It seems unlikely to me as well. While Uber can draw on anyone in a city that has a car and wants to make a little money, a “Tesla Network” can only draw on Tesla owners who are willing to let strangers ride in their car and who are willing to pick up the grunt work of cleaning out other people’s trash from their vehicle every service day for a moderate amount of money. Most Tesla’s are expensive, luxury vehicles - a Model Y with FSD is a $60K+ car. Neither the cars nor their owners are going to be especially well matched to this.
I mean…maybe? Sure, easy to do in theory - but is it easy to do well enough to take on Uber? I mean, the app is probably simple enough to code - but Tesla’s going to be way behind everyone else in terms of the algorithms under the hood.
Uber’s been doing this for more than a decade. They’ve got data on ridership patterns and fleet usage and pick-up/drop-off times and pricing in every major city. To say nothing of the brand recognition and the fact that the Uber app is on almost everyone’s phone already from years of marketing and expansion.
There’s probably a pretty good reason why Waymo has partnered with Uber, rather than trying to build their own service from the ground up. It’s not like Alphabet/Google couldn’t figure out how to code a ride-sharing app.
But perhaps the biggest reason of all: ride-sharing is a customer services business. It’s not a tech product, though that’s an important input. You’re not just programming software or engineering a product - you’re providing a service to people. You have to have the infrastructure to manage complaints, deal with unhappy customers, maintain and clean vehicles, manage relationships with local regulators (traffic and otherwise), handle people who have lost or forgotten belongings, etc.. You’re running a consumer facing business, not just selling a product or offering software.
I think Waymo has decided that since Uber’s had about 15 years learning exactly how to efficiently do all that, it’s better not to reinvent the wheel and instead focus on the self-driving software and tech. Tesla’s choosing instead to take on Uber on that customer service side. I’m not sure that’s a choice that really fits with Tesla’s core competencies, but we’ll find out…
True, but within a year Tesla (or someone else) could duplicate 99% of that data. I have a couple ideas why Google decided to partner even though they could obviously write the code and certainly have the servers to host and store the data - and it’s that they want to avoid the Betamax/VHS/DVD/BlueRay/LaserDisc multiple standards scenario (*most recently visited upon EV charging plugs).
Instead of convincing everyone to download the special Google app along with the Robotaxi app and the Waymo app, they’d like to force it all into one neat package where you get anything from anyone. There will be a back room price to pay, obviously, depending on which carrier gets the fare but it would simplify thing tremendously for the consumer and help the industry grow faster.
It wouldn’t be the first time a company decided to do something better for the industry and the customer than for themselves, but…
..mmmmm OK, I can’t think of another right now but there’s probably one or two that have happened in the long history of industry.
Elon just told today on TV, ramp up will happen by late 2027… the past is, the timeline always slips. So, why are you assuming within a year, millions??
I get that you are bullish on TESLA, but have realistic assumptions and a dose of skepticism.
I don’t think so. Uber’s got a massive amount of data - at least a decade’s worth with the largest number of drivers in virtually every city in the U.S. (and globally). Tesla’s got few ways to generate that data until they start operating a ride-share service…and it will take time for them to scale up that ride-share service. Uber’s almost certainly got a really good “road map” for how ride-share behaves: what demand in Atlanta on a weekday afternoon at 3:30 p.m. looks like, where vehicles need to be to keep wait times down, what type of rider patterns are out there, how long a typical PU/DO interaction takes, what percentage of passengers leave objects behind, etc. A rideshare algorithm has to “solve” for a lot of different variables to keep service running smoothly - and you need data to make that experience smooth.
It’s going to take Tesla (or any competitor) far more than a year to compile that kind of data. Which gives Uber and Lyft a leg up on both maximizing their efficiency and UX. That’s not going to matter much during the pilot phase of their robotaxis (if you’ve got 10-20 cars serving an invite only list of screened customers, this isn’t really relevant). But if they actually want to start scaling up in the next few years, they’re going to be hampered by not having a full partnership with one of the incumbents.
As pointed out upthread, that’s inconsistent with Tesla’s stated timeframes. Tesla they’ve indicated that they’re targeting end of 2026 as when volume production of the Cybercab will begin. If they actually hit that date, it would not be until 2028 at the earliest before they had millions on the road. And while Tesla is good at many things, one thing they are not good at is hitting their announced time frames.
No, you don’t. I mean, by definition - you can’t “drive one daily” if the car is driving itself. You’re still driving it, because you drive a car with an advanced driver assistance system. It would be both illegal and extremely foolhardy if you let the car drive itself, rather than maintaining full attention and control of the vehicle and only letting the ADAS assist you with driving (though not legal advice).
That has nothing to do with it. You know yourself that your car isn’t self-driving, which is why your initial comment was that you “drive one,” rather than that you have a car that drives you.
Tesla’s FSD is not yet capable of driving itself. It cannot be safely operated without a human being attentively sitting behind the wheel performing part of the driving function - at best actively monitoring and being fully apprised of all relevant circumstances so that they can take over all functions from the car on a second’s notice, and at worst driving for the car because it has already disengaged.
You know that. You know that Tesla has never said otherwise. You know that whenever Tesla has made a statement about FSD’s capabilities in a context where it was a legally binding representation, they have pretty much said what I’ve said.
The fact that there are moments where it looks like the car is capable of driving itself does not mean the car is yet capable of driving itself - because to be materially safer than a human, it would have to be able to do that for years without disengaging. Not mere minutes or hours. And it’s nowhere close to that yet.
Not a chance. Not even a small chance. There won’t even be 100,000 within a year, not even 10,000.
We own a HW3 Tesla and a HW4 Tesla, and FSD is truly amazing. I use it all the time, even for 5 minute drives. But it still is neither ‘F’ nor ‘S’ yet. Small example, last week I visited friends of mine in a gated community (probably 1/3 of the communities around here are gated). My car drove me from my driveway (needed a small nudge on the accelerator because for some reason it thought the trashcan on the grass swale was “obstructing the path”) all the way to the entrance of their development, but it didn’t know how to stop at the guard gate to announce our arrival and to get the guard to open the gate. So I had to take over. I stopped the car at the guard, I told him who I am visiting, he checked his computer, saw my name there, opened the gate. Then I drove through into the development, and engaged FSD again. It brought me to the front of their house, then I took over again and pulled my car into their empty driveway and parked it there.
So, 1/3 or more of the developments here are gated. And trash cans appear on the swale 4 days a week (our days are Monday and Thursday). How will robotaxi handle all that? The answer is that FSD is not robotaxi. FSD and robotaxi are two different things right now, and will likely remain two different things for quite some time. One of the huge differences is that robotaxi software will have a constant connection to some control center somewhere such that if the car ends up in a scenario that it can’t handle (“red hands” for example), it’ll request help from the control center. Then a human (or maybe a powerful AI someday) at the control center will “drive” the robotaxi out of that scenario until it can take over the driving itself again.
Finally, it is also possible that robotaxi will require a front bumper camera (right now only found in Cybertruck and in the new model Y has one). Otherwise it has no way to determine if a child’s toy or a small piece of hand luggage (or God forbid an infant in an infant carrier seat) was inadvertently left on the ground right in front of the front bumper. Ever see a crowded arrivals pickup area at a typical airport? Cars right up near each other in a row, and people walking between them every which way, carrying all their luggage, their kids, and other stuff. All while being somewhat discombobulated from the travel.
Then it should be easy for you to find something to back that up, rather than just hand-waving that you don’t agree with me.
Again, every formal statement by Tesla about the capabilities of FSD emphasize that it is not capable of driving on its own. Heck, that’s why they formally changed the name to FSD (Supervised). That’s why Musk has not publicly claimed that FSD has achieved full autonomy, but instead keeps making forward statements that it will come soon (even self-deprecatingly declining to repeat that it will come at the end of this year again). That’s why your owner’s manual contains the following statements:
Warning
Always remember that Full Self-Driving (Supervised) (also known as Autosteer on City Streets) does not make Model 3 autonomous and requires a fully attentive driver who is ready to take immediate action at all times.
Warning
Full Self-Driving (Supervised) is a hands-on feature. Keep your hands on the steering wheel at all times, be mindful of road conditions and surrounding traffic, and always be prepared to take immediate action. Failure to follow these instructions could cause damage, serious injury or death. It is your responsibility to familiarize yourself with the limitations of Full Self-Driving (Supervised) and the situations in which it may not work as expected.
(Well, if you have a Model 3. The one for the Model Y is the same, they just refer to the Model Y).
There’s no claim whatsoever by Tesla or Musk or anyone with the company that FSD is already autonomous.
Gretzky: “Skate to where the puck is going to be, not where it has been.”
When I was buying Tesla stock 6 years ago, many naysayers told me that Tesla was going bankrupt. The market cap then was $50B. EVs would never work. Where would they charge ? They cannot scale
Today, the market cap is $1.1 Trillion and Model Y is the best selling car on the planet. The naysayers are still there.
Great you believed in some thesis and it worked. Wonderful. But that doesn’t make you can claim that Tesla has FSD.
Our views are very clear, you think Tesla can make robotaxi work and that will be worth Trillions more. I am skeptical that Musk can deliver it on the timeline he mentioned, and there are lots of challenges from technology, regulatory, & operations.
No one will have much problem with your views, where the challenge is you are claiming there is FSD today, robotaxi’s tomorrow, million will be in road 2016, etc. There are no regulatory approval today and you expect there will be millions of robotaxi’s in 2016 doesn’t add up.
Only time can tell how it will play out. So, we can stay out of this conversation for another 2 years and come back and discuss…
