EVs: If you build them, will they sell?

[Goofyhoofy]

@buynholdisdead

If that was true than Tesla would no longer be gathering data.

No, they still pick up some edge cases, probably, but by now they’ve got 99% of everything, I’m sure.

But here’s why I say that the car needs to learn to drive, not just within a certain area. There’s a well traveled 2-lane road near my house. In the last few months it’s been down to one-lane for hours at a time. Power line tree clearing. Pothole fixing. Water main break. Private crew pouring a concrete driveway. Even Waze doesn’t know about these changes timely, which in any even only last a few hours. The dataset won’t include these, so the car needs to know what to do with a guy holding a “slow” sign, or a non-functioning traffic light, or a car that’s dead in the middle of the road, or …. It doesn’t matter if that’s in Nashville or Miami, these are edge cases that must be dealt with, not to mention the “light blue truck crossing against a light blue sky”.

@buynholdisdead

You assume because you have a steering wheel you can go around the branch? Exactly what a level 5 car would do only it would do it for you.

So the guy is chainsawing the branch. Does the car turn around, even if I don’t want it to? Or does it wait, even if it’s clear to me that it will take a half hour?

@jerryab2

anybody? It is already known, once Tesla has a documentable Level 5 AV, they will stop selling cars to the public.

Another of Musk’s predictions that will be famously and fabulously wrong. People will still buy cars for the same reasons they buy washing machines even if there’s a laundromat around the corner. Most people want to own. Not everybody, but I’ll give significant, gigantic, life changing odds on this bet.

@jerryab2

TaaS will become the de facto standard because it will be cheaper than buying a car (along with paying all the associated costs of owning and operating a car).

Oop, there it is again. I’ll make the odds even better if that’s possible. This is right up there with “vacation hotels will close because of: timeshares. Or AirBnB. Or something.”

(I’m trying to envision how this works. Musk just announces “I’m not selling cars to the general public anymore”, even though it will take tens-of-millions to replace private cars? Or does he sell some to people but start his taxi service at the same time?)

@jerryab2

TaaS will become the de facto standard because it will be cheaper than buying a car (along with paying all the associated costs of owning and operating a car).

I’m also sure this isn’t true. I drive a 26 year old Infiniti. Mrs Goofy has a 13 year old Infiniti. Our van is 17 years old. All have reached near zero terminal value, our only expenses are fuel, registration, insurance, and rare but occasional maintenance. A rental service isn’t going to get away with running 20 year old cars around, so they’re going to have significant depreciation in addition to other “business” expenses like billing, dispute resolution, insurance on newer cars (we have only liability) and more. It’s simply not possible that someone can match my transportation costs with any for-profit scheme. (They might win on fuel, but by the time this gets going I’ll have an EV too, powered by the solar on my house.)

[buynholdisdead]

@Goofyhoofy

So the guy is chainsawing the branch. Does the car turn around, even if I don’t want it to? Or does it wait, even if it’s clear to me that it will take a half hour?

“Alexa please stop the car.” “Alexa please find an alternate route”

Andy

[jerryab2]

@Tamhas

What about the word “geofenced” do you not understand?

I understand it very well. I have been following PRT for about 50 years now. Well-accepted travel routes will be geo-fenced because it would be irrational to NOT include them. Remember: These will NOT be “long distance” trips but rather relatively short ones that connect to another geo-fenced area. Defining “long distance” is impossible because a trip LA-SF would be fine. As would a trip Dallas-Houston and maybe also Austin. Chicago area would be fine. And so on.

NYC could possibly be connected to Long Island and points north (if adequate vehicle space, but getting on/off Manhattan is massively congested, even for AVs–if you go south or west (bridges/tunnels).

[jerryab2]

@Goofyhoofy

I’m trying to envision how this works.

Once Level 5 AV is achieved, then it is TaaS–not car sales. You pay a monthly fee to be driven anywhere within the capability of the vehicle. How it is done is up to the TaaS provider. Unlikely you would be allowed to keep/use the same vehicle all the time because that is not how TaaS works. It is a vehicle sharing network. The customers are NOT the owners leasing their vehicles to others because that leaves a LOT of money “on the table”–which Tesla won’t do.

[mschmit]

@jerryab2

Defining “long distance” is impossible because a trip LA-SF would be fine. As would a trip Dallas-Houston and maybe also Austin. Chicago area would be fine. And so on.

I recently did the (approx) SF to LA and back via Southwest. The Uber to/from the airports at each end cost more than the airfare. Doubtful many people will want to robotaxi this trip based on cost alone.

Mike

[jerryab2]

@mschmit

Doubtful many people will want to robotaxi this trip based on cost alone.

Maybe not the full distance is practical for a car, but going to a point on the route, such as a refueling station in a town, would likely also be reasonably geofenced.

[Goofyhoofy]

@buynholdisdead

Once airplanes became available how long did it take for people to adopt them for travel?

In any significant numbers? From 1903 to the 1950’s. If you want to get past the “already wealthy” then not until the 1980’s.

@buynholdisdead

Alexa please stop the car.” “Alexa please find an alternate route

Dang. Cell service is down again. Guess I’ll just sit here, since I don’t have a steering wheel.

@jerryab2

Once Level 5 AV is achieved, then it is TaaS–not car sales. You pay a monthly fee to be driven anywhere within the capability of the vehicle.

So Musk stops selling vehicles to the general public? But Ford, VW, and others keep selling them? I understand “??? as a Service” is a big deal now. With software for instance, I guess is it possible to buy software anymore? How about storage? Know anybody or any companies which still purchase storage and keep it themselves? With Accounting as a Service, are there still private accountants? Any big companies have their own in-house staff? (The answers: yes, yes, yes, and yes.)

@jerryab2

It is a vehicle sharing network

I prefer to have my own car. With my own sunglasses on the dash, and my mask and hand sanitizer in the console, and my own stations programmed on the audio. Crazy idea, I know.

[38Packard]

@bjurasz

Personally, I’m not holding my breath. And I’m not entirely sold on wanting the feature for me, honestly.

That’s a bummer. As a 66 year-old I look forward to (and truly hoping for) technology helping me as I grow even older

'38Packard

[buynholdisdead]

@Goofyhoofy

In any significant numbers? From 1903 to the 1950’s. If you want to get past the “already wealthy” then not until the 1980’s.

That is incorrect.

@Goofyhoofy

Dang. Cell service is down again. Guess I’ll just sit here, since I don’t have a steering wheel.

That is also incorrect, you do not need a connection to use Alexa.

[btresist]

@albaby1

But on the TaaS side, you have systems that are technologically able to operate autonomously within the geofence.

What is not being considered is that the “geofence strategy” of autonomous driving requires constant updating of the geofenced area. Neighborhoods and roads change. The Tesla self-driving strategy does not, as it is neighborhood independent. Tesla Level 5 will have a significant cost advantage over the Waymo-type Level 5, to the point where if both was available, I doubt the latter would be competitive.

A couple of other points.

Tesla’s FSD strategy is based on machine-learning, a method by which a program can rewrite itself based on new data. Tesla appears to have an enormous data advantage over its competitors, probably an order of magnitude difference. If the Tesla data-driven strategy is the correct one, then it will have the best self-driving system with the best safety record. Given the importance of safety when it comes to driving, I doubt that very many people will buy the second best software.

@albaby1

…hitting 20 million units with only three models depends on a sharp and dramatic change in consumer behavior.

Technological paradigm shifts are frequently associated with dramatic changes in consumer behavior. If self-driving becomes the norm along with autonomous taxis, then it becomes much more cost effective to own a model 3/Y type of EV and simply rent the larger vehicle types when needed. This becomes increasingly the case as birthrates decline. The WWII generation and Boomers needed station wagons, vans, and 6-passenger SUVs for their large families. The Millennials, not so much.

[mschmit]

@buynholdisdead

That is also incorrect, you do not need a connection to use Alexa.

But to use ANY voice navigation in a Tesla you do need a cell connection. Not only that if you are outside cell coverage (i.e. a National park) and you have planned for having no cell coverage and have pre-entered your navigational route you’ll keep your map…for a couple of stops but then they’ll disappear.

Mike

[Goofyhoofy]

@btresist

Technological paradigm shifts are frequently associated with dramatic changes in consumer behavior.

Probably not in the desire to use something different than the neighbor. Even when Henry Ford was making the Model T at a stunning low price, GM took the other half the market with a variety of styles that often cost twice as much. Steve Jobs found that he couldn’t just offer his “perfect size” phone, that people wanted variety.

There are now hundreds of brands and models of cars for sale, each with multiple configurable option packages - not to mention pick up trucks, vans, and so on. SUV’s, sedans, sports cars, rag tops, and from manufacturers in Korea, Japan, Sweden, France, Italy, England, Germany, and the US, as well as China and elsewhere. Suddenly everyone is going to settle on 3 or 4 makes and models? Seems incredibly unlikely.

@btresist

The Millennials, not so much.

How about families in the 1920’s? People want different . Our entire economy is based on a multiplicity of choice. Look at housing. You could bring down the cost of houses if you made them all the same. (Levittown did it for a while.) Vacuum cleaners. And yes, automobiles. Corporations may make rational, uniform choices, like Boeing v Airbus, but people want choice. Always have.

[Tamhas]

@jerryab2

Well-accepted travel routes will be geo-fenced because it would be irrational to NOT include them.

I still think you are massively oversimplifying the problem. What about doing the LA to SF and deciding it would be nice to detour to Carmel for the day. You have to do everything or have a fall back mode, which means no level 5.

[buynholdisdead]

@mschmit

But to use ANY voice navigation in a Tesla you do need a cell connection. Not only that if you are outside cell coverage (i.e. a National park) and you have planned for having no cell coverage and have pre-entered your navigational route you’ll keep your map…for a couple of stops but then they’ll disappear.

Mike that will easily be overcome with Starlink. Also I just went on a 3000 mile trip from Nevada into Montana through Idaho to Seattle Back down through Idaho to Nevada and didn’t have any problems with my navigation. Did it get spotty in areas, yes, but not to any degree where it was bothersome. Verizon all the way and not any problems to talk about.

Andy

[mschmit]

@buynholdisdead

Mike that will easily be overcome with Starlink.

Sure that is possible. But they would have to design a new antenna, of course. It doesn’t really fit into all the current models with a glass roof.
And do they (SpaceX) really want to do this? (Track and communicate with millions of cars) When they could be charging $100+ per month for others?

[rainphakir]

@mschmit

And do they (SpaceX) really want to do this? (Track and communicate with millions of cars) When they could be charging $100+ per month for others?

The rumor is that Elon wants X to be “The Everything” app.

Ie, X is the “umbrella” under which Starlink delivers data/media/communications/etc to and from Tesla vehicles… And homes, and RVs, and cell phones, and boats, and planes, and …

So. Yeah. I can see Tesla’s using Starlink.
Will new tech and new designs be required?
Sure. But isn’t that Elon/Tesla’s reputation?

Will Elon’s vision come to pass?
Only TPTB knows.

ralph

[MarkR]

@buynholdisdead

That is also incorrect, you do not need a connection to use Alexa.

My Alexa requires a connection to do ANYTHING at all.

[buynholdisdead]

@MarkR

My Alexa requires a connection to do ANYTHING at all.

That’s called an Alex. You need to upgrade.

Andy

[btresist]

@Goofyhoofy

Steve Jobs found that he couldn’t just offer his “perfect size” phone, that people wanted variety.

Apple currently produces all of five hardware iphone models. Even if you count the variants within each model class the number of iphone choices is 10 at most. Yet the iphone accounts for over half of smart phone sales in the US.

Products that are software dependent generally have fewer model options as differences in functionality are driven more by software than hardware. Car companies will move in the same direction as software becomes a bigger part of automobile functionality. Tesla dominates US EV sales with just four models. People want functionality more than variety.

@Goofyhoofy

and from manufacturers in Korea, Japan, Sweden, France, Italy, England, Germany, and the US, as well as China and elsewhere. Suddenly everyone is going to settle on 3 or 4 makes and models? Seems incredibly unlikely.

There will still be car manufacturers from different companies. But like most everything else involving computers, these companies will be offering a far smaller model lineup than we see traditional OEMs selling today. That’s because hardware variation becomes less important when a computer is the focal point of the product. There are lots of sizes and shapes for regular vacuum cleaners. In contrast, there are relatively few distinctive hardware differences between computerized robotic vacuums. The primary difference between makes and models is the software.

When cars become robotic there will be little need for hardware differences in performance or safety. If you aren’t driving do you really care about acceleration and handling? The need for different models is much reduced.

@Goofyhoofy

…but people want choice. Always have.

Sure, but people will tend to choose the product that provides the most bang for the buck. That’s why Americans who want EVs overwhelmingly choose Tesla’s despite the company only offering four models. Variety isn’t necessary if one has a better product.

[ItsGoingUp]

@Goofyhoofy

Last year Tesla sold 1,300,000 cars. That would mean in

2024: 1,950,000 and in…

@XMFRob

If you were to hunt for the specific guidance/goal, Musk was clear that this was a long term average which might be exceeded or fall short in any year to year comparison. I forget when it was said, but that timing is important to determine any projected future path. I know they grossly exceeded that guide in the past but can’t say (without digging up the info) where 2023 will fall on that.

The specific guidance Rob mentions is from 2020. Here’s an article that talks about it.

Tesla To Increase EV Sales By 50% A Year On Average, More In 2021

The relevant quote is:

“We are planning to grow our manufacturing capacity as quickly as possible. Over a multi-year horizon, we expect to achieve 50% average annual growth in vehicle deliveries. In some years we may grow faster, which we expect to be the case in 2021. The rate of growth will depend on our equipment capacity, operational efficiency and capacity and stability of the supply chain.”

In 2020, Tesla delivered 499,647 vehicles. For simplicity, let’s call that 500,000. So, 50% average growth is:

2021: 750,000 (936,222 actual)
2022: 1,125,000 (1,313,851 actual)
2023: 1,687,500 (889,015 actual first half)
2024: 2,532,250
2025: 3,796,875
2026: 5,695,313
2027: 8,542,969
2028: 12,814,453
2029: 19,221,680
2030: 28,832,520

Goofy gets much lower because not only is his starting number off, but his 2024 prediction should actually be 2023. As can be seen, in these first couple of years Tesla is well ahead of their predicted numbers. We shall see whether than continues. But it’s worth pointing out that Tesla has been growing its annual deliveries by 50% yearly since 2012, so it’s not as though they don’t know how to do this. And the past three years that growth has been despite the COVID related supply chain issues.

@albaby1

We’ll find out in 2025, when Tesla has to sell more than two million of a single model in one year.

Despite this constant drumbeat from albaby of too few models, the only thing that has proven a real challenge so far is getting enough batteries. Making the cars more affordable has not so far been an issue, nor has homogeneity. The price of Tesla vehicles will continue to go down, the choices will continue to go up (CyberTruck, “Model 2” platform).

I believe that unless Tesla runs out of raw materials they will continue to grow at the rate they predicted. But there’s no doubt it will be a challenge.

@Goofyhoofy

The fact that Musk is (allegedly) talking about licensing the FSD tells me that he recognizes that it’s a perishable advantage

Musk has quite clearly said that he expects Tesla to be the first to solve FSD, but that within a short time thereafter everybody would have it. He’s been saying that for years now. It’s only recently that he has mentioned possibly licensing it, but I imagine that has been a possibility all along. But it’s all meaningless until it actually works.

Of course, so far pretty much everything Musk has said about FSD has proven to be wrong, or “overly optimistic” if you prefer. Me, I’ve bought it three times since 2017, and almost six years later it isn’t even close to what Musk predicted for the end of 2016. It’s good enough that I would hate to be without it, especially for long trips, and I use it all the time, but it’s not what was predicted. My older son, who has said he would not buy a car until it could drive him around, is still carless (and doesn’t drive). So to him it continues to be worthless.

-IGU-