Robotaxi Cybertruck fleet with Income Generation Purpose?

[Yakuza]

No, no. If you are going to become an AI fan you have to learn to speak the language. It is a year away. It has always been, and will be for the rest of our lifetimes, a year away. Go back (a few years at least) and review Elon's comments on this.

Have you seen any fsd beta videos? How could it be that far off given it can now drive fairly well in city streets?

Advertisement

 

[Stickboy46]

Have you seen any fsd beta videos? How could it be that far off given it can now drive fairly well in city streets?

Fairly well is good enough when you have a driver to intervene. Fairly well, isn't close to enough for robotaxi.

 

[ajdelange]

Have you seen any fsd beta videos?

Yes, I have and that strengthens my conviction.

How could it be that far off given it can now drive fairly well in city streets?

If you knew anything about AI and controls you wouldn't ask that question. The "within a year" thing is a sort of joke in the industry as we've been hearing that since the days of Norbert Wiener". See my response at
https://www.cybertruckownersclub.com/forum/threads/fsd-beta-videos.1857/#post-29090

 

[Doggiewumpus]

It takes 50% of the effort to get 90% of the features. The remaining 10% of features/reliability takes the remaining 50% of effort. Normal software can be "ready" at 90% feature set. This HAS to be 100% done. There is no edge case that can happen. The car can NEVER be confused for robotaxi.

FSD does not have to be able to handle all edge cases. It has to be able to handle enough of them that it drives better than a human driver. Admittedly, this will take a lot of work, but you don't have to hold it to a perfect standard.

Regulators are not against autonomy. On the contrary, they recognize that 40,000 people a year are killed in accidents (with countless others injured) and they understand the potential of autonomy to save lives. They also recognize that people (being irrational) will unduly weight the death of a person by an AV and ignore the deaths caused by human drivers. So they are ready to step in, as they have with other safety technology, and make it difficult or impossible to file lawsuits, and other ways of ensuring that the technology is not stymied.

Here's an interesting statistic that shows how irrational people are. When surveyed, people say they would not accept autonomous technology unless it is 10 times safer than human drivers. So if it saved 40,000 lives, but kills 5,000 they won't accept it. That kind of crazy "reasoning" is why regulators are already planning how to step in.

 

[Yakuza]

Fairly well is good enough when you have a driver to intervene. Fairly well, isn't close to enough for robotaxi.

We have a new very good baseline to improve from. Especially once it’s rolled out to more vehicles and they collect more data. To the other guy, I don’t care about sayings.

 

[Stickboy46]

FSD does not have to be able to handle all edge cases. It has to be able to handle enough of them that it drives better than a human driver. Admittedly, this will take a lot of work, but you don't have to hold it to a perfect standard.

Regulators are not against autonomy. On the contrary, they recognize that 40,000 people a year are killed in accidents (with countless others injured) and they understand the potential of autonomy to save lives. They also recognize that people (being irrational) will unduly weight the death of a person by an AV and ignore the deaths caused by human drivers. So they are ready to step in, as they have with other safety technology, and make it difficult or impossible to file lawsuits, and other ways of ensuring that the technology is not stymied.

Here's an interesting statistic that shows how irrational people are. When surveyed, people say they would not accept autonomous technology unless it is 10 times safer than human drivers. So if it saved 40,000 lives, but kills 5,000 they won't accept it. That kind of crazy "reasoning" is why regulators are already planning how to step in.

10000% disagree if you are talking about robotaxi. I'd agree if you are talking about FSD in the current state (with a required driver sitting behind the wheel). People are willing to accept that other people are not perfect .. i agree. They aren't willing to accept that a machine .. that someone is making money off of ... is not perfect and kills someone random.

Sorry, you aren't going to convince me that regulatory approvals will be granted to full scale national driverless cars (i.e. Robotaxi) unless they have a track record is damn near perfect. It's not going to happen.

 

[Stickboy46]

We have a new very good baseline to improve from. Especially once it’s rolled out to more vehicles and they collect more data. To the other guy, I don’t care about sayings.

Agreed .. Just saying the baseline to the level of required "perfection" required for regulatory approvals for robotaxi on a mass scale are HUGE and will not be closed within 2 years.

 

[Crissa]

No, no. If you are going to become an AI fan you have to learn to speak the language. It is a year away. It has always been, and will be for the rest of our lifetimes, a year away. Go back (a few years at least) and review Elon's comments on this.

Huh, it's down to a year now?

I know you're trying to pretend this is a Friedman Unit, but AI skeptics are often like the evolution deniers: Every new piece of evidence or ability just moves the goalpost. New transitional fossil? That's two mire missing links! Oh, you taught a computer to do something, that's not AI, it's just (insert name of new technology).

We have more and more capable autonomy. No one knows when it will cover all cases, but dang, it's covering alot more than it did last year. Or five years. Or ten. It's pretty close now.

-Crissa

 

[59sjordan]

Just for reference, I have one Dual Motor CT on order with FSD checked. I can't see how the FSD check mark is a problem at this point. I will evaluate the viability vs cost when I am required to actually front the money. In fact, I may even reserve a second CT with FSD if Tesla says they will allow the FSD to move with the driver.

What I don't hear people discussing for the whole Robo-Taxi debate is that Elon (Tesla) is planning on making smaller and cheaper cars. Wouldn't these actually be a much better solution for the Robo-Taxi fleet. Wouldn't these cars kind of make the CT a very expensive option? Although, possible a niche market for Robo-Taxi.

I was just wondering if anyone has put this piece of the puzzle into their thinking?

 

[Doggiewumpus]

What I don't hear people discussing for the whole Robo-Taxi debate is that Elan (Tesla) is planning on making smaller and cheaper cars. Wouldn't these actually be a much better solution for the Robo-Taxi fleet. Wouldn't these cars kind of make the CT a very expensive option? Although, possible a niche market for Robo-Taxi.

I was just wondering if anyone has put this piece of the puzzle into their thinking?

If there’s a party of 6 or someone who has cargo they’ll want to order a Cybertruck on their Robo app. And if they want a luxury experience. So if by expensive you mean expensive for the end customer then maybe, but if it’s a little more people will pay.

if you’re thinking that as a Tesla buyer you should wait for the cheap car, then no, because FSD will be really expensive then. And also, if robotaxis are a thing before the 25k car comes out, Tesla won’t sell it but will just churn them out to be used as robotaxis and owned by Tesla.

 

[Yakuza]

Just for reference, I have one Dual Motor CT on order with FSD checked. I can't see how the FSD check mark is a problem at this point. I will evaluate the viability vs cost when I am required to actually front the money. In fact, I may even reserve a second CT with FSD if Tesla says they will allow the FSD to move with the driver.

What I don't hear people discussing for the whole Robo-Taxi debate is that Elan (Tesla) is planning on making smaller and cheaper cars. Wouldn't these actually be a much better solution for the Robo-Taxi fleet. Wouldn't these cars kind of make the CT a very expensive option? Although, possible a niche market for Robo-Taxi.

I was just wondering if anyone has put this piece of the puzzle into their thinking?

Premium cost for CT rides + it’s just icing on the cake. If your truck can be making you money passively to get your monthly note to 0, that’s amazing.

 

[Doggiewumpus]

I had a weird thought. Bear with me here.
Let’s say robotaxis are approved by the time Tesla gets to building our Cybertrucks. At that point it will be cost effective for Tesla to stop selling cars and put them all on their robofleet. The sooner they do this the better because for s while they’ll have no competition. So they’ll want to ramp up fast. So if I do the math correctly, it would be cheaper for them to buy prospective owners out of their reservations and use those Cybertrucks for their robotaxi fleet. Elon said that FSD would be worth 100k so it would be worth it for them to pay owners the difference between the 7k that was in the reservation and the 100k that it’s worth then. If Tesla does this then they can pocket the 30k per year that the owner would have made and they can guarantee that the vehicle will be used 24/7 for robotaxing (thereby making their half of the profit full time). So will Tesla buy us out of our reservations for 93k?

 

[FutureBoy]

I had a weird thought. Bear with me here.
Let’s say robotaxis are approved by the time Tesla gets to building our Cybertrucks. At that point it will be cost effective for Tesla to stop selling cars and put them all on their robofleet. The sooner they do this the better because for s while they’ll have no competition. So they’ll want to ramp up fast. So if I do the math correctly, it would be cheaper for them to buy prospective owners out of their reservations and use those Cybertrucks for their robotaxi fleet. Elon said that FSD would be worth 100k so it would be worth it for them to pay owners the difference between the 7k that was in the reservation and the 100k that it’s worth then. If Tesla does this then they can pocket the 30k per year that the owner would have made and they can guarantee that the vehicle will be used 24/7 for robotaxing (thereby making their half of the profit full time). So will Tesla buy us out of our reservations for 93k?

The only reason that I can think for them to do this would be that the battery shortage issue was not any where near getting solved. In which case the CT would probably just get "delayed" and the 25k vehicle or more M3 or MY vehicles would get prioritized. There would be almost no reason to start using CT for robotaxis at the beginning. Robotaxis would not be going off-road. And if someone needs the extra storage, they can just go to u-haul.

 

[Doggiewumpus]

The only reason that I can think for them to do this would be that the battery shortage issue was not any where near getting solved. In which case the CT would probably just get "delayed" and the 25k vehicle or more M3 or MY vehicles would get prioritized. There would be almost no reason to start using CT for robotaxis at the beginning. Robotaxis would not be going off-road. And if someone needs the extra storage, they can just go to u-haul.

If Tesla is the only autonomous taxi service they’ll need all the vehicles they can get because demand will exceed supply.

 

[FutureBoy]

If Tesla is the only autonomous taxi service they’ll need all the vehicles they can get because demand will exceed supply.

Maybe that is why the Austin plant seems to be building out the buildings for the MY first. Get the MY line up and running for the robotaxis. CT's for robotaxis would just be a waste of batteries in the near term if the priority was robotaxis.