts disputes.

In summary; people suck and have always sucked. The end.