Once you go full retard at RH....you fall off the cliff.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
Or at least more than one car per family. This assumes that five minutes from placing the request to seeing the car is about the normal maximum. Anything longer than ten minutes makes for a very hard sell.
I expect a big jump when the software that is being rewritten now gets into widespread use and has a few months under it's belt. No idea if that will be this year though.
Knightshade
Well-Known Member
Tried to consolidate all replies to 1 post since the thread is already crazy big 
Do you have a source for the 22 mph being typical? I'd imagine it's even slower in say NYC, but it's be a LOT higher here in NC where I live (and pretty much the entire country outside dense urban cities if we're gonna pretend robotaxis will replace lots of "normal" cars too not just in dense cities).
It also appears to ignore that a good portion of the country, especially some of the larger denser cities in the US, are in the north where it's very cold a lot of the year- and efficiency to keep the cabin warm for riders causes a significant range hit (I suppose the Y will do better here than the 3 at least)
It also appears to assume all the RTs will be long-range vehicles... an SR+ already can't make it 264 miles.
Sure, as soon as someone invests the money to build those things.
Which I specifically mentioned as an option in my post.
In fact the option that probably makes the most sense (well, SOME charging stations with an attendent, be it SCs, urban SCs, or just regular 240 service of some kind).
This will be even more needed as you start trying to replace things like trucks, where someone calls an RT to transport a truck full of dirt or gravel or something for them... suddenly someone needs to clean the bed out.
I guess you also missed where I specifically mentioned none of these problems were show stoppers, they'd just require some $ outlay to fix because the infrastructure as Tesla built it for long-range road trips is vastly different from that needed for lots of short-range taxi trips, and even WITH that spend would still easily be a cash cow for the company.
Not really though.
Because a robotaxi can run 24/7. I know uber drivers that only drive a few hours a day. So One RT likely replaces like 3-5 rideshare hail drivers at least.
So cut that number down quite a bit for how many RTs are actually "needed" instead of uber/lyft cars.
(note too your source uses "probably" a lot in citing number of drivers- a quick search shows a pretty wide array of guesses at the real #)
FWIW I don't disagree at all Tesla is well ahead of anyone else trying the vision-only approach (and that that is the ideal approach)- I just disagree with the idea the current HW is sufficient to reach L5. (the fact they've already changed it multiple times since originally claiming it was sufficient seems pretty good evidence they don't actually know how much is sufficient until they get there.... see also Elon himself saying you shouldn't trust any prediction he makes about a thing he hasn't ever done before- and how many such predictions he's gotten wrong before)
Nope.
I grew up in NY.
I'm very very very familiar with dense urban situations and taxis in those environments.
The fact I'm in NC now gives me the other side of that perspective. (and I've lived in both "cities" in NC, and now "out in the country"). I've also spent time in quite a lot of major cities elsewhere in the US (and western Europe).
There's very very very little in the way of taxis here- because they don't make nearly the same economic sense.
There's SOME uber/lyft presence, not so much close to my house but maybe 15-20 minutes away you can reasonably find one within 10 minutes.... but again massively less than in dense cities. Mostly you need to own a car here.... whereas I think anyone owning a car in NYC is insane.
That said- there's still like 2 million NYC residents who own private cars.
Despite MASSIVE availability of taxis, uber, subways, buses, etc.
That's why I think the idea everyone dumps their cars when robotaxis get here doesn't hold up to the facts.
People like owning cars. Even in places it makes NO SENSE for them to do so.
See above. Owning a car in NYC is crazy expensive. Parking alone can be more than a car payment in lots of parts of the city.
Yet a couple million people still do it. Despite easy public transit and a very walkable city on top.
So the idea they'll all stop owning those cars once their taxi doesn't have a driver appears just wishful/magical thinking.
Even moreso the folks in less urban areas.
Again- if someone only wants to get from A to B and doesn't care about owning the vehicle or it being a specific vehicle- we'd see most cars being sold being $15,000 econoboxes.... instead of an average new car price in the US of around $35,000, and pickup trucks and SUVs being wildly popular.
See also all the private car owners in dense urban cities with excellent, cheap, public transit- why do they still own cars?
Why would RTs change their mind?
I agree RTs would likely replace SOME private vehicles. There's some % especially where when their existing car is on its way out, going full RT might be a reasonable replacement.
I just think some estimates of how many are wildly, wildly, optimistic. See again how many folks in places like NYC where owning a car makes 0 sense still own a car. Millions.
And increasingly so as you get outside dense urban cities.
So I agree with your points 1-3.... the question is- where I live for example I'm roughly 15-20 minutes from any retail other than a gas station type store.... I'm ~30 minutes to a couple of decent sized cities... and 45-60 minutes from several other major cities.
But the population density is so sparse there's no high speed internet available. (HURRY THE HELL UP STARLINK!)
I wouldn't ever consider replacing my car with a robotaxi unless 1-3 were all true.... not just at my house, but anywhere I'd be likely to travel TO so I can get back easily... but I don't see the economic model where someone buys a $40,000 robotaxi and then has it sit within 5 minutes of my house for the maybe couple of trips a day they'd ever get from me and my very sparse neighbors sitting there.
The fact it's CHEAP only works for the guy who paid for the RT if it's also making trips often.
Easy in LA, Chicago, Atlanta, etc downtown.
Less easy in the burbs.
Pretty hard in the "country"
Do you have a source for the 22 mph being typical? I'd imagine it's even slower in say NYC, but it's be a LOT higher here in NC where I live (and pretty much the entire country outside dense urban cities if we're gonna pretend robotaxis will replace lots of "normal" cars too not just in dense cities).
It also appears to ignore that a good portion of the country, especially some of the larger denser cities in the US, are in the north where it's very cold a lot of the year- and efficiency to keep the cabin warm for riders causes a significant range hit (I suppose the Y will do better here than the 3 at least)
It also appears to assume all the RTs will be long-range vehicles... an SR+ already can't make it 264 miles.
Sure, as soon as someone invests the money to build those things.
Which I specifically mentioned as an option in my post.
In fact the option that probably makes the most sense (well, SOME charging stations with an attendent, be it SCs, urban SCs, or just regular 240 service of some kind).
This will be even more needed as you start trying to replace things like trucks, where someone calls an RT to transport a truck full of dirt or gravel or something for them... suddenly someone needs to clean the bed out.
I guess you also missed where I specifically mentioned none of these problems were show stoppers, they'd just require some $ outlay to fix because the infrastructure as Tesla built it for long-range road trips is vastly different from that needed for lots of short-range taxi trips, and even WITH that spend would still easily be a cash cow for the company.
Not really though.
Because a robotaxi can run 24/7. I know uber drivers that only drive a few hours a day. So One RT likely replaces like 3-5 rideshare hail drivers at least.
So cut that number down quite a bit for how many RTs are actually "needed" instead of uber/lyft cars.
(note too your source uses "probably" a lot in citing number of drivers- a quick search shows a pretty wide array of guesses at the real #)
FWIW I don't disagree at all Tesla is well ahead of anyone else trying the vision-only approach (and that that is the ideal approach)- I just disagree with the idea the current HW is sufficient to reach L5. (the fact they've already changed it multiple times since originally claiming it was sufficient seems pretty good evidence they don't actually know how much is sufficient until they get there.... see also Elon himself saying you shouldn't trust any prediction he makes about a thing he hasn't ever done before- and how many such predictions he's gotten wrong before)
Nope.
I grew up in NY.
I'm very very very familiar with dense urban situations and taxis in those environments.
The fact I'm in NC now gives me the other side of that perspective. (and I've lived in both "cities" in NC, and now "out in the country"). I've also spent time in quite a lot of major cities elsewhere in the US (and western Europe).
There's very very very little in the way of taxis here- because they don't make nearly the same economic sense.
There's SOME uber/lyft presence, not so much close to my house but maybe 15-20 minutes away you can reasonably find one within 10 minutes.... but again massively less than in dense cities. Mostly you need to own a car here.... whereas I think anyone owning a car in NYC is insane.
That said- there's still like 2 million NYC residents who own private cars.
Despite MASSIVE availability of taxis, uber, subways, buses, etc.
That's why I think the idea everyone dumps their cars when robotaxis get here doesn't hold up to the facts.
People like owning cars. Even in places it makes NO SENSE for them to do so.
See above. Owning a car in NYC is crazy expensive. Parking alone can be more than a car payment in lots of parts of the city.
Yet a couple million people still do it. Despite easy public transit and a very walkable city on top.
So the idea they'll all stop owning those cars once their taxi doesn't have a driver appears just wishful/magical thinking.
Even moreso the folks in less urban areas.
Again- if someone only wants to get from A to B and doesn't care about owning the vehicle or it being a specific vehicle- we'd see most cars being sold being $15,000 econoboxes.... instead of an average new car price in the US of around $35,000, and pickup trucks and SUVs being wildly popular.
See also all the private car owners in dense urban cities with excellent, cheap, public transit- why do they still own cars?
Why would RTs change their mind?
I agree RTs would likely replace SOME private vehicles. There's some % especially where when their existing car is on its way out, going full RT might be a reasonable replacement.
I just think some estimates of how many are wildly, wildly, optimistic. See again how many folks in places like NYC where owning a car makes 0 sense still own a car. Millions.
And increasingly so as you get outside dense urban cities.
So I agree with your points 1-3.... the question is- where I live for example I'm roughly 15-20 minutes from any retail other than a gas station type store.... I'm ~30 minutes to a couple of decent sized cities... and 45-60 minutes from several other major cities.
But the population density is so sparse there's no high speed internet available. (HURRY THE HELL UP STARLINK!)
I wouldn't ever consider replacing my car with a robotaxi unless 1-3 were all true.... not just at my house, but anywhere I'd be likely to travel TO so I can get back easily... but I don't see the economic model where someone buys a $40,000 robotaxi and then has it sit within 5 minutes of my house for the maybe couple of trips a day they'd ever get from me and my very sparse neighbors sitting there.
The fact it's CHEAP only works for the guy who paid for the RT if it's also making trips often.
Easy in LA, Chicago, Atlanta, etc downtown.
Less easy in the burbs.
Pretty hard in the "country"
"Without that short hedge (and our relatively small Tesla short) "
Given that, his fund itself is quite small, and what does a 'relatively small' mean? I am guessing he is now short with around $10,000 or so.
His investors should be total fools, or he is being bankrolled by vested interests to keep the charade going. They probably believe his attacks on Tesla and Musk in social media and his media interviews and quotes he gets from main stream media with the title 'Short Seller", is worth the money he is losing.
Most likely the latter.
I keep hearing this name and i see his twitter profile says he is a Space photographer. I follow a lot of people in twitter and youtube and read up a lot of articles, but I haven't heard this name. Who is he and what is the important work he has done?
StealthP3D
Well-Known Member
It would be more than annoying, it would be illegal for funds to trade on material, non-public information, even if the information didn't come from Tesla. There's a reason they announce S&P inclusion to all market participants at the same time.
If as expected the cost of a ride drops to a dollar per mile or less then urban and even suburbanites will own fewer cars.
If the 2nd car is used for shopping, dr appointments, taking children to school etc. And instead of paying a monthly car payment and insurance payment they would spend less. $300 a month in car/insurance payments would equate to 300 miles a month. With no additional depreciation costs.
Talking about how many Uber/Lyft and taxi drivers is meaningless. As prices drop usage will increase.
Where will we find land for robo taxies wait? We will have empty parking lots because it is cheaper to take the robo taxi than drive.
I am wondering about the effect on commercial real estate as this additional land becomes empty and available for development.
If the 2nd car is used for shopping, dr appointments, taking children to school etc. And instead of paying a monthly car payment and insurance payment they would spend less. $300 a month in car/insurance payments would equate to 300 miles a month. With no additional depreciation costs.
Talking about how many Uber/Lyft and taxi drivers is meaningless. As prices drop usage will increase.
Where will we find land for robo taxies wait? We will have empty parking lots because it is cheaper to take the robo taxi than drive.
I am wondering about the effect on commercial real estate as this additional land becomes empty and available for development.
To make the donations list I am pretty sure your credit card has accepted the charge. To be spoofing Elon or anyone for ten grand would be a pretty expensive joke.
An amusing anecdote...
Oil giants' production cuts come to 1 million bpd as they post massive writedowns
Oil giants' production cuts come to 1 million bpd as they post massive writedowns
Robo Taxis - my foot. I am sorry I am not drinking that cool aid
Good question. All those upcoming major break through software that we keep hearing, will end up being another small improvement to an existing feature. I am sure we will get lot better in city driving than where we are today in the next few years, but automated driving without any driver presence? give me a break..
Knightshade
Well-Known Member
Part of the issue is a lot of folks are just guessing at a ton of factors in order to get to whatever conclusion they prefer.
$30,000 from a Tesla Robotaxi?…not as crazy as it sounds.
These guys do some math and find charging less than $1.11 per mile and a Tesla owner loses money on operating an RT.
This still allows RTs to majorly undercut driven-taxis and rideshares that run 2-3 bucks a mile...by more than enough to make RTs highly profitable in dense areas that current rideshares and taxis work in... but doesn't really get you to where we need millions of em in the US or they massively replace normally owned cars.
But upthread some folks insist costs will be like $0.10-$0.20 cents a mile and they'll replace tons of normal private cars.
Even Tesla on autonomy day cited $1/mile but then admitted it was an arbitrary number not based on much of anything.
Nobody really knows what the real cost will be... or what # of cars would be legit "replaced" with RTs in any given area for any given cost... Or even the best ways (or real costs) to address the various infrastructure issues RTs present (refueling, cleaning, idle areas, acceptably high availability in non-dense areas, etc...).... because it's a product that doesn't yet exist, and we don't really know what adoption rates would look like in different areas, and all the surrounding costs would build off of those.
I think Tesla even managing to offer L3 highway driving would make FSD a huge financial boon for the SP, as that's still way in advance of what anyone else offers- and seems a lot more in reach too...and doesn't really require any guesswork or making up a lot of #s regarding disrupting the entire transport system.
To me it makes sense to include that in thinking about SP going forward, as I think they're not especially far away from being able to do something like that...
L5 RTs? See the previous poster about the black swan thing and why heavily figuring that into any near/mid term SP calculations is...dubious.
StealthP3D
Well-Known Member
In The ARK analyst who models robo-taxi revenues uses an average of 22 mph over the duration of the day - it's actually a very optimistic assumption if you have ever used the "Avg. Speed" function of a GPS while doing any kind of intermittent driving in an urban area. To achieve 22 mph you need to spend a significant amount of time above 50 mph to make up for all the stopping. In Mid-town Manhattan you can only expect 4.7 mph average (and this doesn't include any time spent parked at the curb). I'm not sure why you doubt this - the actual number for a robo-taxi in actual service will probably be a lot lower than 22 mph average in urban areas, even including a mix of suburban driving. Every time they stop to let someone in or out really kills the average speed number. And a robo-taxi will not be equally busy throughout the day, there will be peak periods and slack periods. These slack periods must be averaged in to get an average that represents the total miles travelled during a "shift". Which is what the point of this "average speed" exercise was about.
A few years ago Uber released Uber Movement, a traffic app designed to track traffic speeds around the world. It utilizes a dataset (called "Speed") to derive all kinds of traffic movement metrics at all hours of the day in a number of cities around the world. The average speeds returned are very low and this doesn't even include the amount of time the vehicle is parked waiting for the busy periods.
Uber Movement: Let's find smarter ways forward, together.
When I used 22 mph as the average speed over 12 hours, I was being extremely generous - to question that is to show that you really have no idea about robo-taxis. I'm sure there will be some atypical robo-taxi usage that will exceed this number but we are talking averages. I think 22 mph may eventually be an achievable average if autonomy and transportation as a service significant reduces congestion and more people in out-lying areas adopt transportation as a service but in the first 5 years that won't be achievable as a fleet average.
One tool you can play with to get a better feel for this is the Avg. Speed on your car's trip meter. Keep in mind even this number will be very generous because it's only functioning when the car is in "Drive", as soon as you put it in park the "Avg. Speed" function stops averaging. We are interested in the average speed of the entire shift in order to derive the number of miles travelled. I have a couple of GPS that provide this number (doesn't stop working when the car is in "Park" and the results are shockingly low whenever the driving is even a little bit intermittent.
He lives in Brownsville and has taken a lot of photos of the SpaceX facilities and rockets. I dunno if it's that important, but he's pretty well-known in SpaceX fan-circles.
If you want a break from the robotaxi talks then here are some new vids to watch!
Jeff Roberts visited Giga Texas:
and so did Terafactory Texas:
The Wolfpack Berlin visited Grüneheide:
And Jason Yang visited Giga Shanghai:
Jeff Roberts visited Giga Texas:
and so did Terafactory Texas:
The Wolfpack Berlin visited Grüneheide:
And Jason Yang visited Giga Shanghai:
Knightshade
Well-Known Member
Sure.
In fact, I specifically already mentioned it's slower in places like NYC.
But MUCH faster in non-urban areas.
Hence I asked for the source on the 22 mph number.
Once again you appear to be repeating my own points as if you're correcting me. It's weird, since you seem to repeatedly do it too.
Doubt what? That traffic in NYC is slower than average? I told you that.
I questioned where one gets 22 mph as an average for taxis everywhere though.
Remember- many are suggesting RTs aren't just for dense urban downtowns, right?
If YOUR argument is RTs will dominate in dense urban areas- I've agreed with that from the start.
But there's no need for "millions" of them to accomplish that goal in the US.
Once again you're repeating back something I told you as if you came up with the idea.
Very odd method of discussion.
Right.
So where does 22 mph come from then?
You say ARK. How'd they come up with that #? Why are you confident it's correct?
Clearly in places like NYC or other dense urban areas, 22 is quite high.
Around where I live 22 would be quite low. Excluding stops you're almost never driving that slowly anywhere near my home. 95% of my drive to work or back is more like 70 mph.
So a robotaxi replacing my car will be averaging a lot more than 22 mph.
So again 22 is feeling like it's either VERY specific to a unique sub-set of places rather than nationwide... or simply a made up out of the air guess number like "what the real cost per mile for an RT will be"
Had high hopes when I saw this link initially... then not so much.
There's only 13 US cities on the list. Mostly the largest and densest population ones.
Where I already mentioned from the start speeds would be slow.
Even worse- when I actually clicked on most of those 13 I got "Speeds unavailable in this city"
Even the few that had it didn't seem to show any obvious (average speed of all trips)- just the average speed on a given road. Possibly a deeper dive into downloaded data would let you get an average based on # of trips and such- but even if it did this is only for a few, very dense, cities.
It doesn't get us to any 22 mph national average though.
And yet can't actually support your 22 mph number with real data.
Especially in the wide use case outside places like NYC.
Weird.
It's almost like NOBODY really has an idea about robotaxis with legit numbers.
Probably why even Tesla said on autonomy day they were just picking an arbitrary cost- since even THEY don't have real numbers to use.
We are. Nationally.
Hence why I pointed out it's gonna be a lot slower in NYC... but a LOT faster in say rural areas.... and asked if you had sources supporting the 22 mph claim nationally.
It doesn't appear you do.
Because the fleet will mostly be in dense urban areas.
Which is what I said in the first place...and why you don't need "millions" such cars because such areas don't need millions of them TODAY and one RT in such a place can replace 3-5 uber guys who don't run all day/night like an RT can.
Once again you appear to have put a lot of effort into reaching the same conclusion I originally did while trying to appear to be "correcting" the person who beat you to the conclusion.
Very very strange.
Yeah.
Mine is a lot higher than 22 mph since I don't live in a dense urban downtown.
95% of my driving is either 2 lane roads with a 55 mph speed limit, or 3-6 lane interstate roads with a 65-70 mph speed limit.
Other than slowing for an occasional intersection on the 2 lane roads, or pulling into my driveway or destination, I'm almost never doing 22 mph or slower.
Neither is anybody else around here.
Nobody ever gained anything by dreaming small.
I think most people are too optimistic about L5 FSD autonomy in the general case. Tesla and other companies can do it with geofencing, but I doubt the fundamental tech exists for the fully general case.
For example, those in software development for awhile may remember that voice recognition had accuracy of maybe 90% with *trained* data sets for decades before 2010. When NNs came along, accuracy jumped orders of magnitude on *untrained* data in just a couple years.
Sometimes the right approach is a process. There were email and document sharing systems for decades. After the internet came along with open RFCs, these took off like a rocket and use expanded 4 to 5 orders of magnitude.
The point is that you only know when you're going to solve really hard problems when you have a solution that scales quickly and works. Until then, the basic paradigm probably needs to be invented.
Maybe Tesla's rewrite will get there. But when EM says he can almost drive home to work without intervention, that doesn't sound optimistic.
I would not feel comfortable riding in a car twice daily if the driver had an accident once a year. Yet, that is 3 orders of magnitude improvement over FSD that has an intervention each trip. We need 4 or 5 orders of magnitude improvement which is 10 to 100 times better than 1000 times better than the current state.
We'll know FSD in the general case has a chance when interventions quickly go from 1 per trip to 1 in 10 trips in 6 month or a year. We've seen gradual improvements over the past couple years, but it's not nearly so dramatic.
For example, those in software development for awhile may remember that voice recognition had accuracy of maybe 90% with *trained* data sets for decades before 2010. When NNs came along, accuracy jumped orders of magnitude on *untrained* data in just a couple years.
Sometimes the right approach is a process. There were email and document sharing systems for decades. After the internet came along with open RFCs, these took off like a rocket and use expanded 4 to 5 orders of magnitude.
The point is that you only know when you're going to solve really hard problems when you have a solution that scales quickly and works. Until then, the basic paradigm probably needs to be invented.
Maybe Tesla's rewrite will get there. But when EM says he can almost drive home to work without intervention, that doesn't sound optimistic.
I would not feel comfortable riding in a car twice daily if the driver had an accident once a year. Yet, that is 3 orders of magnitude improvement over FSD that has an intervention each trip. We need 4 or 5 orders of magnitude improvement which is 10 to 100 times better than 1000 times better than the current state.
We'll know FSD in the general case has a chance when interventions quickly go from 1 per trip to 1 in 10 trips in 6 month or a year. We've seen gradual improvements over the past couple years, but it's not nearly so dramatic.
Last edited:
Lot's and lot's of anecdotal evidence there.Tried to consolidate all replies to 1 post since the thread is already crazy big
Do you have a source for the 22 mph being typical? I'd imagine it's even slower in say NYC, but it's be a LOT higher here in NC where I live (and pretty much the entire country outside dense urban cities if we're gonna pretend robotaxis will replace lots of "normal" cars too not just in dense cities).
It also appears to ignore that a good portion of the country, especially some of the larger denser cities in the US, are in the north where it's very cold a lot of the year- and efficiency to keep the cabin warm for riders causes a significant range hit (I suppose the Y will do better here than the 3 at least)
It also appears to assume all the RTs will be long-range vehicles... an SR+ already can't make it 264 miles.
Sure, as soon as someone invests the money to build those things.
Which I specifically mentioned as an option in my post.
In fact the option that probably makes the most sense (well, SOME charging stations with an attendent, be it SCs, urban SCs, or just regular 240 service of some kind).
This will be even more needed as you start trying to replace things like trucks, where someone calls an RT to transport a truck full of dirt or gravel or something for them... suddenly someone needs to clean the bed out.
I guess you also missed where I specifically mentioned none of these problems were show stoppers, they'd just require some $ outlay to fix because the infrastructure as Tesla built it for long-range road trips is vastly different from that needed for lots of short-range taxi trips, and even WITH that spend would still easily be a cash cow for the company.
Not really though.
Because a robotaxi can run 24/7. I know uber drivers that only drive a few hours a day. So One RT likely replaces like 3-5 rideshare hail drivers at least.
So cut that number down quite a bit for how many RTs are actually "needed" instead of uber/lyft cars.
(note too your source uses "probably" a lot in citing number of drivers- a quick search shows a pretty wide array of guesses at the real #)
FWIW I don't disagree at all Tesla is well ahead of anyone else trying the vision-only approach (and that that is the ideal approach)- I just disagree with the idea the current HW is sufficient to reach L5. (the fact they've already changed it multiple times since originally claiming it was sufficient seems pretty good evidence they don't actually know how much is sufficient until they get there.... see also Elon himself saying you shouldn't trust any prediction he makes about a thing he hasn't ever done before- and how many such predictions he's gotten wrong before)
Nope.
I grew up in NY.
I'm very very very familiar with dense urban situations and taxis in those environments.
The fact I'm in NC now gives me the other side of that perspective. (and I've lived in both "cities" in NC, and now "out in the country"). I've also spent time in quite a lot of major cities elsewhere in the US (and western Europe).
There's very very very little in the way of taxis here- because they don't make nearly the same economic sense.
There's SOME uber/lyft presence, not so much close to my house but maybe 15-20 minutes away you can reasonably find one within 10 minutes.... but again massively less than in dense cities. Mostly you need to own a car here.... whereas I think anyone owning a car in NYC is insane.
That said- there's still like 2 million NYC residents who own private cars.
Despite MASSIVE availability of taxis, uber, subways, buses, etc.
That's why I think the idea everyone dumps their cars when robotaxis get here doesn't hold up to the facts.
People like owning cars. Even in places it makes NO SENSE for them to do so.
See above. Owning a car in NYC is crazy expensive. Parking alone can be more than a car payment in lots of parts of the city.
Yet a couple million people still do it. Despite easy public transit and a very walkable city on top.
So the idea they'll all stop owning those cars once their taxi doesn't have a driver appears just wishful/magical thinking.
Even moreso the folks in less urban areas.
Again- if someone only wants to get from A to B and doesn't care about owning the vehicle or it being a specific vehicle- we'd see most cars being sold being $15,000 econoboxes.... instead of an average new car price in the US of around $35,000, and pickup trucks and SUVs being wildly popular.
See also all the private car owners in dense urban cities with excellent, cheap, public transit- why do they still own cars?
Why would RTs change their mind?
I agree RTs would likely replace SOME private vehicles. There's some % especially where when their existing car is on its way out, going full RT might be a reasonable replacement.
I just think some estimates of how many are wildly, wildly, optimistic. See again how many folks in places like NYC where owning a car makes 0 sense still own a car. Millions.
And increasingly so as you get outside dense urban cities.
So I agree with your points 1-3.... the question is- where I live for example I'm roughly 15-20 minutes from any retail other than a gas station type store.... I'm ~30 minutes to a couple of decent sized cities... and 45-60 minutes from several other major cities.
But the population density is so sparse there's no high speed internet available. (HURRY THE HELL UP STARLINK!)
I wouldn't ever consider replacing my car with a robotaxi unless 1-3 were all true.... not just at my house, but anywhere I'd be likely to travel TO so I can get back easily... but I don't see the economic model where someone buys a $40,000 robotaxi and then has it sit within 5 minutes of my house for the maybe couple of trips a day they'd ever get from me and my very sparse neighbors sitting there.
The fact it's CHEAP only works for the guy who paid for the RT if it's also making trips often.
Easy in LA, Chicago, Atlanta, etc downtown.
Less easy in the burbs.
Pretty hard in the "country"
Nice to speculate but we will all have to wait and see.
I'm putting my money on the guy who lands rocket's on ship on the ocean.
dc_h
Active Member
I think we will know when competitors are getting close to L5, when they begin investing more in reducing their hardware costs. Tesla’s design incentives focused on cost efficiency from day 1, while the others assumed their hardware costs were not relevant in the long term.
Building the network software could be more challenging then many here assume, but the cost to Tesla should be much lower then Uber and Lyft. Integration of the car operating system and the TaaS platform, as much as any potential lead in full autonomy may be as important for market share and cost advantage. Built in GPS, detailed call logs, a could platform already connected to the cars are all tools that will need to be developed by other autonomous service providers.
Similar threads
- Locked
- Locked
- Locked
