Make your robotaxi predictions for the 8/8 reveal

So Elon says that Tesla will reveal a dedicated robotaxi vehicle on 8/8. What do you think we will see? Will it look like this concept art or something else?

I will say that while this concept drawing looks super cool, I am a bit skeptical if it is practical as a robotaxi. It looks to only have 2 seats which would be fine for 1-2 people who need a ride but would not work for more than 2 people. I feel like that would limit the robotaxis value for a lot of people. Also, it would likely need a steering wheel and pedals for regulatory reasons even if Tesla did achieve eyes-off capability.

So I think this is concept art for a hypothetical 2 seater, cheap Tesla, not a robotaxi.

Could the robotaxi look more like this concept art but smaller? It could look a bit more like say the Zoox vehicle or the Cruise Origin, more futuristic box like shape IMO and seat 5-6 people.

Or maybe the robotaxi will look more like the "model 2" concept:

Other questions:
- Will the robotaxis be available to own by individuals as a personal car or will it strictly be owned by Tesla and only used in a ride-hailing network?
- What will cost be?
- Will it have upgraded hardware? Radar? Lidar? additional compute?
- Will Elon reveal any details on how the ride-hailing network will work?

Thoughts? Let the fun speculation begin!

 

[aataskin]

I think this is essentially what Tesla is doing and will likely do for awhile. The robotaxi narrative is just hype to create positive marketing. Robotaxis sound more exciting than L2 point to point.

I wonder what they do after another two-three years down the road they throw in the towel on their Vision only attempt to get them to an L4/L5 (probably even an L3)? Do they ditch the robotaxi stuff and will be overtaken by the Merks and Bimmers as premium EVs with L3 or L4/L5? And won't be able to license their tech because, frankly, nobody will want to license an L2, everybody already has it. Or do they go back to radars/lidars etc like everybody else and play catch-up?

 

[JulienW]

Because it is? I'm still struggling to see the point of city L2. Doesn't free any time. You're still driving. Highways and traffic jams is another story. There L2 is great.

In my scenario IF Tesla prefects L4 robotaxites and puts the same software/hardware on our cars but keeps them L2 then we could in effect go point to point the same as if it's L4 and we were in siting in the backseat.

We would just have to be seated in the driver's seat and be legally in control.

 

[spacecoin]

We would just have to be seated in the driver's seat and be legally in control.

And not be allowed to do anything else but to vigilatly monitor the system... Again, value proposition?

 

[aataskin]

In my scenario IF Tesla prefects L4 robotaxites and puts the same software/hardware on our cars but keeps them L2 then we could in effect go point to point the same as if it's L4 and we were in siting in the backseat.

We would just have to be seated in the driver's seat and be legally in control.

So if we were drunk then police would still write us up, right? How much more would people then pay for such an L4 that is legally an L2?

And again, who will license such a solution from Tesla?

 

[spacecoin]

So if we were drunk then police would still write us up, right? How much more would people then pay for such an L4 that is legally an L2?

And again, who will license such a solution from Tesla?

I would pay nothing (at least not over 30 USD per month) for such a a system (including current FSDS). City L2 is of no to little value to me. It's fascinating/impressive technical work though.

 

[Daniel in SD]

The previous version was hydraulics based vs Optimus being electric motor based. The new Atlas appears to have switched to electric motors also, but appears to have far more complex articulation.

Depends on how much volume is required to reach the price point. But I think the challenge of reaching that price with a humanoid robot that can do parkour and flips is going to be much harder than one that can just walk regularly.

Why? Are powerful electric motors too expensive?

What Tesla is doing is called marketing and how successful companies are able to sell products in high volume vs companies stuck selling a product in very narrow niche applications.

Neither will be able to do any human job for years so they're both marketing. The only niche application that exists is as a toy.

 

[JulienW]

So if we were drunk then police would still write us up, right? How much more would people then pay for such an L4 that is legally an L2?

And again, who will license such a solution from Tesla?

Then you call a Tesla robotaxie (or now a Uber) and pick up your car the next day.

 

[DanCar]

From Dec 23, from Elon:
We’re working on a low cost EV that’ll be made in very high volume. We’re quite far advanced in that work. I review the production line plans for that every week. I think the revolution in manufacturing that’ll be represented by that car will blow people’s minds! It’s not like any car production line that anyone’s ever seen … It’s a level of production technology that is far in advance of any automotive plant on Earth … The first production line will be in Giga Texas.

https://twitter.com/x/status/1732124790327054586

 

[brkaus]

Spot is not useless, it can be used to go into dangerous areas, just like how police have been using specialized robots in the past, except it is even more flexible, given it has legs instead of tracks. But the question is if it is $75k useful when there is similar product that costs just $2700 (actually lower, gen 2 is even less expensive at $1600) that can be used for the same task? Does the extra capabilities justify the higher price tag?

And the price tag of the mass manufactured version makes it more viable to be used in less specialized tasks, like for example for patrolling areas, doing performances for entertainment, and for hobbyists to buy one unit themselves for fun.

I see a similar thing for Atlas (which old estimates are probably like $150k, now it's probably going to be more like $200k) vs the goal for Optimus (under half price of a car or under around $20-30k). Atlas certainly is far more capable in term of acrobatics and the articulation of the body, but is it going to be worth ~10x more? Of course, I'm not going to count any eggs before they hatch (neither Atlas nor Optimus are officially selling yet).

To me that sounds like the $40k cybertruck.

And yes, none of the robots have really hatched and we don't know what they are capable of doing in a productive (earn $$) sort of way.

 

[brkaus]

I wonder what they do after another two-three years down the road they throw in the towel on their Vision only attempt to get them to an L4/L5 (probably even an L3)? Do they ditch the robotaxi stuff and will be overtaken by the Merks and Bimmers as premium EVs with L3 or L4/L5? And won't be able to license their tech because, frankly, nobody will want to license an L2, everybody already has it. Or do they go back to radars/lidars etc like everybody else and play catch-up?

I'd think throw more horsepower and the computing and continue vision.

Or use their fleet and do hyper accurate mapping.

I do think they may need more cameras for more stereo vision.

 

[DrChaos]

I'd think throw more horsepower and the computing and continue vision.

Or use their fleet and do hyper accurate mapping.

I do think they may need more cameras for more stereo vision.

Short range lidars are getting cheaper too, but 360 stereo vision (plus IR) and more compute is a more guaranteed win, plus you could train to be reasonably robust for partial obstruction of some cameras.

 

[DrChaos]

Because it is? I'm still struggling to see the point of city L2. Doesn't free any time. You're still driving. Highways and traffic jams is another story. There L2 is great.

I've found some value in the new FSD city. If I'm going to, or from, somewhere unfamiliar at night, I might not be able to see the appropriate route or turns soon enough and the FSD system tends to be better than I am if it's dark and I'm tired. Also it lets me look further ahead while it takes care of following the car in front. Prior to 12.3 the performance of the system was poor enough that I wouldn't be able to let it drive without very continuous monitoring of everything, so it was not valuable.

Of course when I know the routes as the system can fail to get into the proper lanes soon enough vs experienced human who also understands typical traffic patterns at certain times of day.

 

[JulienW]

Posted earlier about if same hardware/software on robotaxie Tesla would keep our cars L2. Another angle that may be even more plausible:

The robotaxie will have a different hardware/software suite than the consumer cars. There is circumstantial proof this will happen.

Back when Tesla touted the Robotaxie app they said anyone who leased a Model 3 couldn’t buy it because Tesla was going to take it back and use as a robotaxie. Well now there are 100,000s of Model 3/Y for sale at a low price and leases turned in every day that Tesla resales. So why doesn’t Tesla just keep and buy these cars and turn them into robotaxies? So no need to waste time developing/building a detected robotaxie model since all the robotaxies Tesla could need are available. There MUST be a reason and that is likely the new Robotaxie will have different hardware.

 

[brkaus]

Short range lidars are getting cheaper too, but 360 stereo vision (plus IR) and more compute is a more guaranteed win, plus you could train to be reasonably robust for partial obstruction of some cameras.

True. Maybe add some future low cost lidar for up close accurate maneuvering.

 

[ThomasD]

Would Doppler Radar help?

 

[brkaus]

Would Doppler Radar help?

For the wipers?

 

[HitchHiker71]

I get they are a much lower density than Uber drivers but all I know is that every time I’m out in the avenues at my favorite shabu-shabu restaurant I see them all over the place…and I haven’t seen even one driverless Tesla.

Precisely, I made the exact same observation earlier in this thread. All of the Tesla robotaxi talk is just that, talk. Because there is not one single driverless Tesla vehicle certified to be on the roads today. Zero, not one. It's all theory, yet to be proven. Will we see a true driverless Tesla on the roads within the next year even? I doubt it. Manned robotaxis? Likely. But driverless? It's anyone's bet as to how long it will take for this to become a reality. It's almost laughable seeing the fanboy base on X with various posts stating that Tesla has solved autonomy. We used FSD(s) 12.3.4 for a few local errands yesterday evening. Not a single one was intervention free in suburbia where we live. On the last leg back home, we had seven, count em, seven interventions, half of which were due to the lane happy changes for no reason, including two where it outright chose a turn lane when we were going straight - all over a mere 12 mile route. FSD(s) is certainly better than 11.4.9 ever was, and it's certainly headed in the right direction, but it still has a ways to go before any "driverless" vehicle becomes a reality. Granted, my observations are anecdotal, but when we consider that, unless geofencing is utilized, automony means 99.99% of all edge cases are solved, for everyone, including anecdotal cases, we simply aren't anywhere close to this reality just yet.

 

[DrChaos]

unless geofencing is utilized, automony means 99.99% of all edge cases are solved - and we aren't anywhere close to this just yet.

Its completely foolish financially to drop the Model 2/C for vague robotaxis. The only buyer of robotaxis is a robotaxi service and as you have found out, those are years away from deployment even once the hardware and software works reasonably. They would geofence as there would be regulatory problems and need for human quality control on the routing and performance.

So the market for those taxis will be small for many, many years even given FSD performance well beyond what's accomplished now.

This is Elon trying to pump the stock with hype and panicking about a cyclical downturn.

 

[HitchHiker71]

Its completely foolish financially to drop the Model 2/C for vague robotaxis. The only buyer of robotaxis is a robotaxi service and as you have found out, those are years away from deployment even once the hardware and software works reasonably. They would geofence as there would be regulatory problems and need for human quality control on the routing and performance.

So the market for those taxis will be small for many, many years even given FSD performance well beyond what's accomplished now.

This is Elon trying to pump the stock with hype and panicking about a cyclical downturn.

Agreed - given Tesla is a $100bb+ company - and supposedly has some of the best engineering talent on the planet (so they claim repeatedly), it would seem to be to be feasible to do more than one thing at a time - and do it well. There's no good reason not to do both/and. Overall, as is typical of Musk, if he pushes off the M2 next week - it's likely not because of robotaxi - it's because of other constraints - that he doesn't want to discuss openly and honestly - such as 4680 production ramp delays - energy density issues - or perhaps the unboxed process isn't realizing the savings they had hoped and promised - ori is to difficult to scale. Basically, it's not as a simple as just going "balls to the wall" with robotaxi. That's probably just a bright shiny object to distract from the real constraints and reasons for delaying the mass market EV platform. We will hopefully find out more info on next week's investors call.

 

[DrChaos]

Agreed - given Tesla is a $100bb+ company - and supposedly has some of the best engineering talent on the planet (so they claim repeatedly), it would seem to be to be feasible to do more than one thing at a time - and do it well. There's no good reason not to do both/and. Overall, as is typical of Musk, if he pushes off the M2 next week - it's likely not because of robotaxi - it's because of other constraints - that he doesn't want to discuss openly and honestly - such as 4680 production ramp delays - energy density issues - or perhaps the unboxed process isn't realizing the savings they had hoped and promised - ori is to difficult to scale. Basically, it's not as a simple as just going "balls to the wall" with robotaxi. That's probably just a bright shiny object to distract from the real constraints and reasons for delaying the mass market EV platform. We will hopefully find out more info on next week's investors call.

It's quite possible unfortunately that this is one of Elon's emotional ill-considered decisions resulting from seeming greater success from FSD. Once before he wanted to jump straight to no-driver no-wheel no-pedal robotaxis years ago and had to be talked out of it (Isaacson's book).

Today he's more addled and less disciplined than before.

Already the C/2 was likely going to use CATL LFP cells, not 4680s, and Chinese makers have figured out how to manufacture cheaply enough. This platform deserves balls to the wall, because the limiting factor on robotaxis is not sheer effort or capital expenditure, but actual scientific development and experience that cannot be hurried by substantially increased spending.

By contrast, sustained effort on known technology could get a C/2 platform ready.