-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
My Experience With EAP Suggests Not Only Won't FSD Go Up In Price, It's Going To Go Down
- Thread starter FredMertz
- Start date
For the price of FSD as it stands, you can afford a lot of Uber / Taxis to pick your drunk self up from the bar. That's when you really need FSD
To be honest, what I want level 3+ for is being able to make longer drives on my own again. Due to brain surgery the constant concentration with visual stimulation when driving creates brain fatigue after a few hours for me and I become a very unsafe driver at that point. Meaning doing a long drive to visit friends becomes impossible if I am alone in the car.
And even so having some sort of eyes of driving would make the 15-16 hour road trips very accessible to everyone even if they don’t have the same problems as I have. At that point you can relax in your car most of the way instead of driving.
I have no use of FSD for going around town, drunk or not.
Go watch the thousands of videos on YouTube of people abusing Uber and Lyft drivers and their vehicles. That’s with a driver in the vehicle and on camera. You think a sticker stating a camera is on is going to mean anything?cue turning interior camera ON. And "Smile you're on Camera" sticker on the seat backs
If there's anything that Elon and Tesla are guilty of "getting wrong", it's not making it plainly obvious what you are buying when you pay for FSD because there seems to be a huge disconnect in understanding as evidenced in all of these posts.
Although the current value of FSD is in some part attributable to a 'safety feature', the amount Tesla charges for FSD is, effectively, a software license for an income-producing future functionality. The price they charge will go up and continue to increase the closer they get to achieving true autonomy. In other words, as their confidence increases that they will achieve an autonomous-capable robo-taxi business model within the life of the vehicles currently being sold, they will charge more because there is less risk that the investor (car buyer) will not have the opportunity to, at least, recoup his/her investment.
When full autonomy is reached, the value of FSD will 100% be a function of discounted future cash flows like any other business. If this is a vehicle designed and built to be driven one million miles, as planned, this should be something like a 10-year life on average, earning its owner around $30k per year. Ignoring the price of the hardware (the vehicle itself), this places the FSD software value at WELL OVER $100k. This value will be reflected in the price of buying a Tesla vehicle with FSD at that point. It remains to be seen whether Tesla will even continue selling cars at that point. Becoming the first autonomous transportation-as-a-service provider has truly astonishing valuation effects, making the profit margin per vehicle built something more like 90%+. All this while simultaneously reducing the cost of transportation to significantly less than owning your own non-FSD vehicle (and, consequently, disintegrating the value of EVERY non-FSD on the road - good luck trying to sell your car, any car, without FSD after it exists;-).
The fact that Tesla offers an opportunity to buy-into this future functionality at a discounted price says volumes about their confidence in the timeline needed to achieve it. The fact that they periodically raise the price is putting the public on notice that confidence is increasing. Considering the fact that Tesla currently only produces vehicles with a maximum expected life of 300k-500k miles and, for earlier models, realistically probably much less, full autonomy is obviously expected within the next few years. FYI, there are major accounting implications to this estimate in order to ratably recognize FSD revenues on the P&L and that means Tesla's auditors (PWC) must be sufficiently satisfied that the underlying assumptions are, in fact, based on realistic progress (not that PWC is considered to be an expert in autonomy - just that assurance of a minimum threshold exists). If Tesla vehicles for which FSD was purchased begin to be retired in large numbers prior to the accomplishment of full autonomy, there may be some accounting changes, refunds, and possibly rights to port the FSD 'license' to a new Tesla vehicle using the amount previously paid as credit toward the new price.
So, IMO, the safety benefits/features and automation of highway tasks on long roadtrips from the current FSD package are easily worth $4k for my family, right in line with the latest EAP upgrade option. The opportunity to participate first-hand in the real world testing and development of one of the greatest technological innovations/disruptions in human history is easily worth $2k to any self-respecting tech geek who upgrades his/her smartphone more than once every few years. An additional $2k for the fairly likely opportunity to see my vehicle become the first mass-produced car become an ASSET INSTEAD OF A LIABILITY??? That's a no-brainer.
Although the current value of FSD is in some part attributable to a 'safety feature', the amount Tesla charges for FSD is, effectively, a software license for an income-producing future functionality. The price they charge will go up and continue to increase the closer they get to achieving true autonomy. In other words, as their confidence increases that they will achieve an autonomous-capable robo-taxi business model within the life of the vehicles currently being sold, they will charge more because there is less risk that the investor (car buyer) will not have the opportunity to, at least, recoup his/her investment.
When full autonomy is reached, the value of FSD will 100% be a function of discounted future cash flows like any other business. If this is a vehicle designed and built to be driven one million miles, as planned, this should be something like a 10-year life on average, earning its owner around $30k per year. Ignoring the price of the hardware (the vehicle itself), this places the FSD software value at WELL OVER $100k. This value will be reflected in the price of buying a Tesla vehicle with FSD at that point. It remains to be seen whether Tesla will even continue selling cars at that point. Becoming the first autonomous transportation-as-a-service provider has truly astonishing valuation effects, making the profit margin per vehicle built something more like 90%+. All this while simultaneously reducing the cost of transportation to significantly less than owning your own non-FSD vehicle (and, consequently, disintegrating the value of EVERY non-FSD on the road - good luck trying to sell your car, any car, without FSD after it exists;-).
The fact that Tesla offers an opportunity to buy-into this future functionality at a discounted price says volumes about their confidence in the timeline needed to achieve it. The fact that they periodically raise the price is putting the public on notice that confidence is increasing. Considering the fact that Tesla currently only produces vehicles with a maximum expected life of 300k-500k miles and, for earlier models, realistically probably much less, full autonomy is obviously expected within the next few years. FYI, there are major accounting implications to this estimate in order to ratably recognize FSD revenues on the P&L and that means Tesla's auditors (PWC) must be sufficiently satisfied that the underlying assumptions are, in fact, based on realistic progress (not that PWC is considered to be an expert in autonomy - just that assurance of a minimum threshold exists). If Tesla vehicles for which FSD was purchased begin to be retired in large numbers prior to the accomplishment of full autonomy, there may be some accounting changes, refunds, and possibly rights to port the FSD 'license' to a new Tesla vehicle using the amount previously paid as credit toward the new price.
So, IMO, the safety benefits/features and automation of highway tasks on long roadtrips from the current FSD package are easily worth $4k for my family, right in line with the latest EAP upgrade option. The opportunity to participate first-hand in the real world testing and development of one of the greatest technological innovations/disruptions in human history is easily worth $2k to any self-respecting tech geek who upgrades his/her smartphone more than once every few years. An additional $2k for the fairly likely opportunity to see my vehicle become the first mass-produced car become an ASSET INSTEAD OF A LIABILITY??? That's a no-brainer.
Yes, it is going to mean I don't have to front the cleaning bill!
For the record, I would never robo taxi my own personal vehicle. I'm a germaphobe
Knightshade
Well-Known Member
This is only true in a world with infinite demand for robotaxis.
A world which obviously does not exist.
Tesla already will have 1 million allegedly robotaxi capable cars on the road THIS YEAR- which is more than all taxis and uber/lyft drivers in the US combined.
And Tesla has plans to significantly continue to ramp up # of cars produced year over year- aiming at 50% YoY growth on average for another decade- with perhaps as many as 20 million cars per year made by 2030.
The idea "everyone" with an FSD Tesla will be able to earn $30,000 a year is nonsensical.
Watch the Battery Day replay (scrub into 1:52 where he talks about the $25K car)
He didn't say "and 8K more for being autonomous"
Think so? Although I would agree that the entire population can't be expected to always act in their best interest and there will be a transition period during which it may be cheaper and more convenient to drive the car you already own, the cost per mile that the robo-taxi fleet will be able to offer ride-sharing (i.e. much, much less than the current Uber/Lyft options) will ensure near-capacity usage of all participating vehicles.
The use of current rideshare options isn't driven as much by cost as it is convenience. That will change dramatically once it costs significantly less to use a robo-taxi than it does to own/operate your own car. After all, at that point each participating Tesla vehicle becomes a for-profit business and pricing equilibrium will maximize utilization and, therefore, total profitability of the asset (not profitability per mile). Regardless of the $30k/yr income estimate, the point is that the cost to participate (current FSD price) is MUCH less than it has the quite likely potential to earn the owner, making it an attractive investment. And the owner, in the future, is likely to be Tesla or a fleet operator who was willing to pay the premium value of FSD at that point.
So as an individual, quite literally, this is your only chance to get a piece of that pie with the few Teslas you might buy during this transition period... Demand for robo-taxis will rise exponentially as the user's cost/mile drops.
Knightshade
Well-Known Member
I find that.... exceedingly optimistic...to put it mildly.
People REALLY like owning their own cars.
Even in places it makes no sense to like NYC.
I absolutely think RTs would reduce total # of privately owned cars- for many for example it'd no longer make sense to own a second or third car in a household.
But replace even half of PRIMARY cars? Don't really see that happening.
It costs an insane amount to own a car in NYC (many other places too but it's the most obvious example). A city with outstanding (by US standards) public transit, and very short distances between most places in a dense populated area.
Yet almost half of all households in NYC own at least one car.
Clearly there's lots of folks more than happy to pay MORE to have their own car than have rely on others to do so.
And that's even more massively true in suburban and especially rural areas where density is much lower and distances between places much greater.
Joe Rural Iowa isn't gonna give up his truck so he can stand around outside waiting for a robotaxi in the next actual town over to show up and take him to the store for some parts for his tractor.
Except when literally anybody else can do the same thing for the same cost.
At which point you get a bunch of people investing, and discovering the market for their product is suddenly nowhere as large as optimistic folks thought.
Which brings us to-
Yeah- the fleet operator is gonna be able to throw a ton of robotaxis into an area, and likely have even lower running costs with scale.
Speaking of scale.... where you do L3 charge 20 million robotaxis? The supercharger network has been expanding far slower than the size of the fleet has- with no change in site for that as the current Tesla model is that superchargers are for roadtrips- not normal daily charging.
That's crap for robotaxis though.
Because if the car has to stop robotaxing and drive back to your house and sit on an L2 charger for 6 hours every 3 hours of driving that's cutting into your imaginary 30k a year pretty hard.
(here again a fleet operator has an advantage in that he can still be running a shift of cars while his other 2 shifts are charging)[/QUOTE]
Ah, the good 'ol ego driving people to spend 100x more to get from point A to point B than they need to in order to look good while doing it! Yes! You are absolutely correct about people's obsessions with their cars (Americans in particular). And the same decision already exists - spend $300k on a Ferrari? Or $15k on a Focus? For most, those comparison price points entirely depend on what they can afford, obviously. There will undoubtedly be hold-outs during this transition, which is why it's considered a transition. But it will happen. Ask your local Millennial. And, interestingly, your ego isn't quite as reluctant to get into a ride-share vehicle that's "beneath you" because it's obvious to observers that it's not yours. Not that anybody is ashamed to be seen getting in/out of a Tesla
All good points about urban and rural differences. Although I don't know where you get your statistic about NYC residents' car ownership being nearly 50%??? The adoption of the TAAS model will be faster in certain locations than others and I make no claim about car ownership going away entirely. There will be niches. But that is exactly what they will be relegated to in the long run.
Not sure what you mean by "anybody being able to do the same thing for the same cost". The fully autonomous TAAS market is going to be largely a first-mover advantage, especially if your competition (Tesla) also manufactures EVs with the lowest cost to produce. Seeing as how the hardware (car) value will pale in comparison to the software (FSD) value, it will behoove competitors to license the software from Tesla rather than continuing to develop their own (unless they have a clear and present path to a feature complete version of their own).
As for fleet operators, yes, this may very well be the best model for all parties. But I wouldn't discount the possibility of Tesla being the local fleet operator or franchising it. I think your assumptions about charging infrastructure are premature. We don't have the push for the L3 charger saturation yet because, well, right now the EV use case only needs it for roadtrips. The first robo-taxis will be in urban envrionments, with rides averaging much less than 45mph, which means the range of an EV is significantly increased (versus pushing air at 55+mph on a highway). The vehicle ranges that Tesla appears to be shooting for should last the entire lot of average waking hours, at which point the demand for RTs drops dramatically and charging begins. There will be plenty of L3 (or better) charging options by the time they are needed.
All good points about urban and rural differences. Although I don't know where you get your statistic about NYC residents' car ownership being nearly 50%??? The adoption of the TAAS model will be faster in certain locations than others and I make no claim about car ownership going away entirely. There will be niches. But that is exactly what they will be relegated to in the long run.
Not sure what you mean by "anybody being able to do the same thing for the same cost". The fully autonomous TAAS market is going to be largely a first-mover advantage, especially if your competition (Tesla) also manufactures EVs with the lowest cost to produce. Seeing as how the hardware (car) value will pale in comparison to the software (FSD) value, it will behoove competitors to license the software from Tesla rather than continuing to develop their own (unless they have a clear and present path to a feature complete version of their own).
As for fleet operators, yes, this may very well be the best model for all parties. But I wouldn't discount the possibility of Tesla being the local fleet operator or franchising it. I think your assumptions about charging infrastructure are premature. We don't have the push for the L3 charger saturation yet because, well, right now the EV use case only needs it for roadtrips. The first robo-taxis will be in urban envrionments, with rides averaging much less than 45mph, which means the range of an EV is significantly increased (versus pushing air at 55+mph on a highway). The vehicle ranges that Tesla appears to be shooting for should last the entire lot of average waking hours, at which point the demand for RTs drops dramatically and charging begins. There will be plenty of L3 (or better) charging options by the time they are needed.
Daniel in SD
(supervised)
One thing I've never understood about the robotaxi story is that the Model 3 would be a horrible taxi. It's a sports sedan, the ride is rough, there's no headroom in the backseat, and you have to teach people how to use the door handles. Also, one of the seats has a useless steering wheel in your lap! Elon was obviously not following his "first principle" design philosophy.
The GM/Cruise robotaxi design seems way better (if they ever release it!).
The GM/Cruise robotaxi design seems way better (if they ever release it!).
The Model 3 will NOT be the robo-taxi of the future for many reasons, including the points you make. But it's one of the vehicles that bridges the gap to the TAAS reality. Once function over form rules the form factor decision because of the underlying economics and the eventual end to having to appeal to the egos of car 'buyers', the robotaxi of the future will look something more like that model, for which Tesla will have models ready for production... That's quite a ways off, however. Perhaps GM is smart in working on this design now, expecting to license FSD from Tesla?
Knightshade
Well-Known Member
I get them from facts.
Where you do get the support for your claims?
New Yorkers and Their Cars | NYCEDC!).
The article even points out that most of them own cars OTHER than for just routine commuting. Meaning the exact best use case for a robotaxi wouldn't change their wanting or needing to own a car at all.
I mean literally anybody can buy the same Tesla with FSD you did and compete against you.
And fleet operators can likely do so even better and cheaper than individuals.
So markets will be vastly oversupplied with taxis- far more than most areas actually need or want.
Making the idea you're gonna make 30 grand a year for EVERY tesla owner hilarious.
Just to point out... Waymo has working robotaxis in service today
Tesla does not.
I don't personally believe the current sensor suite is capable of L5 FSD on Teslas anyway.... right now even L2 NoA shuts itself off in moderate rain, and basic AP does in very heavy rain or snow.
I think L3 highways is very very very likely, and maybe L4 on highways, and with the operational domain of either defined as clear to perhaps light/moderate rain at best.
But I think there's a pretty good chance you're gonna need more sensors to get even L3 in the city let alone 4 or 5.
God I hope not... Tesla is horrendously bad at logistics and customer service.
They do absolutely fantastic engineering and some very clever software though.
So...much like yours on L5 robotaxis, except chargers at least exist today
Those are literally the hardest environments to get a safe working system in though.
Much more likely you'll see driverless appear first in long-haul trucking platoons and the like, where the vast majority of the trip is on freeways- that's a much easier case to solve for.
Tesla has already discussed the need for megachargers to support the semi... but hasn't really built any yet.
Likewise they originally promised to build special supercharger stations for the cybertruck that could handle charging while towing a trailer- those too appear to remain imaginary (though there have been recent plans for wider stalls at least.... you'll still not be able to charge with a trailer but at least naked CTs will be able to use em).
See again Tesla isn't great with logistics.
This is the company that was surprised when they made their first very mass production car (the 3) and suddenly realized "Huh- apparently we need a ton of trucks to transport these... guess we should've thought of that" after the fact.
Then a few months later suddenly realized "Huh- apparently we should've already built a TON more service centers to handle this vast increase in fleet size... our bad"
Now- financially- kicking the can on this stuff down the road was probably needed... but it's not a great look to keep admitting you didn't bother planning ahead for the obvious results of the thing you're doing.
Current range on 3/Y is around 300ish miles. More like 225-250 if it's cold. Also you need enough range to actually get to your first pickup... and between pickups... and back to the cars home. And you don't want to run the battery below 10% if you can help it... So maybe 150 usable miles on average for paid taxi work if you're lucky.
At say 30 mph that's 5 hours.
So paid attendants- or will the cars somehow charge themselves? (which obviously current ones can't do at all)
Attendants make more sense, since they can also do things like clean interiors, put air in tires, clean sensors off if needed, etc... but then they need to be paid which'll eat into profits.
Here again a fleet operator might beat the individual owner by having their own L3 station too.
GM is a traditional car maker where they test/validate things well before they release them. Sure delays happen, and enhanced supercruise did get delayed a bit because of COVID-19.
But, we already know what features enhanced supercruise has on them because they are being sold this quarter.
So I'm actually more confident that enhanced supercruise will get released this fall, and will work than I am about the Tesla rewrite being
released by the end of this year.
Right now the auto-lane changes aren't consistent enough (in all circumstances) to go up against something that uses the combination of visual+radar like the Cadillac Enhanced Supercruise does.
Don't get me wrong. Tesla is the team I'm rooting for, but we have to be realistic here.
Knightshade
Well-Known Member
To be fair- FSD has been sold since 2016 and we're still not sure what the final product will look like
Essentially TODAY GM is offering capabilities not significantly better than Tesla in 2014.
it's certainly possible whenever Enhanced SC actually is available to use it'll catch them up to roughly 2017/18 Tesla (though on a much more limited set of roads), we'll have to wait and see.
Sorry, I don't appreciate the implication that my claims aren't based on facts, especially when I haven't made any such statistical claims. I am trying to spell out the case for why Tesla is justified in charging for FSD and why a current Tesla buyer should seriously consider paying for it.
I made no claim that NYC is the perfect microcosm for the robo-taxi business model. And I do not believe individual car ownership is going away anytime soon for anyone willing to pay the premium to do so. But how about China? Brazil? India? Even if Americans are the hold-outs, generally speaking, the rest of the world has much fewer cars per citizen already and a natural fit for this model.
The likelihood of Tesla reaching full autonomy and, consequently whether anyone else is close (i.e. Waymo) is an entirely different conversation. But for the record, a geofenced solution isn't a solution at all.
Any company that experiences rapid, highly scrutinized growth is going to have logistics issues and these are improving all the time. I don't think anything you've stated is either detrimental or insurmountable and Tesla is working toward solutions.
You are incorrect on the range concerns. A 300mi range M3 will achieve that range +/- when it is driven in the exact manner that the EPA test cycle prescribes. This includes a lot more freeway driving than many robo-taxi trips will require. And since it takes 8 times the power to push a vehicle at 55mph than it does at 25mph, the increase in range can be substantial at lower speeds. See the 606mi M3 hypermiling range record for evidence of that effect. Obviously that can't be expected, but it illustrates the point. Throw in continued effciency improvements via OAT updates and, for future models employing the latest battery tech, that range can/should realistically reach 10hrs+ of service time before requiring a charge.
Finally, I think we disconnected somewhere on who would be running the fleet of robo-taxis. Individual owners who opt-in to the Tesla fleet will not have pricing authority. They will not be in competition with each other and if their robo-taxi isn't properly cleaned each night, verified by rider feedback, they'll risk getting kicked out of the fleet. Instead, users will use a Tesla app just like they do for Uber/Lyft with the same dynamic pricing factors, just cheaper to use because of the obvious lack of a driver and using highly efficient EVs that are designed to last one million miles. Again, individual ownership is a transitional phase opportunity. Not the long-term model. Fleet operators and, yes, Tesla, will be the likely long-term solution.
Nothing you've presented is entirely incorrect or not a concern. They're just all surmountable and FSD makes sense to buy now if you can. I won't respond again. So the last word is yours if you want it and feel the need to have it. Thanks for the discussion.
I made no claim that NYC is the perfect microcosm for the robo-taxi business model. And I do not believe individual car ownership is going away anytime soon for anyone willing to pay the premium to do so. But how about China? Brazil? India? Even if Americans are the hold-outs, generally speaking, the rest of the world has much fewer cars per citizen already and a natural fit for this model.
The likelihood of Tesla reaching full autonomy and, consequently whether anyone else is close (i.e. Waymo) is an entirely different conversation. But for the record, a geofenced solution isn't a solution at all.
Any company that experiences rapid, highly scrutinized growth is going to have logistics issues and these are improving all the time. I don't think anything you've stated is either detrimental or insurmountable and Tesla is working toward solutions.
You are incorrect on the range concerns. A 300mi range M3 will achieve that range +/- when it is driven in the exact manner that the EPA test cycle prescribes. This includes a lot more freeway driving than many robo-taxi trips will require. And since it takes 8 times the power to push a vehicle at 55mph than it does at 25mph, the increase in range can be substantial at lower speeds. See the 606mi M3 hypermiling range record for evidence of that effect. Obviously that can't be expected, but it illustrates the point. Throw in continued effciency improvements via OAT updates and, for future models employing the latest battery tech, that range can/should realistically reach 10hrs+ of service time before requiring a charge.
Finally, I think we disconnected somewhere on who would be running the fleet of robo-taxis. Individual owners who opt-in to the Tesla fleet will not have pricing authority. They will not be in competition with each other and if their robo-taxi isn't properly cleaned each night, verified by rider feedback, they'll risk getting kicked out of the fleet. Instead, users will use a Tesla app just like they do for Uber/Lyft with the same dynamic pricing factors, just cheaper to use because of the obvious lack of a driver and using highly efficient EVs that are designed to last one million miles. Again, individual ownership is a transitional phase opportunity. Not the long-term model. Fleet operators and, yes, Tesla, will be the likely long-term solution.
Nothing you've presented is entirely incorrect or not a concern. They're just all surmountable and FSD makes sense to buy now if you can. I won't respond again. So the last word is yours if you want it and feel the need to have it. Thanks for the discussion.
I would put even the current version of Supercruise as being superior to my HW3 FSD Tesla.
Why? The roads I use TACC/AP on are covered by Supercruise and the hands free solves a major annoyance of EAP. Plus I would love to be able to be in the right lane without my car re-centering everytime the lane widens during a merge point.
What would I miss? Auto-Lane change because I enjoy it despite some of its embarrassing misses.
So to really leapfrog what I have now would require enhanced autopilot.
I do acknowledge that even enhanced supercruise doesn't have features EAP/FSD have, but I haven't been able to successfully use those features on a consistent basis. So I haven't gotten any enjoyment out of them.
Daniel in SD
(supervised)
Tesla's price was far too low when they were actually selling Full Self Driving. People purchasing now are only promised "Autosteer on city streets" as a future feature. The price is too high given the risk that that is all you'll get.
Similar threads
