California’s AV testing rules apply to Tesla’s “FSD” | Hacker News Reader

California’s AV testing rules apply to Tesla’s “FSD”

cyberlaw.stanford.edu

105 pointscodechicago277148 comments

> Tesla's "FSD" has the "capability to drive a vehicle without the active physical control or monitoring by a human operator," but it does not yet have the capability to do so safely. Hence the human drivers. And the testing. On public roads. In California. For which the state has a specific law. That Tesla is not following.

I couldn't find it spelled out in the article what portions of the law Tesla is not following. But upon reading it[0] it is clear Tesla could not possibly comply. There are requirements such as reporting in advance which roads the cars will be tested on. (I assume saying "all roads" would be a non-starter with the DMV.) And each test-driver needs an individual permit. So if it is found that Tesla's FSD falls under this law, they have to disable or severely limit FSD in California.

I found the argument in the article quite strong that indeed FSD falls under that law. So I'm wondering what will happen next. Will the California DMV dare anger those who sunk money into FSD? Or will the DMV just turn a blind eye until there's a crash with a stroller and that will force them to act?

[0] https://www.dmv.ca.gov/portal/file/autonomous-vehicles-testi...

The reason Tesla does not fall under this legislation seems pretty simple to me, but IANAL…

The law applies to vehicles with the "capability to drive a vehicle without the active physical control or monitoring by a human operator".

Tesla FSD does not have this capability. If the driver does not consistently demonstrate that they are actively monitoring the vehicle and in physical control, then the “AutoPilot(TM)” feature deactivates.

”An autonomous test vehicle does not include vehicles equipped with one or more systems that provide driver assistance and/or enhance safety benefits but are not capable of, singularly or in combination, performing the dynamic driving task on a sustained basis without the constant control or active monitoring of a natural person.”

So, in summary Tesla "fully self driving" does not actually have the capability to "fully self drive." It seems like the hazy marketing around this feature hasn't been a great idea.

I'd say the marketing has been perfect, the haze confuses both the consumers and regulators believe whatever they want to believe.

Other than the latest, bleeding edge beta ("FSD Beta") FSD is vaporware.

Autopilot on the other hand is a pretty good driver assist (lane keeping, TACC, etc.).

From one of the previous threads on FSD

> Tesla has so many versions of "FSD" or Full Self Driving because they keep shifting the post for what exactly is Full Self Driving.

> What even is FSD compared to SD, where did the need to tack on "Full" come from? And I bet it's because of them calling their lane assist cruise "Autopilot" which implies self driving.

This is just a semantic argument about what "capability" and "autonomous" means. The system is obviously capable of autonomously navigating in an engineering sense, because that's what it does.

The system as deployed ("FSD beta") also requires a human driver, and actively monitors for driver attention. So it is not actually operating autonomously.

It seems that the spirit of the law is to disallow unregulated testing of cars that are actually operating autonomously. Cars with a human in the loop don't really seem to be what the regulations were thinking of.

The letter-of-the-law argument would seem to disallow any sort of uncommanded control by the vehicle, including lane assist and 1970's cruise control. And that's not right either.

The article addresses that explicitly.

Basically, the DMV is saying "oi, if you're testing a system that's capable of driving by itself, you have to follow certain rules"!

Then you reply "No, we're still testing it, and it requires human supervision, so it's not capable of driving by itself, so the rules don't apply!"

> "FSD" is aspirationally an automated driving system. The name unequivocally communicates Tesla's goal for development, and the company's "beta" qualifier communicates the stage of that development. Tesla intends for its "full self-driving" to become, well, full self-driving, and its limited beta release is a key step in that process.

> Tesla's "FSD" has the "capability to drive a vehicle without the active physical control or monitoring by a human operator," but it does not yet have the capability to do so safely.

It’s that bit that seems like it would make Tesla’s actions fall even further under the DMV’s purview. Is their argument really going to be that they can’t be held to safety regulations because it’s too unsafe to be regulatable? That’s like saying you’ve stolen too much money to be guilty of robbing the bank.

I think the "loophole" here is that - despite the marketing claims - Tesla is NOT testing a system which is designed to run without active monitoring by a human. This is unlike Cruise, Waymo etc. who do fall under this law.

Marketing is admissible in court as evidence of intent.

That would be the case if they didn't also have controls in place to prevent the driver from disengaging while autopilot was active.

The marketing is defense against any claims that you bypassed a safety, since marketing is saying publicly the safety systems are not needed.

My company has lost in court about similar cases where the court decided that the marketing picture showing someone closer than the warning label says is safe means we don't believe the warning label and so we are liable when a customer gets too close to cutting blades.

are you suggesting any developer of autonomous driving systems can straight up ignore this law as long as they make sure the system requires the presence of a human?

Yes? Tesla isn't the only company making driver assist systems. I have the equivalent functionality in my Genesis right now and no one is making noise about it.

I suppose it depends on what "capable" means.

If you mean that a Tesla with FSD can be used to drive without human control, then that is definitely true since the tests for driver alertness can be bypassed.

If you mean that a Tesla with FSD has the level of technological sophistication needed to drive autonomously, that is also true: if you bypass the checks for driver alertness, the Tesla can continue driving (albeit not safely). This is well beyond the range of most "driver assistance" technologies.

If you mean that a Tesla can drive autonomously without bypassing any checks for driver alertness, then no, it cannot, but only assuming these checks never fail and are always sufficient to prevent it from being used without active monitoring.

Tesla's case seems weak here, in that it relies on an extremely narrow interpretation of the law, according to which a system that can drive autonomously but tries to prevent you from doing so is not considered truly autonomous.

But IANAL, so perhaps Tesla has a better case than I'd assume.

> Tesla's case seems weak here, in that it relies on an extremely narrow interpretation of the law,

To be clear: there is no case, nor has Tesla tried to make one. The linked article is an opinion essay by an academic in the field. The CA DMV hasn't said anything.

For myself: I'd argue that we have ~23k FSD beta vehicles on US roads now that have been driving for 3-4 months and that if we're really concerned about the safety of the system (vs. getting into semantic arguments about the statute language of some random law) we should just go get the data and see. By my math that's going to be something in the tens of millions of miles now. That should be well above the noise floor.

How are "we" going to go get the data? Tesla isn't exactly open about these things. (no, marketing numbers aren't enough)

The subject being discussed is literally the regulatory authority of the California Department of Motor Vehicles. You don't think they track accidents statistic?

I mean, if someone was coming along from the DMV and saying "here's a report showing a 95% confidence of a higher than average accident rate among FSD beta vehicles" then I'd be first in line to admit that, yes, we should probably halt deployment while we figure out what's going wrong.

But that's not what's happening here at all. The people with the data are silent, which tells me something that confirms my priors as a FSD beta tester: this system as deployed is safe. It is.

Tesla isn't complying with the law, they're keeping the data to themselves. How would the CA DMV be getting accurate statistics?

What form are random cops going to fill out for "Tesla fucked it again and this was FSD Beta version whatever" ? Or will cops exactly and reliably fill out accident reports in exactly the same way with impossible precision, and then who is going to notice? How many accidents will they miss? How many _near_ accidents will they miss? How many takeovers by the driver will they have no idea about?

How does the CDC get covid infection statistics? How does the NTSB do aviation safety studies? How does OSHA report on occupation risks? Of course the government can do this. In fact in this very area there was big news over the summer that the NHTSA had opened an investigation into Tesla autopilot after finding a collection of just 13 accidents over six years where a vehicle on AP hit an emergency vehicle! If they can pick out needles from the haystack like that, they can surely run an analysis over FSD beta.

Look, if you're arguing that we need a unified system in place for reporting accidents and correctly tagging them to autonomous vehicle software stacks, then I'd tend to agree. But that's not an argument about Tesla safety, nor even with Tesla's compliance with existing regulation.

All I'm saying is that (1) my personal experience with this system as deployed is that it's quite safe, frankly safer than driving personally, (2) there is a ton of data out there being generated by all these cars, and (3) let's go find it instead of pursuing brand flamewars like this.

> "[...] NHTSA had opened an investigation into Tesla autopilot after finding a collection of just 13 accidents over six years"

If it takes them 6 years, they're not getting enough actionable data quickly enough.

> How does the CDC get covid infection statistics? How does the NTSB do aviation safety studies? How does OSHA report on occupation risks? Of course the government can do this.

By requiring that key industries and people report the data, then sometimes with inspectors. It's not magic, and it's not going to go well if you're expecting the data to just appear for them.

> If it takes them 6 years, they're not getting enough actionable data quickly enough.

The phrase "over 6 years" means that the sample set was six years long (since AP was first deployed), not that the data is six years old! This doesn't seem like a serious argument.

My point, to repeat, was that regulators absolutely look at this data, to the extent that a cluster just 13 accidents was enough to trigger an investigation. The data is absolutely available! And to the extent that it's not, let's talk about specific reforms you'd like to see to improve the quality.

But none of that is an argument about whether or not Tesla FSD beta is currently operating safely. Again, it's quite clear that it is.

> But none of that is an argument about whether or not Tesla FSD beta is currently operating safely. Again, it's quite clear that it is.

eyeroll I'm glad you've conceded the argument on my behalf. Any other decisions you'd like to make for me?

> And to the extent that it's not, let's talk about specific reforms you'd like to see to improve the quality.

I'd start with Tesla complying with the law and providing data about how many accidents, how many disengagements, how many miles, etc. and not having to depend on what gets noticed over 6 years.

CHP has alleged cases of drunk or sleepy drivers who disabled the dead man’s switch and were somehow in a full self-driving mode. Because people who presumably aren’t mechanics or automative engineers can enable an unsupervised self-driving mode, it seems like it might fall under this law.

CHP has even dealt with one guy who has been arrested multiple times for letting his Tesla drive itself down the interstate while he (the only occupant) was sitting in the back seat. https://www.mercurynews.com/2021/05/11/backseat-driver-chp-a...

Yes perhaps to further analogy at the extreme end - it could be seen analogous to if a caffeinated-weed-vodka energy drink company claimed its caffeinated-weed-vodka enhanced your driving ability and claimed to have internal metrics suggesting it worked, marketed the heck out of their "super alert" drink and wink-wink encouraged people to test it on the road but first to go to e-sign some click through waivers on their website.

just reading this now... this person wasn't even using FSD, but was using autopilot...

Put an ankle weight on the steering wheel.

This is how it fails:

Turn the system on, keep a finger touching the steering wheel, then close your eyes. Boom.

In fact no. It actively monitors driver attention with a camera. If you close your eyes or look down at your phone for more than a few seconds, the car will shriek at you, lock you out of autopilot for the drive[1], and if you don't take manual control will bring the vehicle to a stop. Recent versions have in fact been tuned to be too sensitive and people have been complaining about false positives.

I do wish people would spend some time actually investigating how this system works before pontificating about it. It's not done. But it's far, far closer to done than you seem to want to believe.

[1] And after three such events kick you out of the beta program entirely.

Is this a recent change? There are many videos online of Tesla's "driving" with no-one in the driver seat. One tester in the bay area even had his car seized[1]

[1] https://www.nbcbayarea.com/news/local/backseat-tesla-driver-...

It's specifically enabled for those enrolled in the full self driving beta (and only while it's enabled). As far as I know it's not yet made its way to those who just use autopilot which is what was used with that remarkably dangerous backseat stupidity.

Edit: Zaroth's comment makes it sound like it's now enabled for normal autopilot use as well https://news.ycombinator.com/item?id=29598520

I am in the FSD Beta queue but I have not gotten access to it yet!

I wonder if the more aggressive monitoring and alerts turn on when you enter the queue for FSD Beta as a way to monitor your alertness before granting entry into the Beta?!

I can follow the argument on how FSD does not fall under the law because of its requirement for driver attention. I do wonder how much water that holds, but that is for another thread.

What I'm curious about is what you mean by "done". You mean an autonomous Tesla that can drive without human supervision? Is Tesla close to that in your view? On what type of roads?

I guess Tesla feels you can "actively monitor" from the back seat? Or did those people defeat some safety?

Yes they weighted the seats, slipped out from a seatbelt, and weighted the steering wheel.

Newer software released in the last couple months also adds some kind of gaze tracking using the interior camera - because the car now alerts constantly if I am looking at my phone and will disengage auto-pilot and put me in the penalty box if I keep doing it. I don’t know how that tech would react if it didn’t see a human in the drivers seat at all, but I imagine ultimately that will be another safety measure. Tesla has a lot of incentive to cut down on these types of pranks.

For those that don’t know, disengage keeps autopilot running but with an extremely annoying alarm and massive red warning on the display, and as soon as you take over the penalty box means you can’t re-engage “AutoPilot(TM)” until the next drive.

There is another way to get it to stop alarming. Don't look at your phone while you're driving.

Please don't look at your phone while actively driving.

Thank you, I know that. I’m just explaining how I know the technology works, which is interesting and relevant in this context, not looking for driving tips.

But you bring up a great point! Very interestingly, the current version of the software will alert just as much and in exactly the same way when it detects the driver looking at their phone, whether the car is at a complete stop or driving freely down the highway.

> whether the car is at a complete stop or driving freely down the highway.

I’ve seen the police on off ramps looking for people texting or whatever while they are stopped to give them tickets.

Not that it matters in the context of TFA but in certain (most?) jurisdictions it isn’t legal at all.

My comment was more condescending then I intended, I appreciate the additional context.

One guy had this foot on the wheel from the back seat. Not sure how/if they defeat the face/eye detection.

that was old tech, and it's not part of the BETA topic is about

I have to have my hands on the wheel and looking at the road near-constantly otherwise autopilot turns off, even if I look at my phone the car warns me to pay attention to the road. I am 100% a safer driver with autopilot on (on the highways - FSD on side streets isn’t there yet) because I can focus my attention on higher level details of my surroundings, such as a semi merging onto the highway 3 lanes over or a driver swerving in and out of traffic several cars away, instead of micromanaging the wheel and pedals to stay in my lane and a safe distance from the car in front of me - the autopilot handles that flawlessly.

Comments from owners are very important. It seems like the people who do not use FSD/own a Tesla have more complaints than people who use FSD/own a Tesla.

Anecdotally, I know over a dozen people who own a Tesla, use FSD daily, and have zero complaints. I use FSD daily with no issues. I get updates every other week or so that makes little improvements that I actually notice and feel while driving.

My biggest stress while using FSD is deciding what music to listen to.

People who have spent thousands of dollars to do X are biased towards thinking X is a good idea.

Last I checked, Tesla "FSD" was still a plain old SAE Level 2 system. It needs close monitoring and physical control by a human operator at all times, no exceptions, no ifs or buts. They're slowly inching towards adding SAE Level 3 in uninteresting, passively safe situations (like enabling the car to roll forward in a traffic jam). We're nowhere near a system that can actively "drive itself" in any real sense. The attempts we've seen (by Tesla and others) have been ludicrously ineffective and unsafe.

> [FSD beta] needs close monitoring and physical control by a human operator at all times, no exceptions, no ifs or buts.

It requires it, for sure, because there are still bugs and edges. It really doesn't "need" it, not for routine driving. My car drives me around now. Zero-disengagement drives are the rule now, not the exception. And where I do take control it's never really been a "safety" situation in the sense of an impending collision. It's because the car is being a jerk and not letting someone merge, or making a poor lane change, or trying to "go around" cars stopped at a light, etc...

Really, it's pretty amazing. No, it's not done. But the kind of hyperbole you're deploying there just isn't correct anymore.

Have you not had a ride yet? Seriously, call a friend and get a ride.

GM has promised an SAE Level 3 system in 2023 but so far it's still vaporware.

https://www.gmfinancial.com/en-us/financial-resources/articl...

Tesla requires active monitoring by a human operator despite the misleading name that FSD implies, and despite their boasts.

Don’t forget to read the Terms of Service!

From the linked regulatory text:

>(a) “Autonomous mode” is the status of vehicle operation where technology that is a combination of hardware and software, remote and/or on-board, performs the dynamic driving task, with or without a natural person actively supervising the autonomous technology's performance of the dynamic driving task. An autonomous vehicle is operating or driving in autonomous mode when it is operated or driven with the autonomous technology engaged

The "active monitoring" part doesn't seem to exempt them from this law. What I'm not sure is - wouldn't this also apply to high cruise control / following?

Yeah, I don’t think they mean to ban cruise control, so it is confusing.

This law seems to be vague in critical ways:

>"performs the dynamic driving task, with or without a natural person actively supervising the autonomous technology's performance of the dynamic driving task"

The terms 'dynamic driving task' and 'supervivising' are critical here, and don't seem to be clearly defined. Supervising could mean an absence of active physical control, maybe meaning a 'supervisor' without a hand on the steering wheel, but with access to a kill-switch. 'Dynamic driving task' also seems a bit vague; does this mean turning, changing speeds, or something else?

This could be interpreted for or against Tesla; maybe that's what they wanted.

It specifically defines 'dynamic driving task' in the definitions section:

> (g) “Dynamic driving task” means all of the real-time functions required to operate a vehicle in on-road traffic, excluding selection of final and intermediate destinations, and including without limitation: object and event detection, recognition, and classification; object and event response; maneuver planning; steering, turning, lane keeping, and lane changing, including providing the appropriate signal for the lane change or turn maneuver; and acceleration and deceleration.

And don't forget to read the comment you're replying to!

Then Tesla should disable it in California.

If I understand his argument correctly, Tesla would be in compliance if they allowed anyone who buys FSD to use it immediately, just like Ford allows anyone to use Blue Cruise on Mach-E.

One of the essential parts of the argument is that Tesla falls under current law because they are selective about who they allow to use FSD.

So Tesla could trivially be in compliance if they ditch the careful rollout. Give the software to anyone who pays for (instead of, as is the case currently, slowly expanding the number of people).

And maybe change the name to Yellow Cruise instead of FSD and stop telling people it's beta software.

And voila, compliant with CA DMV legislation.

I hope that this person understands his own argument in which case it's hard to believe that he is concerned about safety.

If they could roll it out to everyone they would do so in an instant. They cannot though because the system is not ready and needs a lot of testing, which is the point the author is making.

It's 100% not ready. There's so many videos people are posting about it nearly hitting pedestrians, going the wrong way down a 1 way, nearly going straight into a curb, etc - all of these only stopped by the driver quickly grabbing the wheel and taking control.

It also doesn't seem to help that enabling FSD on Tesla makes it use the visual camera only, which seems counterproductive. The ultrasonic sensors seem to be critically important because visuals can be obstructed easier.

Please please be careful when reasoning from cherry picked data. You can go to http://reddit.com/r/idiotsincars and see much worse, every day.

All I can say to comments like this is "call a friend and get a ride". There are ~23k cars on US roads with this system now. They're not hard to find. And it's really pretty mind blowing. For sure the most fun I've had behind the wheel in 25 years of licensed driving.

> Please please be careful when reasoning from cherry picked data.

The very fact that you are able to cherry-pick failure cases from a mere fleet of 23k cars means the system is by no means ready.

In fact, you/Tesla/everyone involved should encourage people to "try your best to cherry-pick failure cases". What you are advocating is the opposite of being responsible (let's try to ignore the failure cases and focus on when it works!)

>For sure the most fun I've had behind the wheel in 25 years of licensed driving.

I'm sorry but the amount of "fun" you have behind the wheel is not remotely close to being a success criteria for a system named "full self driving".

> What you are advocating is the opposite of being responsible

To be clear: what I was actually saying was not to ignore videos on youtube, but to be responsible about interpreting them. When data sets get large, you need to stop reasoning on the basis of anecdotes (because there will always be single data points to support any perspective) and start making your decisions on the basis of data.

Basically: is FSD beta as deployed any more dangerous than human drivers? I think it's clear to everyone that the answer is "no", for the simple reason that the roads are filled with bad drivers already. In fact the one actually documented accident involving a beta car of which I'm aware[1] was one where the car got T-boned by a red light runner.

Really, I'm being sincere: stop being angry on the internet about this and get yourself a ride. You'll see things differently, I all but guarantee it.

[1] Sort of. It was on twitter and I've lost the reference and might be misremembering.

> The ultrasonic sensors seem to be critically important because visuals can be obstructed easier.

The car still uses ultrasonics when parking but AFAIK it has never used them for driving around. Perhaps you meant radar?

Yeah sorry, got the sensors mixed up. But going visual-only seems to be quite the handicap. I think other makers are going to have an edge over tesla if they keep radar.

It was publicly known that Musk was personally involved with requiring all sensors to be completely hidden from view, even to the point of inhibiting the car's view of its surroundings. This was back when other cars had the large LIDAR installations on top of the roof, which was probably better, but obviously didn't give the regular-car look Tesla wanted.

> There's so many videos people are posting about it nearly hitting pedestrians

Can you post a link to a video of FSD nearly hitting a pedestrian? Because searching on YouTube, I couldn't find a single one.

https://www.loopinsight.com/2021/12/13/drive-around-san-jose...

There was a video posted on reddit a couple of days ago. I think it was in San Jose, a number of times it tried to go into the tram lane and once or twice it almost hit pedestrians crossing a road when turning. Maybe that gives you enough information to find the link.

Yeah, regardless of this supposed legal incentive, I don’t see any reason why Tesla wouldn’t roll it out to everyone if they thought it was ready for that.

>I hope that this person understands his own argument in which case it's hard to believe that he is concerned about safety.

I think he understands his own argument. I don't think you don't understand the implications of why they haven't rolled it out.

Isn't Blue Cruise only available on select mapped highways, and only described as a SAE J3016 level-2 system? That seems to be a crucial difference.

```The name ["Full Self Driving"] unequivocally communicates Tesla's goal for development, and the company's "beta" qualifier communicates the stage of that development. Tesla intends for its "full self-driving" to become, well, full self-driving, and its limited beta release is a key step in that process.```

Seems like your analysis skipped that part.

There is a huge difference between goal and actual state. You can't enable FSD Beta without clicking several checkboxes in clear language saying you have to pay attention every second on the road, you're responsible for single action of a car on the road. And it provides more restriction on a driver compares to regular driving, such as distraction camera monitor, phone monitor, hands off the wheel etc..

It's so boring to see how trolls have no numbers, no real issues but just trying attack the marketing name/aspirational goal.

Ooooh. Checkboxes. That will definitely prevent abuse.

Even if the Tesla owner is responsible about it, other people could drive the car and they wouldn’t have to agree to anything. They can go nuts.

I’ve seen plenty of videos online too. I have no interest in driving next to a car being driven by a crazy AI that seems to have the driving skill of a 13-year-old playing GTA and not know the rules.

I don’t have a Tesla. I was forcibly opted into this program. There is nothing to prevent other people from doing stupid and ridiculous things with this system and seriously injuring or killing me.

Heck, they can do the right thing. I’ve seen videos were trying to take over control doesn’t seem to work. Where it tries to drive directly in the traffic in the wrong lane.

This is insanely dangerous and I want no part of it but I don’t have any mechanism to stop it but to hope regulators do something. Tesla put MY life on the line.

They use eye tracking camera, to prevent you from lookin not on the road, sensors of the wheel, and also checking you're not on the phone. They doing way more compares to any car maker to keep you safe as non owner of the tesla. I guess been so emotional you better stay in your basement until Level 4-5 will become a reality.

Tesla added the camera eye tracking very late in the game, just like they didn't add driver hands-off wheel detection and enforcement to autopilot until it had been available quite some time.

I think this is one of the issues with Tesla FSD - the company basically has had to be forced to take safety seriously repeatedly and has not done so from the outset.

Similarly, you and I are both forced to share the roads with folks who are impaired in various ways, whether it's declining reflexes, vision, cognitive function - or just straight up drunk driving. As somebody who comes from a country with very strict anti drunk-driving rules (checkpoints randomly on commonly used roads frequently enough to deter it), it blows my mind that I have to share the roads in the US with drivers who think having "just 4 beers" before their 45 minute drive home is fine.

Since the US hasn't shown many signs for cracking down on drunk driving, and adding better infrastructure for folks who can no longer drive for various reasons is equally a non-starter at the moment - I'm all in favor of any system that speeds up the arrival of genuinely fully autonomous driving. Tesla hasn't killed or injured anyone using this "FSD beta" program to date, despite a large number of miles driven, so their existing precautions appear to be working.

Even in the interim, before we reach fully autonomous driving, "better than the average human driver" is still a huge performance milestone worth aiming for before we start obsessing over "perfect" (Not saying FSD Beta is there yet, but Autopilot on Highways might be).

I don't fully understand your argument - are you saying FSD should be tested publicly, because we allow other impaired drivers as well? Wouldn't the solution be to strictly enforce the ban of impaired driving, on the one hand, and test FSD privately using trained supervisors, on the other hand?

If effectively enforcing the ban on impaired driving was a viable option that was truly on the table then yes, I would 100% be in favor of requiring this testing to happen privately. Breathalyzer for engine start, eye tracking to prevent phone usage (which Tesla has btw), more frequent vision/driving tests, etc would improve road safety significantly. But that's just never going to happen, it's a completely unrealistic scenario that requires a huge culture shift, and would be incredibly unpopular policy. Not to mention the public transport infrastructure that needs to be built to support all of these folks who suddenly are no longer allowed to drive but live in areas where driving is a necessity of everyday life. The demographics who'd be affected hardest by these kinds of changes are also vocal voters with high turnout.

Given the state of the public roads now, full of dangerously impaired driver's as it is, we've clearly decided as a society that it's worth the risk to allow this behavior to continue. So in that light, I think it's worth allowing these public tests of autonomous systems under some strict set of rules. I agree these rules shouldn't be exclusively self-enforced by Tesla and co., and there should be some more stringent open transparent reporting requirements around how many systems are active, miles driven, incident reports etc. But with the right controls and oversight in place I think it's a worthwhile endeavor to allow the testing to continue in the spirit of accelerating autonomous driving becoming a reality. It's the impaired drivers that _make_ it more urgent for this system to exist in the first place.

> Not saying FSD Beta is there yet, but Autopilot on Highways might be

Anecdotally I've found the new FSD Beta a huge leap forward when used like "Autopilot" (basically no destination, drive in a straight line for lane keeping). It understands the roads 500% better than the older builds in terms of lane layout, little or no lines, etc.

> Since the US hasn't shown many signs for cracking down on drunk driving…

Kind of curious where you live because the police seem to only look for drunk drivers these days, for the money.

In Phoenix you have to make the police do their job (like take their bumper off) if they don’t think you’ve been drinking because they know they just have to wait a bit to write up a big-dollar ticket.

I assume you refer to the invite-only beta with FSD that works on city streets, but Tesla's naming is (intentionally?) very confusing.

It would go a long way if Tesla simply renamed the invite-only program to FSD Alpha, and continued referring to "enhanced Autopilot" that's available to everyone for an additional fee as FSD Beta.

Can anyone explain me what is fundamentally wrong with Tesla approach emotions aside. Tesla rolled out FSD Beta(not Autopilot) to a very small group of people. After been in public for almost a year, it has 0 death or injuries. It is not ideal. Tesla clearly understand that and communicate that to every user of FSD Beta. They limit amount of people using it. Also they limit the way you use it introducing lots of points of control so that you pay attention at the road.

EDIT: They will kick you out of the beta for a slight miss-doing such as phone or distraction. You will not get into this beta if your driving skills are not perfect. Even if you press a break too hard all the time, it's a passive sign to them your driving is not perfect.

They invest billions of dollars in infrastructure so that application of this feature can grow. Tesla use testing data to eventually enable it in public. That will occur only once they are confident in their security level. They attract some peoples money who admire their goals to save peoples life.

What's wrong with that? Why people getting mad every time they hear aspirational name Full Self driving. What is the logic behind it?

For all the claims I've read about Tesla supposedly misleading people, I have yet to see any evidence that someone with access to the FSD beta was ever misled by Tesla about it.

> Can anyone explain me what is fundamentally wrong with Tesla approach emotions aside.

The fundamental problem is testing on public streets technology that can very easily kill innocent bystanders if it doesn't work correctly.

I don't have a Tesla, but I'm still subject to the risk just by driving, biking and walking on public streets that Tesla thinks is their playground to test buggy new code.

Testing word processing software on your users via beta is really no problem.

Testing autodrive software on your users, which might kill them, is wrong.

Add in that the messaging that the car will drive itself (ie that it's something other than level 2 automation) and the poor quality of the software and it's not hard to see that Tesla's behaviour is self-serving and repugnant.

Does it not have 0 death or injuries because they kick you out of the beta for not driving perfectly?

yep, that proves their model actually *WORKS*

>Tesla's "FSD" has the "capability to drive a vehicle without the active physical control or monitoring by a human operator,"

Tesla doesn't not have this capability, after few seconds not having your hands on the steering wheel , or not watching on the road , or not having a driver in a sight of camera it will ban you fron FSD Beta. It also monitors if you're using your Phone and warn you for that and after ban. This feature alone makes FSD safer compares to regular human :)

Precisely this. Whether the authors of the legislation intended it or not, this mandatory requirement for continual human input excludes a system from their oversight.

The FSD Beta, as it is supplied to end users, is not capable of performing the dynamic driving task on a sustained basis without the constant control or active monitoring of a natural person.

This interpretation of the law renders it totally useless, because tacking on the requirement for a human presence is trivially easy.

As the author points out, the intent of the law is to enforce certain restrictions on any system that is intended to eventually become fully automated. And clearly Tesla is testing a system, and the system is supposed to be fully automated one day.

I don’t agree that it makes the law useless. An autonomous vehicle is of extremely limited incremental value if it requires persistent driver engagement at every moment.

Technology like we’re seeing in FSD Beta is going to be the next big road safety revolution. Even if it never becomes an autonomous vehicle, repackaged as a passive safety suite, the technology is already good enough to be a collision avoidance system orders of magnitude better than any other on the market.

We should just apply these same rules to driving, full stop. So many people are killed by inattentive drivers every year; even a short timeout from driving (15 min or so) in response to inattention would probably reduce these killings drastically.

Why is it called Full Self Driving?

Unless I'm reading this wrong, this is just a person's opinion, not an opinion or ruling from a government agency. Interesting, but without any legal authority.

Yes, these things will only shake out when someone takes Tesla to court.

However the opinion of a professor of law, who is also an engineer, is not something to so easily dismiss. IMO, it means that Tesla is likely to find themselves in court.

Right - he helped California develop their standards and was one of the coauthors for the "Level 1/2/3/4/5" SAE classification for self-driving. His opinion obviously isn't backed with the force of law, but this analysis is likely enough to get regulators to send Tesla a letter or two.

So Tesla geofences FSD beta to outside California? Like moving their HQ to Austin? California should absolutely act if the law covers FSD, but it’s entirely realistic for Tesla to jurisdiction shop (just as credit card interest rates are mostly set by the laws of South Dakota).

California is one of, if not the largest EV market in the US right now. Consumers by and large don't care where a company has their HQ, but they will be upset if they don't have access to features. And this impacts people that aren't in CA that have FSD but road trip to/through CA. I can't imagine this is an ideal situation for Tesla to be in, and they'd prefer to avoid this as much as possible. There isn't a really good solution here for them.

I don't see how they can jurisdiction shop in any similar way to a credit card company in this instance, these are markedly different scenarios and services provided.

Not only that, but California EV transferable credits are a major part of Tesla's income stream. IIRC, if Tesla stops selling in California, their profits go down by more than 50% (assuming that the other states can absorb the extra Teslas, which at this point is a reasonable assumption.)

They’ll still sell FSD, they just won’t offer the beta functionality in (or if you traverse) California due to “regulatory constraints.” They’ll continue to run FSD beta testing in 49 other states.

In similar fashion, Autopilot has tighter operational constraints in parts of Europe versus the US.

https://www.teslarati.com/tesla-autopilot-europe-restriction...

Yeah, they will still sell it but it wont sell to people living in CA which is one of the larger US markets. that's a non-trivial market to close your doors to for something like FSD. The margins on FSD have to be some of the best on the whole car.

The EU regulations and standards are likely just as annoying for them, but they don't sell the same volume across the EU yet, and the regulations are a bit more consistent across the whole region. So getting a Tesla in one EU country will work the same in another. Which is just easier for a consumer.

I am not saying this would end Tesla, but I can't imagine a world where they want it to come to that. I think they would look at either properly complying or some other workaround before just drawing a fence around FSD in CA.

I think you're a bit confused about what's going on. Autopilot, FSD and FSD Beta are all different things.

Autopilot is advanced cruise control, built-in and sold everywhere with each car.

FSD is a software package you can add. It can do a few things, change lanes on highways, chime when the green light comes on and you haven't moved for a few seconds, etc. You can buy FSD anywhere, but not all the features are available everywhere. Overall, it doesn't do much right now. FSD Beta is a testing program you apply to and can join if you have a good driving record and are selected. Only in the US. There's at most low-digit thousands of people who have access to this, fewer that actually use it.

Article is about the last bit only. Not being able to test FSD in California only is probably a pain, but won't impact much. Allowing people to test the beta only started a few months ago.

The European Union already shut down Tesla's "full self driving" capability

No, they didn't.

Tesla never enabled FSD in Europe. Currently it's only enabled in US. FSD couldn't possibly have been shut down if it was never enabled there.

As far as Autopilot goes, EU limits how much software can steer the car that makes some of the Autopilot functionality not possible (e.g. to take a sharp turn you have, you know, turn the wheels).

For that reason some Autopilot functionality is limited in EU.

This is not anti-Tesla rule but one that affects every car maker and they are in the process of changing that rule.

Yeah I don't know they have a good solution - but based on his analysis, more honest marketing (e.g. stop claiming things like "Drivers are only there to meet a legal requirement, the car will drive itself!") might be enough to stay on the right side of the line.

I'm pretty sure Levels 1-5 classification far predate self-driving classifications.

It’s easy for a legal argument to appear convincing in isolation. If you haven’t heard from the other side, it’s probably appropriate to maintain some reservations, as I dare say that Tesla’s lawyers would see things differently.

I think a critical point to make is that this is not actually about Tesla/Waymo. Its about creating guidelines that are generally applicable to all companies pursuing self driving car technologies.

So it applies to Tesla but perhaps more importantly companies that are trying to copy-cat style mimic Tesla's move fast approach to SDC for tech advantage reasons/marketing PR/save money and perhaps are cuting even more corners than Tesla.

Some examples I hear about are Nuro and maybe even Cruise, Aurora.

And then imagine a host of other companies as well - Faraday Future, Comma.ai, Pony.ai, Xpeng, etc. Its unclear at this point that these companies shouldn't have some oversight/held to some standard operating procedures.

https://techcrunch.com/2021/12/14/pony-ai-suspension-driverl...

The author of the article is incorrect in their understanding of what FSD is. They imply several times that Tesla is running the beta to test an actual driverless system, which is illegal. Despite the name and any marketing claims, the system CANNOT run without active human monitoring and intervention. This isn't just for the testing phase - there will need to be human hands on the wheel at all times at full release as well.

This alone exempts it from the California law, which is written for driverless testing (the kind Waymo and Cruise do).

Are you saying FSD will never be full self driving? Where is indication/evidence that that's Tesla's intention?

I am saying that the current iteration of FSD that they are testing on the streets is not self driving as the California law defines it since there need to be hands on the wheel at all times. If/when they ever remove the requirement we can revisit the legality of their operation, but that is not of concern today.

But then you are back to square one - the point of contention: Without the checks, it's a self-driving system, but it can't do so safely (yet), hence the checks. And as you said, there seems to be an intention to be full self-driving, without checks, at some point in the future.

>... as Tesla's engineering and marketing have become more aggressive.

I get aggressive marketing. What is aggressive engineering?

Willing to take more risks with what they are deploying?

BS from author , Tesla don't market FSD Beta at all. They market autopilot, it's a totally different system

On our team we call them "code crimes".

https://www.smbc-comics.com/comic/fsd

Telsa is one of the most evil, malicious, companies, in that it knows that what it has is well short of self-driving, and yet markets it as such. This leads to deaths. Their marketers have blood on their hands.

Electric cars were a joke until Tesla showed up

My Nissan LEAFs were no joke. Sure they didn't have a fancy 17" LCD display, but they were good cars in their own right. I leased 2 of them.

leaf was a joke as a product in hot and cold weather due to their non working cooling/heating battery system. Even though it's perfectly working in some cases.

OK but to be accurate, the Tesla Roadster came out before the Nissan Leaf.

I think Tesla has helped span the gap between EV hobbyists and the general car consumers, but not all by its self.

Have you personally used Tesla's products in California?

What a strange question. It reads like one of those weirdly-specific tshirts.

And yes. I have personally used Tesla's products in California, in May, when it wasn't raining, after a small lunch I made at home.

I'm asking because I'm surprised you wrote that having been a user. Seems I have a very different experience as a customer than the people who have a problem with them.

I guess what's happening here, is that legacy automakers not able to make EV's with decent profit ( that's real numbers ), hence their production is limited.

Traditional media creating so much buzz to slow down and make tesla look bad. I can see every bad news about tesla re-printed everywhere. 1 car got on fire boom... every car outlet re prints. BMW recalled 1M+ cars because of fire risk and some of them still catching on fire on the road while driving on a highway, no one even heard of it.

Regarding the article. Author probably never tried to drive FSD. If driver doesn't look on a road for few seconds few time it will ban you from this feature. They use camera, sensors of steering wheel etc... It is very aggressively RESTRICTED , locked LEVEL 2 system. It's only available for few thousands of people. Regular autopliot works only on highways and very limited.

FSD name is clearly aspirational. Once they will have a confidence they will switch from level 2 to 3 etc.. By removing/easing restrictions. Only AFTER THAT it will become eligible to a different regulations.

The only debate here could be how aggressive you are about this level 2 restrictions while building LEVEL 5 system.

I am part of this Beta program. My personal feeling after using this system: you pay attention more compares to regular driving. It tuned this way it will rather do a false positive distraction decision.

And yeah, current version is so far away from LEVEL 4 or 5. And it's not usable for every day. The only use of it try, test, provide feedback, turn off , and use the car in a traditional way. There is 0 amount of people who is using this feature not for testings purposes.

Not everything is a conspiracy against Tesla by the media. A lot of folks have legitimate concerns about the way Tesla markets, tests and operates it’s self-driving systems and educates it’s customers about the limitations.

This is a video from 5 years ago describing Tesla’s self driving features.

https://vimeo.com/188105076

I personally feel like Musk is writing checks his engineers can’t cash.

A lot of the criticism of Tesla is invited by a stock price which a lot of folks think are propped up by mistruths and exaggerations about the future of company and the feasibility of it’s vision.

Also, quite frankly I find the term “legacy automaker” infuriating. Toyota builds 9 million cars a year or 18x what Tesla does and Toyota has a legendary reputation for quality and reliability.

Tela's current production rate is at ~1m a year atm. And they doubled amount of their factories in 2022. So not 18x but 9x now and 4.5x a year or two after. They are in the same league nowadays. And they are doing that with 3 factories in 2 countries now. Compare that with Toyota :)

Also Tesla outsold Toyota/Lexus in almost every country they operate in segments/pricing they compete by absolute volume.

Toyota has also a reputation like any legacy car maker a reputation for selling you some crap like GPS for 10x the price. At least Tesla's overpriced autopilot approach works towards safety( the end goal level 5 system). And if you don't believe in it , not buying it will not make you getting from point a to point B more complicated.

Tesla also sell every car "full package" in legacy car makers terms. Every camera, every seat ( except color), every screen detail, speakers , gps is the same in every model from the cheapest to the most expensive. Compare that to legacy makers trying to get 10x of every simple feature like seats heating. You need an App? Pay for it. You need firmaware upgrade go to service and pay for it..

Speaking of safety tesla's cars has always excellent security tests. Toyota don't. There are many things toyota lags behind. Like multi-media, autonomous driving, connectivity, software over the air, battery tech, electric motor design, manufacturing. So yeah, they are legacy atm. Legacy can have a good reputation, even perfect nothing wrong with it.

They are legacy because as of 2021 they bet on legacy tech, their way to sell cars is legacy, their approach to packaging is legacy, their approach to safety is legacy... their way to produce cars is legacy.

Toyota has quite good roadmap for 2-3 years, maybe they will get into modern world eventually.

> They are legacy because as of 2021 they bet on legacy tech, their way to sell cars is legacy, their approach to packaging is legacy, their approach to safety is legacy... their way to produce cars is legacy.

So why does Toyota sell more cars than Tesla?

> Traditional media creating so much buzz to slow down and make tesla look bad.

That's a complicated conspiracy theory. The truth is more simple: Tesla is making itself look bad.

5 years ago Tesla said that all of its cars being produced at that time had "the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver":

https://www.tesla.com/blog/all-tesla-cars-being-produced-now...

That was a lie.

> FSD name is clearly aspirational.

Super. So when will Tesla start refunding everyone who paid for "full self driving"? Until the "aspiration" becomes an actual delivered product giving refunds is the only honest thing to do.

Don't do beta testing with machines that can kill people in public. End of story.

Any person can get distracted and kill people in public. Most of the cars doesn't have distraction mechanisms like FSD Beta. e.g. if you're on the phone, FSD will not work. Lets ban people from roads according to your logic. FSD Beta killed 0 people since 1 year of program, humanity killed and made millions people injured due to human mistakes. Any Level 4-5 system deployed on a large scale with save millions of people, prevent even more people from been disabled in the future.

This is some of the most twisted logic I've seen, kudos to you.

According to you, as long as a piece of software isn't killing millions of people, it's totally A-ok because human are worse?

This requirement is impossible to fulfil because there are no cars that can't kill people on public roads. Regardless of means of driving.

Thousands of people die every year because of human errors. Do you think this should continue forever? If not how do suggest we create software without testing it in real world conditions with a safety driver who can monitor and take over at any time?

Do you know how many people have died because they had a medical emergency while driving? Do you think that should continue forever?

If you actually want to save lives, you might start at the top. Driving is hazardous for operators and for bystanders, and kills 40,000 people in the US alone every year. Build cities such that the majority of people don't need or want to drive, and viola, you're saving lives.

And the tens or hundreds of billions of dollars worth of human effort and engineering put towards making cars that drive themselves can go to something else.

a very good joke, what's your estimate to rebuild every city and road in the world lol. Billions of dollars it's a very tiny small fraction of this money.

It's a matter of priority. Many people have this concept of Amsterdam and Utrecht as having forever been a cycling paradise, but the reality is that they were choked with cars and smog until the 70s. The political will existed to prioritize non-automotive modalities, and a few decades later you have what we see today.

Most deaths from driving occur outside of cities.

could you clarify what you mean by 'outside of cities?'

It would probably be more accurate to say ‘outside of city streets.’ The threat of death tends to scale with vehicle speed.

I think testing these systems on public roads is 100% reasonable if we all have open access to data. I’m not talking about raw driving data, but aggregate metrics about how well the system is operating and limitations are clearly communicated.

Selling a customer a car with a “full self driving” feature while burying the actual functionality in a legal agreement displayed on screen in the vehicle.

This [1] is a screenshot I just took from the "build your vehicle" section of Tesla's website. It took me one click, and scrolling down one page, to get to this point.

How is that "burying the actual functionality"? It's right there!

[1] https://imgur.com/9IUJW1B

That misrepresents how Tesla sells FSD. The actual current functionality is shown in clear bullet points at the time of sale. It is not buried in a legal agreement.

FSD? Tesla sells full self-driving? Does Tesla have a product that does actual full self-driving?

Given that “full self driving“ isn’t a formally defined term, your question is not one with a concrete answer. It’s just a name. It’s clear to anyone buying the product what it does. Nobody need to explain that Starburst candies don’t literally cause stars to burst.

(For what it’s worth, while I think criticism of the “autopilot” name is patently absurd, I do agree that FSD isn't a very good name and it would have been better if it were something else.)

> Given that “full self driving“ isn’t a formally defined term, your question is not one with a concrete answer.

Don't make me laugh. Everyone knows what it's supposed to mean and Tesla hasn't delivered it. It's embarrassing to witness you even trying that on.

Reply on news.ycombinator.com