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

[neroden]

You are reversing the meaning of nines. It is a guarantee of performance, not a guarantee of failure. 99.999% guarantees all but 5.26 minutes worth of a year's (continuous, non-stop) driving. Those remaining 5.26 minutes might be fine, but they are not guaranteed.

Flip it around, how many minutes is a typical driver "guaranteeing"? That is, what is the average failure rate for a human driver. If Tesla can really achieve five nines (which I doubt, except for limiting the circumstances) five nines is almost certainly better than any human driver.

I don't have the data to back it up, but if someone drove 60 hours a week, fifty weeks a year do you really think they would only make 1.8 minutes worth of mistakes? I think a reasonable person can accept that human drivers make errors far more frequently than that. The question is, when will Tesla be able to demonstrate 99.999% performance?

[edited to make more evident the part I was responding to in the quote]

I'vd said this before, but due to an extremely well-documented human bias, people will tolerate massively higher failure rates from humans than they will from robots. At least two orders of magnitude worse. Nobody really seems to know why, but it's a fact of human nature and we have to deal with it.

So, given two options:
-- A robot in charge, where 1 person dies for every 1,000,000 instances
-- A human in charge, where 100 people die for every 1,000,000 instances

People will collectively choose to have the human in charge. :shrug: It is what it is. Musk has made some statements indicating that he understands this.

My point is "better than a human driver" is irrelevant legislatively.

You can get around this psychology by only offering "driver assist" features, leaving the human clearly in charge.

 

[Lycanthrope]

You know, when everyone starts complaining about margins, production issues, delivery concerns, stock worries, etc, etc, etc, maybe we should remember that Tesla could have chosen to go THIS direction. Thank God they chose to do the difficult thing to bring this technology to the masses.

Dan

The $12.5 Million Bugatti La Voiture Noire Is the World's Most Expensive New Car

God, that's ugly... as an aside the following link popped-up at the end of the article:

 

[homer214]

Looks like maybe, just maybe this will be a Green Day. Methinks the shorties protest too much.

 

[GOVA]

How about some one (car maker) joining the supercharger network ?

Unfortunately, ALL have their heads stuck deep into their other ends.

 

[jerry33]

How about some one (car maker) joining the supercharger network ?

Joining the Tesla SC network implies they believe Tesla is better than whatever they have (or don't have). I doubt any car manufacturer will do that except as a last resort.

 

[AudubonB]

You know, after the announcement about them prepping to build Superchargers in Iceland, the more I think about it, the more I think we're about to become a Tesla Nation (sales are currently low because you can't service them here, there's no Supercharger network, and people have had to import their own) - potentially one of Tesla's highest market penetrations in the world. We have:

• Norway-like EV incentives (a gas car costs ~40% more than an EV of the same base price)
• High per-capita income (not as high as Norway, but still high on a global scale)
• The fastest "competing" chargers in the country are only 50kW.
• Low speed limits mean that the time between stops is so long that most people would want to stop anyway after going through a Tesla's range. E.g. no inconvenience.
  
• Surprisingly long travel distances (the Ring Road takes about 15 hours nonstop to drive all the way around)
• Everyone wants to at least be able to drive the Ring Road; it's sort of a minimum requirement for owning a car here. And why PHEVs have been much better sellers than BEVs thusfar.
• Long "side trips" away from the Ring Road (Snæfellsnes, Vestfirðir, Austfirðir, etc) call for long ranges
• Winters not super-cold, but long and snowy. Very windy. Mountainous. Having a good range-buffer for adverse conditions is important here.
• AWD is very popular here
• Lots of apartment-dwellers in-town with no access to charging (boo!); those who own cars want to be able to minimize how often they have to stop at a fast-charger in town (aka, range), and how long they have to be there (Superchargers). 50kW and Leaf-like ranges don't cut it for most people.

Things Tesla could do to significantly improve their appeal to Iceland:

• Tow package
• Air suspension (more clearance when needed)
• Greater native ground clearance (snow, unpaved roads)
• True offroad vehicles (not "city SUVs")

I didn't see on that list the answer to: "And your vehicular hydrocarbons come from....where?"

 

[ZachF]

 

[StarFoxisDown!]

Seems like the model y event combined with the model 3 35k announcement has really taken the hype out of the EV onslaught that was supposed to be the Geneva auto show. Audi won't even talk price on their Q4 EV prototype and it already has worse specs in both range and acceleration........0 to 60 - 6 seconds lol. I thought i misread that at first lol

 

[Cobbler]

Let's pump some of that NIO money back into $TSLA

 

[wayner]

It's not about the harm. It's about the technological credibility of Elon Musks claims wrt FSD. @Fact Checking defended him by saying that there is not a single claim of Elon about the capabilities of their hardware that does not eventually come true.

I doubt this is true. I don't know if you specifically call it hardware but Teslas, at least the Tesla Model S, were supposed to have third party apps. That never happened,.

 

[mongo]

I'vd said this before, but due to an extremely well-documented human bias, people will tolerate massively higher failure rates from humans than they will from robots. At least two orders of magnitude worse. Nobody really seems to know why, but it's a fact of human nature and we have to deal with it.

I do. humans are known fallible. Robots run code. If a robot kills someone due to an error, and that error can be corrected with code, then the robot should have had that code to begin with, and the death would not have happened. If it cannot be fixed with code, then fixed mechanically, if not mechanically, then fixed by not having the robot.

Robots run on physics, they don't 'just screw up'. So there is traceable/ correctable/ should have been done culpability (back to fallible humans).

 

[printf42]

OT

So yes, I concur, first evidence was that FSD could be accomplished with the processing power of H2, then it became apparent that H3 was required, so everyone that bought FSD gets H3.

I think HW3 is not really required, it’s just easier.
I remember before they reveal HW3, Elon said there are two ways to make FSD work, an easy way and a hard way.

Now I believe the hard way means to make it work on HW2, which is very much possible but means a lot of performance optimization to extract every last bit of performance out of it without sacrificing too much of accuracy.

And the easy way is HW3, where it could run the network that works as is, at the cost of a new computer board, they save significant unnecessary engineering efforts. And could redirect those resources to speed up the actual feature development.

HW3 is the way to go, don’t waste your time on getting 2x better runtime performance to save some money, instead, use that time to develop a new feature that provides 10x more value.

 

[lklundin]

Yeah, I also hope FactChecking is correct.

Not to put on any pressure, but if any of you guys are wrong, then a bunch of us here are screwed. Regardless of the disclaimers.

 

[jerry33]

I doubt this is true. I don't know if you specifically call it hardware but Teslas, at least the Tesla Model S, were supposed to have third party apps. That never happened,.

When it was announced Apps weren't as prone to malware as they are now. Apple and Google spend a bundle on keeping the bad apps out, it would very hard for Tesla to commit those kind of resources.

 

[SpaceCash]

I doubt this is true. I don't know if you specifically call it hardware but Teslas, at least the Tesla Model S, were supposed to have third party apps. That never happened,.

Also 30-second battery replacement instead of charging

 

[Artful Dodger]

I don't know, but I'd guess they're going to bring in $130 million+ from people buying AP and FSD now (that it has more functionality).

Agreed. This really is a wildcard for 2019Q1 revenue. Tesla has lots of unbooked revenue in the bank from past sales of FSD (for the features they haven't delivered yet). By moving the dividing line between EAP and FSD (ie: switching NoA from EAP to FSD), they can in effect unlock some of that revenue. Good move.

Together with banked ZEV credits, it could be enough to make Tesla show a 'tiny profit' in 2019Q1. In fact, I'm quite sure they will, and in all Quarters going forward. No Equity raise needed.

Cheers!

 

[RFernatt]

I'vd said this before, but due to an extremely well-documented human bias, people will tolerate massively higher failure rates from humans than they will from robots. At least two orders of magnitude worse. Nobody really seems to know why, but it's a fact of human nature and we have to deal with it.

So, given two options:
-- A robot in charge, where 1 person dies for every 1,000,000 instances
-- A human in charge, where 100 people die for every 1,000,000 instances

People will collectively choose to have the human in charge. :shrug: It is what it is. Musk has made some statements indicating that he understands this.

My point is "better than a human driver" is irrelevant legislatively.

You can get around this psychology by only offering "driver assist" features, leaving the human clearly in charge.

With the potential for great societal benefit, I believe the federal government should shield manufacturers that follow the established rules and own up to any problems promptly when there are issues. The feds do something similar with the Vaccine Injury Compensation Program as a "no fault system designed to compensate individuals, or families of individuals, who have been injured by childhood vaccines".

Vaccines offer such societal benefit that allowing companies that develop and manufacture such vaccines to be sued out of existence or shy away from developing useful vaccines because of fear of litigation would not be in the best interest of society. Do the life and cost savings to society from automated vehicles rise to that level?

BTW, I also think all pharmaceuticals should be handled in the same way. If a company follows the rules, gets FDA approval, and owns up to anything they discover after the fact quickly and do not willfully try to cover up issues, then I think the maximum awards should be severely capped (5x projected life earnings?). None of this billion dollar award garbage that just drives up the costs for everyone else. If they do try to cover it up or falsified data submitted to an authority, etc., then throw them to the wolves.

 

[jerry33]

Also 30-second battery replacement instead of charging

It worked but no one wanted to pay the price. It should still be an option for vehicles that need to run 24/7.

 

[Cartegena]

"Tesla.com gets 13M monthly website visits versus 1M visits for The international BMW Website | BMW.com , 3M for Mercedes-Benz Luxury Cars: Sedans, SUVs, Coupes and Wagons ,10M for New Cars, Trucks, SUVs & Hybrids | Toyota Official Site (still less). Clearly future of ordering cars is Tesla and it is online! "

Ray Dargham on Twitter

Ray may be rationalizing a bit to support his conclusion. Most serious U.S. car buyers visit the local dealers' websites in their metropolitan areas to shop prices with the internet sales person (after checking invoiced costs with Edmunds, KBB, etc. for the configurations they are considering.) Visits to the manufacturers' corporate sites are probably mostly to check more detailed information on the configurations' specifications and features.

I've visited multiple times 3 of the 4 Tesla's physical locations in our area (one a storefront in a multi-story mall, another a store in an upscale shopping district and the third a service center with a display/waiting area in the front--the fourth is a service center on the far other side of the metropolitan area) I've also visited a service center with display area in another large city. The two service centers were like small OEM dealers without much (if any) of a Parts Department (mainly apparel with logos) and no rows of late-model trade-ins. Most dealers around here out-source collision repair work, probably because of the vapor and liquid emission treatment systems and frame-bending equipment required. Tesla seems to be expanding it in-house collision repair capabilities.

There are trade-offs to both selling approaches, but mega, single brand dealership locations are long gone.

 

[Cobbler]

Audi won't even talk price on their Q4 EV prototype and it already has worse specs in both range and acceleration........0 to 60 - 6 seconds lol. I thought i misread that at first lol

I suspect the EV line from Audi is deceiving. They've invested a lot of money in a hydrogen platform together with Hyundai group.
Since the big air intakes in the front are still present (and bad for efficiency in battery powered vehicles), I suspect we'll see the same models with Hydrogen powered drivelines soon