Tesla tells investors it’s being investigated by the Justice Department
arstechnica.com
arstechnica.com
But whether it's due to the embarrassment of being swindled or being drunk on the Kool-Aid, the amount of Tesla owners I know that tell me they are happy for paying up to $15k for what is essentially cruise control, and who honestly believe 'real FSD is only a year away' is shocking.
Hmm, 285,000 [0] sales of FSD * say $8,000 [1] to be conservative = $2,280,000,000
That's more than I expected.
[0] https://electrek.co/2022/12/29/tesla-people-bought-full-self...
[1] https://www.notateslaapp.com/tesla-reference/958/tesla-fsd-p...
Also maybe I'm just proving your point by saying that there is one option.
I write code. I notice bugs. I don’t want an in dash embedded device with a poorly designed ui/x.
I want something I can wrench on and when it breaks I can debug and fix.
Currently in a 2015 wrangler. Auto trans, manual locks, manual windows, $90 marine radio with shit bluetooth.
I’d get a cassette deck if I could bluetooth to it.
Point being, where I am, the car is what I need and can deal with my roads and winters and doesn’t have an OTA updated dash-to-drivetrain link that can decide at any time heated seats or cruise control is 30 bucks a month.
This was a much more lucrative proposition as a $2k add-on, I think.
However, the Enhanced Autopilot option was removed in 2019 and came back last year. I think this was done to juice the sales count. If the DoJ agrees, then Tesla might have to honor EAP's 2019 purchase price. This would make the minimum refund something like $3500.
Either way, this would be a blow to Tesla and will definitely worsen the layoffening.
Acting like you know better than actual customers of FSD whether the $8K, $10K, $12K, or $15K they spent on the software is reasonable for them is the height of arrogance.
Also, characterizing the current FSD Beta system as "cruise control" is delusional.
Again, arguing that you know better about what other people do with their money is arrogant.
Full Self-Driving Capability
$15,000
All functionality of Basic Autopilot and Enhanced Autopilot
Traffic Light and Stop Sign Control
Coming Soon
Autosteer on city streets
The currently enabled features require active driver supervision and do not make the vehicle autonomous. The activation and use of these features are dependent on achieving reliability far in excess of human drivers as demonstrated by billions of miles of experience, as well as regulatory approval, which may take longer in some jurisdictions. As these self-driving features evolve, your car will be continuously upgraded through over-the-air software updates.cmd-f "beta"
0 results
hmmm...some further questioning is warranted.
So, for the people actually buying it with the actual contract paperwork and the sales pitch being provided, does the word Beta ever get used and is it fully explained?
to me, it looks like another "smear the DoJ for doing their job" campaign by those having the job done to them. it is so in vogue these days
Full Self-Driving Capability
Navigate on Autopilot: automatic driving from highway on-ramp to off-ramp including interchanges and overtaking slower cars.
Autopark: both parallel and perpendicular spaces.
Summon: your parked car will come find you anywhere in a parking lot. Really.
Coming later this year:
Recognize and respond to traffic lights and stop signs.
Automatic driving on city streets.
If you pay $10K for a paint color, you get the paint color. If you pay $20K for sport brakes, you get the brakes. If you pay $15k for Tesla FSD, you get the hope that Tesla will eventually deliver FSD, much the same as a Star Citizen customer is buying the hope that they'll eventually actually be able to fly their ship in a full game.
You can still argue that the aspiration is worth the money to the customers, or that they're already getting enough value out of the partially delivered product. But I also think it's reasonable for onlookers to think that there's some level of fraud going on, given the constant promises without delivery.
We are talking about the group of people who system16 criticized for being "happy with having paid for FSD." Other peoples' needs and wants differ, and characterizing those needs and wants as if they are plain wrong is arrogant.
In the end, who knows? Who cares? If a customer says they are happy with the purchase, why should someone else tell them they are plain wrong? I can disagree, of course, but I should not tell them they're just wrong for being happy.
I don't tell people they're wrong for paying extra for a special paint color.
That's really my problem. You get used with it and start trusting it until it unexpectedly goes straight into that white van.
The trouble is, nobody can afford to employ a hundred thousand professional test drivers.
> You can still argue that ... they're already getting enough value out of the partially delivered product.
I think it's true that some people feel that way, and that's fine. But I agree with OP that the others "who honestly believe 'real FSD is only a year away'" are being taken advantage of if not outright defrauded.
Agreed, most cruise control systems are considerably better. My cruise control has never tried to swerve me into oncoming traffic.
By the same token, walking will avoid the risk of being in a highway accident. So by this logic, walking is superior to driving.
Yes, a system that doesn't try to do something is better than one that does it poorly.
> By the same token, walking will avoid the risk of being in a highway accident. So by this logic, walking is superior to driving.
Again we're on the same page, walking is superior to driving. Although the pesky cars often make it more dangerous and there are limits that make walking impractical at times.
But they get the brakes and they get the paint. What they get with supposed "full self-driving" is puffery and lies.
Musk's public statements around FSD since 2014 have been lies: https://jalopnik.com/elon-musk-promises-full-self-driving-ne...
In 2016 Tesla claimed that all Teslas being produced "will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver." That was a lie: https://www.tesla.com/blog/all-tesla-cars-being-produced-now...
And these days Teslas can't even do full self-parking: https://www.youtube.com/watch?v=nsb2XBAIWyA
Now the word is that suckers who paid for FSD in a Tesla with Hardware 3 probably won't get Hardware 4: https://www.thedrive.com/news/tesla-cybertruck-gets-new-self...
Tesla was recently done for false advertising around FSD: https://electrek.co/2022/12/12/tesla-ordered-upgrade-self-dr...
Tesla also admits they have failed to deliver full self-driving but they think they don't need to refund any money because "failure is not fraud": https://electrek.co/2022/12/07/tesla-self-driving-claims-fai...
Despite the admission of failure, Musk is still running around claiming that FSD will one day magically become the "biggest asset value increase of anything in history": https://electrek.co/2023/01/26/elon-musk-thinks-not-smart-do...
It's not even bullshit artistry. It's just bullshit.
>the amount of Tesla owners I know that tell me they are happy
Exactly
During a 2 hour drive on an interstate on a clear, sunny day, FSD started steering into another vehicle 3 different times. One was a truck, and another was a wide load house being transported. I'll admit that is an odd case, but "don't accelerate straight into a house that is at windshield level" seems like a pretty simple ask. It did this multiple times and I had to temporarily disable the cruise control.
It also failed miserably on curving state highways, to the point that I turned it off after about 20 minutes.
FSD is supposed to only work on city streets, and was in limited beta in 2021 (you had to explicitly enroll). Autopilot is what runs on freeways/interstates today, and can be used on the street (but is not meant for city streets).
Autopilot, and it's Navigate on Autopilot system have been fantastic for me, but they are absolutely systems you need to monitor because they can do dumb things. It's not meant to be a hands-off driving system like the FSD system Tesla claims it will be one day.
Would you call all of these "unbreakable wine glasses" fraud? https://www.amazon.com/s?k=unbreakable+wine+glasses I think we would all agree that not one of them is actually unbreakable. But if someone sued for fraud because they were able to break one, they would probably be laughed out of court.
Whereas for a "full" self driving car, Tesla obviously has not (and probably will not) deliver a car that can fully drive itself in the context of normal car usage.
I imagine the vendors of unbreakable wine glasses would probably be in trouble if they charged $15k and argued that they'd receive over the air updates to Kevlar as soon as the regulatory approvals had gone through...
So you think the Justice Dept. has a formal investigation ongoing for which the main argument hinges around the name Full Self Driving? Umm..okay.
>So long as everybody knows what they are paying for is not an actual fully autonomous driving system
The expectation is that it will be. And, I admit, it might be, one day. But at what point is selling some potential future promise fraud? Over an infinite horizon, you could get away with any claim.
At the minimum, it's not "fraud, plain and simple" because there are people that currently use it and are happy with it. And it's not just sunk cost fallacy because you can buy it as a subscription for $200/mo, which is much less of an investment, and you can cancel that at any time. I don't know if Tesla breaks out numbers of how many FSD subscribers there are, but I think that would be pretty interesting to know.
Regardless, for something in development like FSD, I think the subscription model makes the most sense, since you know kinda what you're getting and if it's not worth it to you, you can cancel and then try again later after some updates. I kind of think they bumped the price to $15k just to switch more people to the subscription model.
But paying 200/mo USD for practically being a crash test dummy in real driving conditions, I'm not so sure.
When I ordered the car, I did so expecting it to be able to drive while I slept sometime during the life of the car. I have considered suing Tesla for failing to live up to the promises that prompted me to purchase the car. I had never purchased a new car before and wouldn't have except for the phantasmal claims of full-self driving. Refunding the $6k I paid for FSD seems insufficient to me.
genuine question but , were you not the tiniest bit skeptical when only Tesla were able to promise FSD?
like what secret sauce have they got that no one else has?
Also, remember that self-driving cars were all the rage then too. Hype was at its peak then, I think.
A large fleet of networked cars that were built from the ground up to provide real-world data (including cameras and human driver inputs) for AI training.
Honestly I kind of wonder if Tesla's software team (or overall technical approach) wasn't very badly mismanaged since they seem to have thrown away a huge head start. Or maybe autonomous car AI is just really hard.
Keep in mind they are likely spending 10x or less than Waymo on FSD, so a lot less employees/researchers, have been around a lot less time and have decided not to use the sensors and ground truth data that Waymo relies on.
Just having lots of data isn't necessarily very useful, or we'd see the pure FSD trying to get their sensors on everybody's cars.
I don't think that happy customers necessarily mean that it's not fraudulent. FSD provides some value and has been seriously misrepresented by Tesla. The fact that some people are happy with the value they get from it does not diminish Tesla's misrepresentation.
Just because there are people who use it and are happy with it doesn't mean it's not fraud. There likely are people who are being sold fake products (unbeknown to them) and are still happy with those products, they were still sold to them under a fake premise. Regarding subscribers, sure there are people who are happy to pay $200/month for driving assist (I mean people pay significant premium for other makers driving assist technology), however I suspect there are a significant number of people where the promise of "autopilot" was a deciding factor to buy a Tesla over another car. So it is likely not just fraud, but also unfair competition.
Other lane-keeping systems do nothing but try to keep the car in the lane (and some do a poor job of even that). Autopilot is rock solid in this regard. Like "I now actually prefer road-tripping 700+ miles instead of flying" rock solid. (I never road-tripped in the past because driving long distances wipes me out. Not so with Autopilot.)
Navigate on Autopilot is also super reliable in automatically changing lanes and exiting highways, and works on highways that SuperCruise and BlueCruise Hands-Free do not.
Traffic Aware Cruise Control is rock-solid in detecting and acting upon traffic lights and speed limit signs, though some systems already do this. This used to require purchasing FSD (it was a $7,000 option when I purchased our Model 3), but now you just need Enhanced Autopilot to enable it (which is $6,000).
I can easily see people asking for at least a downgrade to Enhanced Autopilot (a $9,000 difference), and I don't blame them. For example, my wife would NEVER use FSD Beta in its current state, but she loves Autopilot and uses it all of the time.
Given that FSD is included in the final price of the car that is used for car registrations, I'm wondering how those refunds will work.
I am completely uneducated on the state of EV charging right now. But for a 700+ miles, you would have to fully charge 2+ times? And how does it compare to a gas car given the availability of superchargers, excessive demands on long weekends, etc? Not trying to dismiss EV because of these factors, just genuinely curious of the compromise.
However, gas stations tend to be very spartan. Maybe you'll be fueling up at Bucee's or Wawa or Publix or something like that, or maybe you'll be filling up at Rod's Country Store which is a murder house that happens to have gas. Superchargers are usually located at the back of malls (which are useless if closed, but very useful when they are not) or large stores. Sometimes they are right in the city center, which is great because you can walk around and explore while you charge. It forces a break, which is great during road trips.
On a 700 mile road trip, we'd probably stop three times or so. However, Tesla Navigation is really really good at route planning with its superchargers. The nav can tell you when a charger is full or about to be full or if it has broken stalls or not. It will actually re-route you if a supercharging station fills up, if possible.
This is simply not true. Drove recently in BMW, Audi, and Volkswagen and all had a very reliable adaptive cruise control + automatic lane change assist. I have two cars, one of them being Tesla, and Tesla does not have advantage in basic cruise control.
Though also he should still be raked over the coals.
>A scorpion wants to cross a river but cannot swim, so it asks a frog to carry it across. The frog hesitates, afraid that the scorpion might sting it, but the scorpion promises not to, pointing out that it would drown if it killed the frog in the middle of the river. The frog considers this argument sensible and agrees to transport the scorpion. Midway across the river, the scorpion stings the frog anyway, dooming them both. The dying frog asks the scorpion why it stung despite knowing the consequence, to which the scorpion replies: "I am sorry, but it is my nature".
Should we accept that the scorpion will always sting its associates, and continue to allow them to hitch rides on frogs? Or should we prevent the scorpions from getting into a position where they can sting others once their nature is readily apparent?
I think GP's position on that question is abundantly clear
This applies to FSD. Why risk your life testing software for the benefit of Tesla? Either FSD frogs are irrationally altruistic given that Tesla is a corporation, with wealthy shareholders, they are unable to judge the risk, or they are being mislead.
Given the large number of FSD frogs enforcement seems inevitable.
Sorry if this reply comes across as mean - not a criticism of you or anything, just personally dislike that way of viewing things.
The part that I feel is on everyone who bought into FSD is that literally every expert in self driving (A field I was in directly at the time) was pretty sure Tesla was never going to get there with there hardware. Tesla supporters were quite rude denying this.
For those unaware, Otto was an inflatable autopilot in the 1980 movie "Airplane!"[2]
[1] https://cdn.arstechnica.net/wp-content/uploads/2022/10/tesla...
[2] https://en.wikipedia.org/wiki/Airplane! (with exclamation mark)
I suspect the same will happen here, the problem here that the prosecution has is the signed purchase agreement is much more legally binding than a tweet. No where in the signed purchase agreement does it say "Statements from our CEO take precedent over this agreement", in fact it says the opposite basically.
That's how they've managed to keep this out of the courts for 10 years.
I think they're an augmentation that can make drivers better.
I also think Tesla's FSD is too lax and allows drivers to be worse.
The safety systems like lane keep, obstacle avoidance, and AEB are not exclusive to Tesla; and Tesla does them really really well.
Thats because FSD now costs more to buy, so these people still think they've got a good deal. It still adds more value to their car 2nd hand than they paid for it.
To do this is a really interesting business move. At first glance, they are throwing away revenue (because a 2nd hand car with FSD presumably sells for more than one without). But they might be able to sell it as an addon or a subscription to the new owner... But the chances of that are probably low to those buying 2nd hand cars.... But also by artificially restricting the numbers of cars available on the market with FSD enabled, those which do have FSD enabled will keep their value better... Which may in turn help with sales of the feature with brand new cars...
I bet there is a really big spreadsheet written by an MBA to make this decision...
Frankly it's far less of a scam than 80% of the crap that gets posted here on HN.
I'm wondering how sharing their internal stats will help - the claim is about whether "Tesla had misled investors and customers by making unsupported claims about the capabilities of its driver-assist features" - do you think their internal stats will prove that they thought their grandiose claims were supported by evidence? Otherwise, I don't know how that will help them.
There have been more than 3 billion miles worth of real-world driving using AutoPilot.
Tesla will almost certainly have to refund people's money, in part because they keep ratcheting up the hardware requirements and they can't bring all the people along from previous generations of hardware. There's also the issue of claiming funding was secured but it wasn't actually ready apparently, according to the infamous tweet from Musk.
Mercedes leapfrogging Tesla on granting a liability umbrella should seal the case on the latter having a moat in self-driving technology.
I don't think anything will upset the narrative that Tesla is the technology leader until it's completely obvious. Not Mercedes taking liability, not Waymo operating real driverless robotaxis, not independent comparisons of tech, not videos of FSD crashing into fake children, nothing.
> Not Mercedes taking liability, not Waymo operating real driverless robotaxis, not independent comparisons of tech, not videos of FSD crashing into fake children, nothing.
Argument is there is no moat. Mercedes is close enough to Tesla that it’s unclear who’s technologically ahead. Mercedes’ level III approvals and liability umbrella are materially ahead of Tesla. That makes FSD another feature, not a barrier to entry.
There was a hypothesis that a decade on, Tesla would have the only or unequivocally best self-driving electric car. (The way it had the unequivocally best electric car for several years.) Evidence is accumulating against that outcome.
Promising no input (until unexpectedly required) is much worse than being ready to take over.
Yes. But it’s more than Tesla has granted. And the car can still self drive in other situations (like a Tesla). You just don’t get the umbrella (like a Tesla).
This is a remarkable level of comparison for a company that only a few years ago looked set, if it focussed, on monopolising the EV supply chain.
Citation? The only thing remotely close I've seen was comparing apples and orange (Autopilot, mostly highway VS all miles without Autopilot).
For sure! All they have to do is release the data showing just how safe FSD is!
...why have they never done that?
Edit: I wasn't being particularly serious, just riffing on the parent comment. That said, there is a long game playing out around FSD.
How is having the leading tech a "trump card" that you don't tell people about? It's literally the best marketing on the planet, if true. More 6d chess from Elon?
> Ambassador de Sadesky : It was to be announced at the Party Congress on Monday. As you know, the Premier loves surprises.
In other words, what is the accident rate of non-AP Teslas on highway/freeway only versus the Tesla-AP rate? I'm assuming Tesla doesn't have an easy way to figure out weather at time of accident (of course, wouldn't be hard for a dedicated researcher). But accidents-per-highway-miles only shouldn't be burdensome to calculate and would provide a much more apples-to-apples comparison