Elon doesn't give a shit about reaching a utopia, merely that people drive (his) cars, and is willing to torpedo all public transit projects to reach that. He is an active threat to society.
Because as it gets better, it provides more value. In its completed form it basically replaces a driver. Calculate how much it would cost to have drivers capable of driving 20 hours/day (rest spent charging).
If they actually make it self-driving at a human level, it's value is clearly over $100K.
To some buyers anyway. It's not really clear that a higher price ends up being the revenue/income maximizing price (if the cost is minimal and many buyers value it much lower than the high price).
If they actually do something impossible, sure, it'll be worth some money. But even then, why would individuals ever be able to participate in that market?
If it's a sure thing, companies (including Tesla) will buy many cars and run the value as low as they can.
Musk once argued that once full self-driving works, it would be more profitable for Tesla to use these cars for an own taxi service than it would be to sell them to customers. They plan to raise the price even more the closer they (claim to) get to their goal.
I don’t understand that endgame. I would never want strangers in my car unsupervised. They will be having sex, puking, changing diapers, and worse. Just imagine all of the fluids and smells that they would leave behind. How is this remotely a good idea?
The first part is the general assumption that the endgame of self-driving cars is robotaxis. I don't personally buy this--most people who own their own cars today clearly prefer to own their own cars over relying on taxi services or public transit--but this is dominant assumption you see in newspaper articles and business planning. And I suspect it's mostly driven by greed for recurring service revenue over one-time sales revenue.
And the second part is the aspersions to Uber (as a stand-in for all the similar services): renting your own car out as a robotaxi is a lot like being an Uber, except for being the driver part. Of course, Uber itself is basically a losing model that doesn't really compensate enough to cover the wear-and-tear on your car from the extra mileage acting as a taxi, so it's not entirely the best model to be emulating. But how could a $10 billion company possibly be wrong?
> I don’t understand that endgame. I would never want strangers in my car unsupervised.
You are misunderstanding the comment you are replying to.
It doesn't say "Tesla utilises the car which it sold to you as a taxi". It says that Tesla doesn't sell you a car. And they don't sell you a car because they can make better money by using their factory output as taxis, than by selling it to you.
I don't understand it at all myself. Assuming you wanted the car to drive 10 hrs a day on average (because of low overnight demand, random charging, maintenance, etc.) a FSD Tesla vs. a new Hyundai Ioniq with a human driver would have the same total costs after 2-3 years (when the hourly salary of the human driver finally makes the Ioniq more expensive). That's what Tesla is talking about competing with in the robo-taxi market.
It was advertised by Tesla and as a Level 5 system. However, given that never happened, it is now known as 'Fools Self Driving' since not only it doesn't work, requires constant full assistance and attention of the driver behind the wheel at all times, it is been admittedly a Level 2 system who's creator gives the endless promises of updates in order to fool the customer that it will 'eventually' be 'Full Self Driving'.
So Tesla can continue to raises prices on this false promise and customers will keep buying, subscribing on the idea that they are hoping that it will be fully self driving, which is the typical pied piper selling snake oil scam.
Unsurprisingly and once again, the Fools Self Driving (FSD) scam raises prices on their customers towards a product that is beyond broken and doesn't work.
It is gotten so bad that even the noble and devoted Telsa fans would do anything to defend this contraption, even if it meant risking the lives of their own children, they will still do it for Twitter points, or for their lord and saviour techno-king Elon to notice them.
Who is going to finally tell the Tesla FSD fans that the piped piper (Elon) has repeatedly over promised and scammed his own fans and customers using FSD to the same tune?
There are many people who have bought FSD from Tesla, and have had their car reach end of life without ever getting to enjoy FSD.
It's possible after being "less than a year away" for... a decade... that it may be less than a year away now. But it looks like a bit of skepticism has been earned.
A conflict of interest accusation is pretty serious, and I think if all you had to go by is your skepticism, you probably shouldn't just make that claim as if it were a known fact: that's just very dishonest.
Sure, be skeptical, but then maybe say something like "I have no evidence for this, but I believe his manifest enthusiasm can only be explained by the fact that he's trying to manipulate the stock price to his own benefit."
Sure, it sounds less rhetorically convenient, but it's more honest.
It is my understanding that he had a substantial amount of shares up to 2021-07-29. After criticism about the conflict of interest he claims to have divested his shares.
I've watched a lot of Munro (and even bought one of their retail reports for fun) and the thing that most bothered me that looked like tesla bias was when he was doing side-by-side stats comparisons of all of the mid-size crossovers like the id.4, mach-e, and tesla y. he kept on harping about the epa range in the comparison and never once mentioned the methodology of the epa estimate and the fact that manufacturers have a lot of leeway to conduct the test themselves - and either overreport or underreport the epa range. this was really glaring to me because tesla is known to stretch their range estimates (as do other manufacturers), but some manufacturers seem to deliberately underreport (mercedes, porsche, ford with some models of the mach-e for example). given that he's in the business of competitive analysis he HAS to know this, but he never said a word about it, which I think made tesla look a little better than it should have. it's one thing to just present the epa range and leave it at that, but he was using it as the basis for a ton of discussion and derived numbers and basically putting down the non-tesla cars... all without mentioning the limits of the number
see this for some real-world testing, there's lots like it and the results usually skew a certain way for each car. for example, there's about a 20% delta in epa vs. tested range for the tesla y LR and mach-e standard range in his chart. the tesla artificially looks 20% better! and this delta was well-known at the time of his video
https://insideevs.com/reviews/443791/ev-range-test-results/
to your question about conflicts of interest, owning a stock is a conflict of interest. nobody needs to show that something is happening inside his head, the circumstances exist and that's it. I think he's usually pretty good and I still watch the videos, and he trades on his reputation and I still regard him as a good source. but I'm still glad to learn that he's sold the stock
It's shocking to me that people are willing to pay to be beta testers for Tesla's entirely-from-scratch manufacturing process and smartphone-with-wheels style of car experience but it doesn't matter what I think - that's the beauty of capitalism.
I have driven a friend's model 3 and while yes its nice, I feel like I would get more value for my buck with the driver assist in for example modern Toyotas.
But I'm also perfectly happy driving my decades old Honda so what do I know.
Based on the videos I've seen I'm not even sure I'd call their software beta. It looks like it's in early development, like where you'd be running it exclusively on closed tracks. Definitely nowhere near folks outside the company.
NFTs don’t have amazing unit economics with gross profit margins of 30% (or any unit economics at all), and incredible discounted cash flows (actually they don’t have any cash flows at all).
It’s like comparing the productivity of a med student on Adderall in a quiet library, and the productivity of a peeled russet potato sitting in the sun.
What are you talking about? I watch FSD progress, and it's clearly getting closer to human level driving. Just go on YouTube and compare FSD from two years ago to the latest. It's night and day. Yes, it still makes mistakes. Humans make mistakes all the time while driving.
As for build quality, we haven't had any major problems with our Model 3. It's a great car.
Tesla has a media problem, where every little problem gets reported as if it's systemic and unfixable.
Does it work in rain, snow, and/or night? How about country roads and roads without clearly marked lanes or misleading lane lines?
It's like hiring a plumber who can only flush toilets, not install them. Self driving in pristine conditions is a party trick. Until it can do the hard part, it's nowhere near human-level.
At autonomy investor day they were saying better than a human in all human drivable conditions by 2020 for the robotaxi network they said they would launch then. They confirmed they meant level 5 autonomy in answer to a direct question.
They knew dojo wouldn't be online by then and they didn't even really train on video at that point, just trying to judge the scene image by image instead of taking into account prior frames (like knowing someone walked behind a bush 1p frames ago, so that sliver of emerging leg from behind a bush that might not be detectable confidently enough from a single image should be weighted higher).
FSD handles those situations well, compared to Waymo etc, because FSD doesn't depend on pre-mapping. Waymo depends on pre-generated 3D maps and it fails if there's a discrepancy. That's why Waymo works only on small areas and uses cherry-picked routes.
> Tesla has a media problem, where every little problem gets reported as if it's systemic and unfixable.
Similar to Apple, people expect more from them than anyone else. Minor issues on Apple products are always scandals. Remember Antennagate? People were sure that it would sink Apple, because "they don't know how to make antennas". People and media handle Tesla similarly, and minor issues are blown out of proportion.
It's the same issue. There is no double standard. Taking advance payment and being 5+ years late is something any company would get massive criticism from, if not sued.
The suspension on the model 3 and Y is definitely not luxury. Nor is the interior build quality. I own a model Y and there’s plenty of creaks and noises in it that other car models don’t have. Also the highway FSD is still ridiculously bad with it’s phantom braking where at 70MPH your car will just slam on the brakes. And highway driving is the easy version that most other manufacturers have now too.
And now it looks like Cruise and Waymo have full self driving in San Francisco which looks amazing. There’s hour long videos of it navigating perfectly and smoothly in a jam packed city. Nothing like the Tesla videos which show it almost slamming into bikes/curbs or just freaking out and stopping in the middle of an intersection.
Has great acceleration though and charging at home is great or on the road with superchargers. But Teslas edge is not as wide anymore.
If phantom braking is reproducible, ie it happens at the same place and time(because sun angle) you should report it to Tesla. Or post a YouTube and link on Twitter. In 15k miles of mostly autopilot, I have had phantom brake happen once or twice. I’m in the Bay Area. It happened when it couldn’t tell wether the shadow of over cross bridge from a physical object. What is your experience?
>It happened when it couldn’t tell wether the shadow of over cross bridge from a physical object.
Who could have predicted that just relying just on cameras was not a good solution ? Definitely not the hundreds of researchers that told Musk that a LIDAR was necessary.
Idk why people preach this a gospel. There are billions of living things, some of which ( peregrine falcon, cheetah etc)do high way speeds and do fine relying on optics. The only animals that rely on sonar are those that live in the dark. There are insects with pea brains which rely on vision. I’m not saying it is proof that only vision works, but to argue against it requires more than platitudes.
They have two eyes close together that do tracking of depth. Plus they have ears and touch, multiple types of sensor synthesis that the Tesla doesn’t even come close to having with a plain old video camera.
Tesla does an extremely dumb and basic depth estimation from two cameras. It still lacks many senses that are critical to movement, and has software that about matches the abilities of a small bird
YouTube can't provide enough information to make a determination about the reliability of a safety critical system. The video that ends up posted could be the best result of multiple attempts and many of the "testers" use hints like pressing the gas pedal or hitting the turn stalk to give the system more confidence in maneuvers that it's otherwise having trouble with. Further, many users who post bad experiences get negative feedback from their audience and sometimes even delete the video and DMCA people who re-host it.
Beyond the problems with YouTube as evidence, its very clear that Tesla isn't operating a proper safety lifecycle as part of the beta and its not clear that their approach and their current hardware is even capable of solving the remaining problems with FSD. Having already sold the hardware as FSD capable, Tesla has created a problem where the hardware drives the validation process of the system, rather than validation process driving the hardware. This is a fundamentally unsafe approach to safety critical systems.
The "media problem" is completely earned by Tesla. You know how you avoid bad press? By consistently under-promising while over-delivering and bending over backwards to keep your customers loyal and happy even after they've handed over their money so that when they need a new car they return to you.
German automakers are a great example of this. Whether it's gas or electric they significantly understate the performance and efficiency of their cars while in the real world they always exceed what they've promised.
FWIW we have a 2018 Model 3 and a 2022 Model Y and both have had very minor or no manufacturer issues. Both have never needed anything other than minor, at home service from Tesla (mostly tire swaps).
A single example does not make for a general rule. Aside from the fact that thank fucking god a less than 1 year old car and a less than 4 year old car should not have some manufacturer issues, Tesla still has a dreadful 226PP100 (as in, each Tesla tested had on average 2.6 problems in a single year), even worse than the industry average (which includes some truly terrible constructors).
These are improving at an astonishing rate, like a Web app rather than a car. Yes, this means the first ones were so bad they really shouldn’t have been shipped. But the newer models (of Y especially, X hasn’t been redesigned) are really a different world.
Hot tip, ship fast and break things doesn't work when you're talking about putting out a 2 ton, gigantic battery out in the world that is expected to last for over 20 years. This is a gigantic waste. Any other company would have been crucified for putting out such a shit car, and the only reason Tesla survived it is because they had funny meme man to do damage control.
> since 2014, and nearly a decade later they aren't that much closer
That is just an absurd statement. AlexNet came out in 2012 and ResNet in 2015 but somehow Tesla had something "close" to the current FSD in 2014? They're heavily using Transformers now which came out in 2017. They have demonstrated leaps in performance since the first version of FSD Beta two years ago.
No it isn't. It is also not early days either and we have given FSD enough time to meet the claims of achieving Level 5 with those strong and 'confident' deadlines even set by Musk himself, which he knew he couldn't meet, but the gullible customers and fans rushed in and fell for it anyway.
Not only they have all been missed, this year, Tesla finally admitted that it was Level 2 and it still doesn't work in the most dangerous time to drive which is at night. FSD (Fools Self Driving) is no where near close in the nearly ten years it has existed.
This is an obvious over-promising of a broken product with the pied-piper selling it to their fans scam.
Everyone from the regulators to Musk himself have known that for years that it is more than just liability and none of the claims of achieving 'Level 5' FSD and the hype around the robotaxis have ever materialized with many customers falling for the scam lead by the pied piper (Musk) and once again increased prices on a broken product.
The driver has to be liable, until the L5 system has lower error-rate than a human. For regulatory and liability reasons, it should be registered as an L2 until it's perfect. No one has ever claimed that FSD is a finished product.
> The driver has to be liable, until the L5 system has lower error-rate than a human.
Which that was promised to be completed in 2020. Where are the Level 5 robotaxis that Elon was confident enough to release by that year?
> For regulatory and liability reasons, it should be registered as an L2 until it's perfect.
And FSD was advertised as Level 5 by Tesla and was due to be complete by 2020 deadline. Even before that, many customers fell for this and were promised that future updates would eventually make it Level 5.
See how the scam works? Not only it was meant to be 'Level 5' as advertised by 2020, Tesla knew that it won't get there all along and continued the deception as their fans bought into the false promise and customers realised that they are paying for a broken contraption that didn't function as advertised and will continue paying for it as the price keeps increasing.
> No one has ever claimed that FSD is a finished product.
Yes. Elon (and Tesla) were the only ones that claimed and promised that it will be finished with robo-taxis on the road, at Level 5 FSD by 2020. This is the pied-piper scam and they have sold a false promise to their loyal fanbase and they will pay for an unfinished product until its 'perfect Level 5 FSD'; whenever that is and will continue to raise prices.
I also don't think it is ok to use unfinished safety critical software on the roads and put many driver's on the roads lives at risk.
FSD is a level 5 system, but it just has too high error rate to give it full control. It's a beta software for a level 5 system.
Level 5 means that the system is designed for autonomous driving in every situation. It's not a metric whether the system actually works. There will be multiple level 5 systems on the road some day, and some of them will make mistakes more than others. Just like all technology and software, it's impossible to make these systems perfect.
Yes, Elon overpromised and they failed to deliver in time. This was compensated by cheaper prices, just like is usual in beta programs. Today, FSD is already insanely good and can drive 99% of time without human intervention.
It's still not good enough, and it's multiple years late from original forecasts. Still, I wouldn't call it a scam. It's a delayed project with a pre-order/beta program, but given it's huge complexity and importance, I'm willing to forgive them. When it's ready, no one cares that it was a few years late.
> FSD is a level 5 system, but it just has too high error rate to give it full control. It's a beta software for a level 5 system.
Level 5 defines that the system is so reliable that there is zero intervention of the driver and that there is no need for them to be fully attentive behind the wheel as defined by the SAE [0]; exactly the requirement of a robo-taxi. This was advertised by Tesla with claims of robo-taxis by 2020, with a fake taxi app for marketing the scam.
> Yes, Elon overpromised and they failed to deliver in time.
And yet he continues to play the pied piper with more false promises. Like these claims and promises:
"Tesla's Full Self-Driving tech will have Level 5 autonomy by the end of 2021" [0]
"FSD will be capable of driving the car anywhere under any conditions with no need for human interaction." [0]
So it is now 2022 and they finally admitted it is Level 2 and requires the full attention of the driver having their eyes on the road at all times. Another false promise.
> Today, FSD is already insanely good and can drive 99% of time without human intervention.
Even the Tesla fans and FSD beta testers dispute this vacuous claim and that figure is significantly lower, given that it completely doesn't work in the most dangerous time to drive which is at night and in other conditions as well. As long as it still requires the driver to be always attentive behind the wheel with FSD on, it is not "Level 5 Full Self Driving" which was advertised by Elon Musk and Tesla but it is absolutely Level 2.
> It's still not good enough, and it's multiple years late from original forecasts. Still, I wouldn't call it a scam.
You need to snap out of the pied piper's scam and realise that FSD is an on-going scam, compensating their false promises for more price increases and forever promises.
Progress this year has been pretty amazing. If they didn't progress or didn't try to make it work, then I'd call it a scam.
You're not reading that SAR chart right. If you look at the features and examples, FSD is level 5 in there. That's what the system is designed to do. It's not a lane assistant or a traffic jam chauffer. Again, it's level 2 only because it's beta and requires supervision, and the driver is liable for any mistakes.
> Progress this year has been pretty amazing. If they didn't progress or didn't try to make it work, then I'd call it a scam.
It should have been complete years ago in 2020, at Level 5 FSD; so what do you mean about 'progress'? After missing more deadlines and claims, the price of FSD increases again despite the multiple malfunctions from lots of testers.
One example of a devoted Tesla fan / investor (Whole Mars Catalog) doing controlled experiments and driving around the same San Francisco street(s) is hardly a generic case of the whole situation.
> You're not reading that SAR chart right
I have read it correctly and Tesla FSD is still not Level 5 as advertised and by definition of the SAE chart.
Even in the video the driver is still paying full attention and has to continuously intervene by touching and guiding the steering wheel. They still have their attention and their eyes on the road which is not required for a Level 5 robot-taxi FSD system supposedly due for 2020.
So Tesla knows they are selling a Level 2 system still requiring the full attention of the driver and still not even close to the Level 5 robo-taxi claims, but was marketed as Level 5 anyway to mislead their customers with the help of their loyal fans doing videos of controlled experiments to fool new customers into purchasing an unfinished product.
Given it frequently malfunctions, the generic excuse is: ‘With each update, it gets better’,‘There is potential, buy it now before the price increases again’ or ‘It will get better soon, its early days’.
It is clearly not early days and we have given it enough time to be Level 5 FSD and it still isn't close even if they were given up to this year to finish it. So it is definitely a piped piper scam selling false promises to their customers. If this was a different company, there would be class action lawsuits all over this.
When I buy my next Tesla wouldn’t pay for it though.
The thing that would make it worth buying at this price is when I can send the car 200 miles away to like go pick up my parents or something. Anything short of not being in the car though is just a party trick.
The beta is out to ~100k cars in NA and growing. Not sure if they’ve sold the package elsewhere in the world but I wouldn’t bank on it coming any time (or car) soon outside of NA.
This is ultimately why I believe Elon Musk isn't the right guy to lead Tesla anymore and hasn't been for a while.
Car companies live and die by attracting new customers but also by bringing back existing customers. Lots of car brands bend over backwards to keep you coming back. When you fail to deliver what you've promised or significantly raise prices without delivering value in return, you're not building a loyal customer base.
Between this kind of thing and the way Tesla handles service, it's very clear they don't really give a damn about you as a customer after you've handed over your money.
This is my take and why I’ve never owned one. I’m in process of buying 2 new “luxury” cars for our family and Tesla’s not even in the realm of consideration.
Musk must be getting pretty desperate as his lawyers tell him he is going to have to follow through on the Twitter acquisition and his financial analysts are predicting the market will not be where he was hoping when it happens.
I have a tern, it's amazing. I can bike at 30mph, most of my errands within a ten mile trip are doable on bike. I do grocery shopping, cargo delivery, a bunch of stuff. It doesn't replace a car for all trips, but you'd be surprised how many trips it replaces.
It doesn't look like it works for any kind of cargo movement. It does not look like it can handle furniture, bags of groceries, or open boxes. A car does all those things.
I have panniers and a cargo rack on the front and back, I've carried six or seven bags of groceries without having to think hard. I've carried a car seat and two large plastic totes at the same time with a bit of Bungie cording.
I recently rented a 2022 Model 3 from Hertz for 9 days. It didn't have the FSD option available, but it did have the preview of what the system would see. Based on that alone, I was alternately impressed and terrified of what the FSD experience would be like. It was mostly accurate in showing other vehicles, but some of their trajectories when turning in front of me looked like a really bad video game with weird orientations and jumpiness. People crossing the road were sometimes picked up and sometimes not. It showed the traffic lights, but didn't display any understanding of turn lane traffic lights. I had the warning for following too close enabled, and it mostly worked but it got confused going through a parking lot once. If this is meant as an advertisement for FSD, I'm not convinced.
The Autopilot visualization is different than the FSD visualization, and the FSD UI shows significantly more detail. Go search YouTube for some FSD videos and you’ll see what it looks like.
Autopilot has basically been shelved for the past 2 years as they've been working on FSD, which will replace the current AP stack. Go have a look on Youtube, there are thousands of videos where you can see the visualizations, it's night and day.
That's crazy then. Is the hardware different, or just the software? I was under the impression that FSD was just basically paying for the software feature. It definitely looks much more dialed in, but still there's a lot of weird glitches with object permanence and physics failures that make me wonder how well this will always work.
I was a little concerned with how fast the car was going on Lombard when the pedestrian darted across. It worked out ok, but if the person had tripped, or dropped something and stopped to pick it up, it would have been harder to stop.
I skimmed through the whole video, but I didn't see any traffic light recognition of anything more than red, amber, green. Are there videos with more complex traffic lights? It's interesting that they managed to drive in that area without going through any of the weird intersections with cable car or Muni traffic lights.
There are 100k people beta testing this, so you'll find any kind of situation imaginable on Youtube, whether downtown NYC, torrential downpour, weird intersections etc. Many Youtubers find edge-cases or difficult routs and then drive that same route for every version to see how it improves (or not). Chuck Cook's famous unprotected left turn is one of those that got (mostly) fixed this release.
> Is the hardware different, or just the software?
It's the same hardware, just software and deep learning improvements. There are higher resolution cameras and better inference hardware coming out, but they're claiming it's not necessary for full autonomy. My guess is eventually they'll hand it out for free just to avoid supporting old hardware. The current cameras are quite low-res.
> Is the autopilot preview meant to help sell FSD?
No idea, but I hope they change it to FSD visualizations soon. For a while they've been trying to join the stacks for freeway (AP), city streets (FSD) and parking into one, the famed version 11. Since AP has been using this old stuff that's probably why they use these visualizations.
And to be clear, this is definitely not done, requires constant supervision. But it's improving pretty fast.
What's the rationale for renting an EV? Isn't it more troublesome because you probably won't have a charger in whatever place your staying, so you can't charge it overnight?
Well, our regular ICE rental car (2020 Hyundai Elantra) got stolen, likely using a low tech viral method [0]. We went back to the rental company and they wouldn't replace the stolen car until we had a police report. We needed a car, and the options at the airport where we rented were limited. We could get a pickup truck or a model 3 for about the same price. We opted for the model 3 to try it out. There was only one supercharger in the area, but it was close to our hotel, so it wasn't bad. I'd been wanting to try a Tesla, so despite the crappy circumstances, it seemed like a silver lining.
I was in Sweden a couple of months ago and rented an EV. In my case the reasons were simple. The main reason is that EV's are just nicer to drive. The second reason is to save money since petrol is much more expensive than charging an EV.
"It's all hype," said Steven E. Shladover, a retired research engineer at the University of California, Berkeley who has been involved in efforts to create autonomous driving for 45 years. "The technology does not exist to do what he is claiming. He doesn't have it and neither does anybody else."
---
It's not close to working at all. Musk must know it.
I used to think FSD without lidar was just too far away (a decade+), but I've been checking every now and then of FSD beta progress for a couple of years now, and at the current rate of progress I changed my mind and I believe that in 2 or 3 years, and with one more hardware upgrade, it will be done.
Their latest 10.69 FSD (released yesterday) is quite impressive already.
I wouldn’t have an issue with this if the upgrade went with the car, but the idea that it turns back off when the person whose driveway it lives in changes is too close to IAPs for me.
If it has an ongoing cost for Tesla, I’m happy to pay for a subscription and perhaps an activation fee. If it’s built-in to the vehicle and requires a configuration for how I like to drive, I’d love to pay an expensive, one-time setup fee to configure my specific vehicle and perhaps a (small) upgrade fee when major versions are released.
But this idea that Tesla wants to have the best of both worlds (massive fees that reoccur) while I get the worst of both worlds (massive fees for a value that I can’t sell/transfer) is one of the two big reasons I can’t see getting a Tesla (even though I would love to).
I was not aware that they had started offering this, and probably ought to have researched a bit before posting. That being said, I think my critique of the non-transferable upgrade still stands. Thank you for sharing the link to the subscription page!
It WILL NOT follow you to the next car though. It’s VIN locked (which is BS, no consumer software is hardware locked like that in a normal computing space).
128 comments
[ 3.6 ms ] story [ 258 ms ] threadWe really need competition in this space (FSD) if we want prices to go down.
Feels like a good opportunity for a class action...
If they actually make it self-driving at a human level, it's value is clearly over $100K.
If it's a sure thing, companies (including Tesla) will buy many cars and run the value as low as they can.
edit: typo
The first part is the general assumption that the endgame of self-driving cars is robotaxis. I don't personally buy this--most people who own their own cars today clearly prefer to own their own cars over relying on taxi services or public transit--but this is dominant assumption you see in newspaper articles and business planning. And I suspect it's mostly driven by greed for recurring service revenue over one-time sales revenue.
And the second part is the aspersions to Uber (as a stand-in for all the similar services): renting your own car out as a robotaxi is a lot like being an Uber, except for being the driver part. Of course, Uber itself is basically a losing model that doesn't really compensate enough to cover the wear-and-tear on your car from the extra mileage acting as a taxi, so it's not entirely the best model to be emulating. But how could a $10 billion company possibly be wrong?
You are misunderstanding the comment you are replying to.
It doesn't say "Tesla utilises the car which it sold to you as a taxi". It says that Tesla doesn't sell you a car. And they don't sell you a car because they can make better money by using their factory output as taxis, than by selling it to you.
Which means it won't be your car.
Before ~6 months ago it was literally vaporware with an nontransferable license.
But in reality: 'Fools Self Driving'
It was advertised by Tesla and as a Level 5 system. However, given that never happened, it is now known as 'Fools Self Driving' since not only it doesn't work, requires constant full assistance and attention of the driver behind the wheel at all times, it is been admittedly a Level 2 system who's creator gives the endless promises of updates in order to fool the customer that it will 'eventually' be 'Full Self Driving'.
So Tesla can continue to raises prices on this false promise and customers will keep buying, subscribing on the idea that they are hoping that it will be fully self driving, which is the typical pied piper selling snake oil scam.
It is gotten so bad that even the noble and devoted Telsa fans would do anything to defend this contraption, even if it meant risking the lives of their own children, they will still do it for Twitter points, or for their lord and saviour techno-king Elon to notice them.
Who is going to finally tell the Tesla FSD fans that the piped piper (Elon) has repeatedly over promised and scammed his own fans and customers using FSD to the same tune?
Between the low reliability, low build quality, and scammy practices, Tesla has turned their lead into a vulnerability.
Once other manufacturers catch up to the battery range, Tesla will have their lunch eaten.
And I say this as someone who has owned a Tesla for over 8 years.
It's possible after being "less than a year away" for... a decade... that it may be less than a year away now. But it looks like a bit of skepticism has been earned.
Sure, be skeptical, but then maybe say something like "I have no evidence for this, but I believe his manifest enthusiasm can only be explained by the fact that he's trying to manipulate the stock price to his own benefit."
Sure, it sounds less rhetorically convenient, but it's more honest.
I think this is the video I'm remembering, but he reused charts like this a lot https://youtu.be/p0kFZ9CVLNg?t=512
see this for some real-world testing, there's lots like it and the results usually skew a certain way for each car. for example, there's about a 20% delta in epa vs. tested range for the tesla y LR and mach-e standard range in his chart. the tesla artificially looks 20% better! and this delta was well-known at the time of his video https://insideevs.com/reviews/443791/ev-range-test-results/
to your question about conflicts of interest, owning a stock is a conflict of interest. nobody needs to show that something is happening inside his head, the circumstances exist and that's it. I think he's usually pretty good and I still watch the videos, and he trades on his reputation and I still regard him as a good source. but I'm still glad to learn that he's sold the stock
But also, the older Tesla's do not and cannot have the hardware for FSD so they really are EoL in that respect.
Plus, there's accidents, etc, that weed out a fair bit of the fleet before 8 years.
But I'm also perfectly happy driving my decades old Honda so what do I know.
It’s like comparing the productivity of a med student on Adderall in a quiet library, and the productivity of a peeled russet potato sitting in the sun.
As for build quality, we haven't had any major problems with our Model 3. It's a great car.
Tesla has a media problem, where every little problem gets reported as if it's systemic and unfixable.
It's like hiring a plumber who can only flush toilets, not install them. Self driving in pristine conditions is a party trick. Until it can do the hard part, it's nowhere near human-level.
They knew dojo wouldn't be online by then and they didn't even really train on video at that point, just trying to judge the scene image by image instead of taking into account prior frames (like knowing someone walked behind a bush 1p frames ago, so that sliver of emerging leg from behind a bush that might not be detectable confidently enough from a single image should be weighted higher).
See this example, where the map data is completely wrong: https://www.youtube.com/watch?v=4gTV9SlutU0
Does it actually work in the real world is the yardstick to apply.
Similar to Apple, people expect more from them than anyone else. Minor issues on Apple products are always scandals. Remember Antennagate? People were sure that it would sink Apple, because "they don't know how to make antennas". People and media handle Tesla similarly, and minor issues are blown out of proportion.
And now it looks like Cruise and Waymo have full self driving in San Francisco which looks amazing. There’s hour long videos of it navigating perfectly and smoothly in a jam packed city. Nothing like the Tesla videos which show it almost slamming into bikes/curbs or just freaking out and stopping in the middle of an intersection.
Has great acceleration though and charging at home is great or on the road with superchargers. But Teslas edge is not as wide anymore.
Who could have predicted that just relying just on cameras was not a good solution ? Definitely not the hundreds of researchers that told Musk that a LIDAR was necessary.
Beyond the problems with YouTube as evidence, its very clear that Tesla isn't operating a proper safety lifecycle as part of the beta and its not clear that their approach and their current hardware is even capable of solving the remaining problems with FSD. Having already sold the hardware as FSD capable, Tesla has created a problem where the hardware drives the validation process of the system, rather than validation process driving the hardware. This is a fundamentally unsafe approach to safety critical systems.
German automakers are a great example of this. Whether it's gas or electric they significantly understate the performance and efficiency of their cars while in the real world they always exceed what they've promised.
FWIW we have a 2018 Model 3 and a 2022 Model Y and both have had very minor or no manufacturer issues. Both have never needed anything other than minor, at home service from Tesla (mostly tire swaps).
These are improving at an astonishing rate, like a Web app rather than a car. Yes, this means the first ones were so bad they really shouldn’t have been shipped. But the newer models (of Y especially, X hasn’t been redesigned) are really a different world.
People put up with the low build quality because at the time Tesla had a unique product not because of some memes.
That is just an absurd statement. AlexNet came out in 2012 and ResNet in 2015 but somehow Tesla had something "close" to the current FSD in 2014? They're heavily using Transformers now which came out in 2017. They have demonstrated leaps in performance since the first version of FSD Beta two years ago.
No it isn't. It is also not early days either and we have given FSD enough time to meet the claims of achieving Level 5 with those strong and 'confident' deadlines even set by Musk himself, which he knew he couldn't meet, but the gullible customers and fans rushed in and fell for it anyway.
Not only they have all been missed, this year, Tesla finally admitted that it was Level 2 and it still doesn't work in the most dangerous time to drive which is at night. FSD (Fools Self Driving) is no where near close in the nearly ten years it has existed.
This is an obvious over-promising of a broken product with the pied-piper selling it to their fans scam.
Yes. Admitted. [0]
> The difference between L2 and L5 is liability.
Everyone from the regulators to Musk himself have known that for years that it is more than just liability and none of the claims of achieving 'Level 5' FSD and the hype around the robotaxis have ever materialized with many customers falling for the scam lead by the pied piper (Musk) and once again increased prices on a broken product.
[0] https://www.news18.com/news/auto/teslas-full-self-driving-cl...
Which that was promised to be completed in 2020. Where are the Level 5 robotaxis that Elon was confident enough to release by that year?
> For regulatory and liability reasons, it should be registered as an L2 until it's perfect.
And FSD was advertised as Level 5 by Tesla and was due to be complete by 2020 deadline. Even before that, many customers fell for this and were promised that future updates would eventually make it Level 5.
See how the scam works? Not only it was meant to be 'Level 5' as advertised by 2020, Tesla knew that it won't get there all along and continued the deception as their fans bought into the false promise and customers realised that they are paying for a broken contraption that didn't function as advertised and will continue paying for it as the price keeps increasing.
> No one has ever claimed that FSD is a finished product.
Yes. Elon (and Tesla) were the only ones that claimed and promised that it will be finished with robo-taxis on the road, at Level 5 FSD by 2020. This is the pied-piper scam and they have sold a false promise to their loyal fanbase and they will pay for an unfinished product until its 'perfect Level 5 FSD'; whenever that is and will continue to raise prices.
I also don't think it is ok to use unfinished safety critical software on the roads and put many driver's on the roads lives at risk.
Level 5 means that the system is designed for autonomous driving in every situation. It's not a metric whether the system actually works. There will be multiple level 5 systems on the road some day, and some of them will make mistakes more than others. Just like all technology and software, it's impossible to make these systems perfect.
Yes, Elon overpromised and they failed to deliver in time. This was compensated by cheaper prices, just like is usual in beta programs. Today, FSD is already insanely good and can drive 99% of time without human intervention.
It's still not good enough, and it's multiple years late from original forecasts. Still, I wouldn't call it a scam. It's a delayed project with a pre-order/beta program, but given it's huge complexity and importance, I'm willing to forgive them. When it's ready, no one cares that it was a few years late.
Level 5 defines that the system is so reliable that there is zero intervention of the driver and that there is no need for them to be fully attentive behind the wheel as defined by the SAE [0]; exactly the requirement of a robo-taxi. This was advertised by Tesla with claims of robo-taxis by 2020, with a fake taxi app for marketing the scam.
> Yes, Elon overpromised and they failed to deliver in time.
And yet he continues to play the pied piper with more false promises. Like these claims and promises:
So it is now 2022 and they finally admitted it is Level 2 and requires the full attention of the driver having their eyes on the road at all times. Another false promise.> Today, FSD is already insanely good and can drive 99% of time without human intervention.
Even the Tesla fans and FSD beta testers dispute this vacuous claim and that figure is significantly lower, given that it completely doesn't work in the most dangerous time to drive which is at night and in other conditions as well. As long as it still requires the driver to be always attentive behind the wheel with FSD on, it is not "Level 5 Full Self Driving" which was advertised by Elon Musk and Tesla but it is absolutely Level 2.
> It's still not good enough, and it's multiple years late from original forecasts. Still, I wouldn't call it a scam.
You need to snap out of the pied piper's scam and realise that FSD is an on-going scam, compensating their false promises for more price increases and forever promises.
[0] https://www.sae.org/binaries/content/assets/cm/content/blog/...
[1] https://www.cnet.com/roadshow/news/elon-musk-full-self-drivi...
Progress this year has been pretty amazing. If they didn't progress or didn't try to make it work, then I'd call it a scam.
You're not reading that SAR chart right. If you look at the features and examples, FSD is level 5 in there. That's what the system is designed to do. It's not a lane assistant or a traffic jam chauffer. Again, it's level 2 only because it's beta and requires supervision, and the driver is liable for any mistakes.
It should have been complete years ago in 2020, at Level 5 FSD; so what do you mean about 'progress'? After missing more deadlines and claims, the price of FSD increases again despite the multiple malfunctions from lots of testers.
One example of a devoted Tesla fan / investor (Whole Mars Catalog) doing controlled experiments and driving around the same San Francisco street(s) is hardly a generic case of the whole situation.
> You're not reading that SAR chart right
I have read it correctly and Tesla FSD is still not Level 5 as advertised and by definition of the SAE chart.
Even in the video the driver is still paying full attention and has to continuously intervene by touching and guiding the steering wheel. They still have their attention and their eyes on the road which is not required for a Level 5 robot-taxi FSD system supposedly due for 2020.
So Tesla knows they are selling a Level 2 system still requiring the full attention of the driver and still not even close to the Level 5 robo-taxi claims, but was marketed as Level 5 anyway to mislead their customers with the help of their loyal fans doing videos of controlled experiments to fool new customers into purchasing an unfinished product.
Given it frequently malfunctions, the generic excuse is: ‘With each update, it gets better’,‘There is potential, buy it now before the price increases again’ or ‘It will get better soon, its early days’.
It is clearly not early days and we have given it enough time to be Level 5 FSD and it still isn't close even if they were given up to this year to finish it. So it is definitely a piped piper scam selling false promises to their customers. If this was a different company, there would be class action lawsuits all over this.
You need to really snap out of it.
Yes, it's a much better toy, but it's still a toy.
And people still buy it because it's the best money can buy right now. It's not perfect but it's there and getting better with every release.
When I buy my next Tesla wouldn’t pay for it though.
The thing that would make it worth buying at this price is when I can send the car 200 miles away to like go pick up my parents or something. Anything short of not being in the car though is just a party trick.
Basically it's an open beta, if you pass their minigame test for 30 days, you can get in and have FSD.
Car companies live and die by attracting new customers but also by bringing back existing customers. Lots of car brands bend over backwards to keep you coming back. When you fail to deliver what you've promised or significantly raise prices without delivering value in return, you're not building a loyal customer base.
Between this kind of thing and the way Tesla handles service, it's very clear they don't really give a damn about you as a customer after you've handed over your money.
(I own FSD on two cars because I like toys)
https://www.ternbicycles.com/en/bikes/472/gsd
Unless you have a trailer, I guess?
To save you the click, here's one released today with the latest version: https://www.youtube.com/watch?v=jCTssX2VdKA
I was a little concerned with how fast the car was going on Lombard when the pedestrian darted across. It worked out ok, but if the person had tripped, or dropped something and stopped to pick it up, it would have been harder to stop.
I skimmed through the whole video, but I didn't see any traffic light recognition of anything more than red, amber, green. Are there videos with more complex traffic lights? It's interesting that they managed to drive in that area without going through any of the weird intersections with cable car or Muni traffic lights.
Is the autopilot preview meant to help sell FSD?
> Is the hardware different, or just the software?
It's the same hardware, just software and deep learning improvements. There are higher resolution cameras and better inference hardware coming out, but they're claiming it's not necessary for full autonomy. My guess is eventually they'll hand it out for free just to avoid supporting old hardware. The current cameras are quite low-res.
> Is the autopilot preview meant to help sell FSD?
No idea, but I hope they change it to FSD visualizations soon. For a while they've been trying to join the stacks for freeway (AP), city streets (FSD) and parking into one, the famed version 11. Since AP has been using this old stuff that's probably why they use these visualizations.
And to be clear, this is definitely not done, requires constant supervision. But it's improving pretty fast.
[0] https://www.today.com/news/news/kia-hyundai-vehicle-theft-in...
https://www.chicagotribune.com/business/ct-biz-tesla-self-dr...
"It's all hype," said Steven E. Shladover, a retired research engineer at the University of California, Berkeley who has been involved in efforts to create autonomous driving for 45 years. "The technology does not exist to do what he is claiming. He doesn't have it and neither does anybody else."
---
It's not close to working at all. Musk must know it.
Their latest 10.69 FSD (released yesterday) is quite impressive already.
If it has an ongoing cost for Tesla, I’m happy to pay for a subscription and perhaps an activation fee. If it’s built-in to the vehicle and requires a configuration for how I like to drive, I’d love to pay an expensive, one-time setup fee to configure my specific vehicle and perhaps a (small) upgrade fee when major versions are released.
But this idea that Tesla wants to have the best of both worlds (massive fees that reoccur) while I get the worst of both worlds (massive fees for a value that I can’t sell/transfer) is one of the two big reasons I can’t see getting a Tesla (even though I would love to).
They do have a subscription based model https://www.tesla.com/support/full-self-driving-subscription...
It WILL NOT follow you to the next car though. It’s VIN locked (which is BS, no consumer software is hardware locked like that in a normal computing space).
The guy just got a free $7500 courtesy of the tax payer for every car he sells.