Without knowing the "full" details of this crash, it is outrageous that Tesla calls it "Full Self-Driving (Supervised)" on their website. This is like naming a product "Healthy Dog Food (Rat poison)". If it requires supervision, it is not "full". Period.
I thought previous lawsuits would've forced Tesla to call it "Advanced Driver Assist" or something.
I agree with the sentiment and don't like Elon at all. But I think there is a case to say the Full refers to features, so not only lane change / parking, but actual driving fully. If it said Perfect FSD or even High Quality FSD, that's definitely worse, but we have to acknowledge FSD can fully drive in many cases, not enough for me to be comfortable enough maybe, but comparing to rat poison seems like a stretch.
As long as Tesla doesn't insure their cars (whenever in self-driving mode) completely for the price of their FSD, it's not FSD, just scammy marketing by a chronic charlatan.
Holy shit, it looks like it was going at full highway speeds (at least) in a residential neighborhood. Doesn't look like it attempts to stop or steer away at any point. This seems like the outcome one would expect from a passed-out driver and a stuck accelerator. I'm honestly amazed that the driver survived.
There are very few (if any?) known deaths caused by FSD accidents.
Note: it’s important to distinguish between Autopilot and FSD here. “Autopilot” is Tesla’s old assisted driving stack that comes free in most vehicles and has had no significant updates in years. “FSD” is an entirely different software stack that only works with newer vehicles and that Tesla charges $$ for. It’s much more advanced and IMO a lot safer.
> There are very few (if any?) known deaths caused by FSD accidents
It’s tough to say given many data sources are aggregated. For what it’s worth, my parents’ car is a Tesla with FSD and I’ve stopped it from, off the top of my head, racing into a red-lit intersection, running over a small dog and running into a closing garage door.
I still use it. It mostly works. But I’m vigilantly monitoring it in a way that isn’t supported by Tesla’s marketing (which frequently shows drivers engaging it hands off).
The garage door I can certainly believe. Its AI brain just won’t be trained to look for that sort of thing.
I’m surprised about the red lights and animals, however: ours seems very cautious around any kind of live animal on the road, even braking and manoeuvring to avoid birds on the road. It’s not so good at avoiding the corpses of already dead ones, however (bump!).
I get the impression that it kind of comprehends the world as a horizontal plane. It’s focused on objects on the surface and isn’t looking for hazards coming from above. Could that be improved? Sure! Is it a priority for them? Maybe not…
At minimum it needs to pay attention to railroad crossing bars and bridges.
They show a 2D representation to the driver because that’s good enough for drivers, but I wouldn’t assume that represents how the system operates internally at every stage. Even navigating requires the concept of bridges crossing roads without intersecting them.
AFAIK, there's no issue with bridges that cross roads, however. It will just ignore a road that's above or below the one you're travelling on that doesn't intersect with yours. Just like a human driver would do.
> It is indeed known to have issues at railway crossings:
At scale you can “have issues” and still work 99% of the time. If it’s completely incapable of handling railroad crossings that’s relatively easy to verify.
> AFAIK, there's no issue with bridges that cross roads, however. It will just ignore a road that's above or below the one you're travelling on that doesn't intersect with yours. Just like a human driver would do.
As a user of the system that’s worked it all out it’s not an issue for you, but getting that behavior from raw sensor data is non trivial.
Both only happened once. But they were shocking when they did. (To be fair, my Subaru tried to push me into oncoming traffic because it was avoiding a “collision” from the guy in the other lane turning weirdly. Turned at collision-avoidance feature off.)
> “Autopilot” is Tesla’s old assisted driving stack that comes free in most vehicles and has had no significant updates in years. “FSD” is an entirely different software stack that only works with newer vehicles and that Tesla charges $$ for. It’s much more advanced and IMO a lot safer.
I'm not sure the point you're making here - that just sounds like "Tesla couldn't care less about updating their software and it's still "not good". People it hits are still dead. "Oh well, it's not like Tesla had updated the software, so you can't blame them".
Oh, I absolutely agree. Ideally Telsa would dump the old Autopilot stack and put all cars (that have the hardware to support it) on a free version of FSD, nerfed so that it's roughly feature-equivalent to Autopilot.
That way everyone would get the safety advantages of FSD, including the really important stuff like better driver attention monitoring. Unfortunately, Tesla keeps Autopilot bad because it forces more people to pay for FSD, and in this case they seem to care more about profits than safety.
> Unfortunately, Tesla keeps Autopilot bad because it forces more people to pay for FSD, and in this case they seem to care more about profits than safety.
Not saying you're wrong, but I think "it forces people to pay for FSD" is a poor decision if true.
More likely it creates market confusion (as we see in this thread) about exactly which of the two products that Tesla advertises as things which drive your car for you is the more dangerous one and which is "the good one" (though from their taxis, I think the scare quotes are still justified even then).
I'm also not even sure if they do care about profits (or safety), given Musk's made about as much from selling Tesla shares as the company has made in lifetime profit, before getting shareholder approval for a deal to give him more shares. (Conditionally give, yes, but the nature of it all suggests share price is more important than profit or safety, and if you don't care about such things that makes it much easier for them to sell a million robots or whatever).
That the reporting is so tortured sort of shows this works. It’s irrelevant in a broad or legal context. But for the narrow audience of his fans, it carries merit.
does it makes sense to compare the two given that waymos are driven in a limited set of circumstances and most of the times below speeds that could kill you on a crash?
If Tesla was bold enough to intentionally auto disengage seconds before an accident (to avoid liability) then can they be trusted to maintain telemetry throughout dicey circumstances?
I also don't trust the motives of a company that names something "Full Self Driving" knowing it's not fully self driving. Never mind their shenanigans around avoiding or disregarding regulations and reporting requirements.
That's how it is now, and (like most things safety-related) Tesla had to be dragged kicking and screaming into it. Just like the initial attention/distracted monitor only required your hand on the steering wheel once every fifteen minutes.
They also considered AEB-activation to mean that meant that FSD wasn't active, even if AEB only kicked in because of FSD's decisions.
Auto disengaging seconds before an acciden't isn't a Tesla-specific thing, it's part of the standard for any driving automation SAE level 3 or lower (https://www.sae.org/news/blog/sae-levels-driving-automation-...). The vehicle will return driving over to the human any time it's unable to determine what to do.
I don’t know about laws, but Teslas automatically record and save everything in the event of a collision. My brother was hit by another car while driving a Tesla a few years ago and it was very easy to retrieve video from the Tesla’s cameras and show who was at fault.
The reddit post from the grandchild says "autopilot", which is really just lane keeping and adaptive cruise control.
I believe autopilot would totally run into a house. It doesn't respect stop signs or red lights. If the house is at a T intersection the autopilot might try to drive right through it. I agree about FSD though
Adaptive cruise control certainly shouldn't run into a house, for the same reason it should detect a stationary (or slow) object in front of you and gradually slow to a stop before it.
Certainly Tesla will. And that will inform how their PR team responds to this collision.
Maybe some plaintiffs will if they can manage to subpoena the data from Tesla in some hypothetical future court case.
The reason I will never buy a Tesla is because it is one of the most advanced surveillance systems against the driver, but there is no one empowered to inspect the car / company (comprehensively, not superficially).
> I seriously doubt FSD would drive into a house.
Strawman. You seem to be insinuating that FSD intended to aim for a house. Usually the chain of events would start with something like “FSD was engaged on the road and in the intended lane” then “FSD lost track of the boundary of the lane” or maybe “FSD identified an obstruction in the lane so it maneuvered out of lane”.
Teslas do glithcy speed limits readings often, from GPS drift (ie. it thinks the car is on the highway) or may be deceived by some sign with numbers on the street.
When that happens and depending on Autopilot [1] settings, the car will shoot out to reach the new limit. If the driver is distracted enough not to notice or react in time, the car could crash. GPS inaccuracy does sound plausible given Teslas slow down if an abrupt change of direction is near (ie roundabout) per GPS and map data.
The house sits at a corner, not a culthesac.
1. It's not clear the driver allegedly said "autopilot" or "FSD"
Previously: A Florida jury found that flaws in Tesla's self-driving software were partly to blame for a crash that killed a 22-year-old woman in 2019 and severely injured her boyfriend.
The jury verdict requires Tesla to pay $243 million in punitive and compensatory damages to the parents of the woman and to her boyfriend.
1) I’ve put enough kms on FSD - it’s taken me across Australia a few times, probably 10,000kms in total - to know that it isn’t going to drive into a house.
2) Even if FSD is enabled, there’s loads of things you can do to create an accident like press the brakes or accelerator pedals, which doesn’t necessarily disengage FSD right away, so let’s just wait for the telemetry to get released.
3) Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
> 1) I’ve put enough kms on FSD to know that it isn’t going to drive into a house.
Is it possible FSD on this vehicle was a different version? Can't FSD change from one drive to the next, based on software updates or even external conditions?
Perhaps you drove in a different region with differing conditions?
> 3) Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
Why is it called "Full Self Driving" if the person behind the wheel must control (or even just monitor) the speed?
That’s exactly the thing. We’re on FSD 13 and America is on FSD 14, we’re always about a year behind. Yours is significantly better than ours (being RHD market), but ours still wouldn’t drive through a house.
You’re extremely confident about this just based on your experience of it not happening to you. It’s good that it hasn’t driven you into a house but that doesn’t mean it can’t fail in a way that does drive someone else into a house.
Well, I’ve seen it very confidently and correctly navigate rural unmarked roads, and drive around rocks and logs across the road and things like this. I find it hard to believe it would drive through a huge stationary object like a house, which I imagine is definitely in the training data.
I guess I’m not saying it’s impossible, but I think at this stage it’s improbable.
> Perhaps you drove in a different region with differing conditions?
Agreed.
He's driven it "across Australia". Which means long straight highways for hundreds and hundreds of miles at a time (then recharge, then do long straight highway for hundreds and hundreds of miles), and often fairly lightly trafficked:
Like... absolutely shocking that FSD performs decently on well maintained, largely straight highway in generally clear desert conditions. Surely this is indicative of its behavior everywhere else! /s
I would agree, and know well, though I lived in Melbourne for two decades. But still, "driving across Australia" doesn't involve a lot of time spent in the Sydney (or Melbourne) CBDs.
> Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
There are decades worth of man-machines UX research to prove this: the more you lean on automated systems to perform a manual task, like specific vehicle operations, the more your reaction time and relefexes for that specific motor skill will suffer.
Non-level 4 driver assistance tech should only be used for helping prevent accidents and not pretend to be actual full self driving
Why do news outlets so often use the phrase "high rate of speed"? Speed isn't a discrete event, it can't really have a rate, unless it is a rate of change, in which case they would be referring to acceleration.
There's absolutely no way this was on autopilot. A) autopilot would not drive this fast in a nieghborhood and B) Autopilot would not drive into a house like this...
Whether if it is either autopilot or Fools Self Driving mode, if this driver did not have their hands on the wheel then he is absolutely in deep trouble.
Tesla themselves got into trouble after previous crashes and are finally telling their drivers to keep their hands on the wheel at all times, and toned down their false advertising.
Whatever they do or don't say in the fine print, they and their CEO clearly communicate to every Tesla owner I've met that you do not have to keep your hands on the wheel. tesla.com/fsd has a splash page showing someone driving the car hands-free. I do agree that the driver should also be held responsible.
Something is terribly broken in the software release process that constantly allows a worse user experience (including autonomous) as Tesla 'matures' its operations. I'm driving a Tesla, that recently added a 'warning' over a center-display showing the three rear-pointing cameras.
1. The cameras only show while operating in reverse;
2. The warning entirely obscures 30-50% of the view in one or more cameras;
3. The warning tells you that there is dirt or debris on the camera.
So, you are warned, that your vision, via the cameras could be better -- by deliberately worsening the view.
Used to work in the hardware industry. Friend worked for a company who was a component manufacturer about the mindset of Tesla - which affirmed to him why he'd never buy one:
"Hey, we sent you over the new firmware for the component, check it out." (The test suite for this component takes approximately 36 hours to execute.)
Three hours later:
"This is working so much better, thanks a lot!"
"???"
"Oh, we just flashed a car we have here and took it out for a drive."
I wonder if they will implement dead zones for self-driving. I was using a Lime scooter in a new city and when I entered a university campus, the scooter slowed to a crawl.
I don’t understand why a person would need FSD in a suburb street.
My nightmare has become reality. Software engineers now rule the world.
Life and death policies being made by people who last week you were arguing about whether using Kubernetes is worth it for a throwaway project, or whether tabs or spaces are more appropriate.
I just wonder how bad it will have to get before they decide we need to be reigned in. Going by how the 20th Century went, if it's anything like that we're in for a rough time before we get there--like, to a place where software engineers have to act like, well, engineers.
It's just not a serious technology. Street-by-street speed limit data exists, and where it doesn't, there are laws across the US as to what speed limits are based on area. Tons of map data to tell you where is residential and commercial, a street and a highway, an on-ramp, and so on.
But FSD doesn't abide by speed limits, and Waymo does, and it is truly self-driving.
So, it's all bullshit. Since day one, it's never been a real attempt at autonomous, legal, safe driving.
Recent lawsuits [1] seem to suggest both are. The driver committed manslaughter and should go to jail. The company sold a dangerous product that killed someone and should pay massive damages.
It's not even a recent thing. Before the turn of the century, jurors holding Cessna liable for incidents cause by pilot error was a common enough occurrence that it basically destroyed the light airplane industry through skyrocketing liability insurance prices.
You might not think so, but there are many people that believe Tesla marketing that FSD is better than paying attention. I struggle convincing my parents that they shouldn't drive if they are likely to fall asleep behind the wheel and it's helpful to categorize an FSD crash separately from a human driver crash.
But the car was fully self driving right? Fully, as completely driving by itself? If the person in the front seat is driving, the feature isn't really "full self driving" now is it?
Without knowing the "full" details of this crash, it is outrageous that Tesla calls it "Full Self-Driving (Supervised)" on their website. This is like naming a product "Healthy Dog Food (Rat poison)". If it requires supervision, it is not "full". Period.
I thought previous lawsuits would've forced Tesla to call it "Advanced Driver Assist" or something.
My SO just complained about the same when she found out the jug labled weed killer she bought in fact affects all plants and grass, not just weeds.
"Why isn't it called plant killer rather than weed killer", she griped. Why indeed.
I agree with the sentiment and don't like Elon at all. But I think there is a case to say the Full refers to features, so not only lane change / parking, but actual driving fully. If it said Perfect FSD or even High Quality FSD, that's definitely worse, but we have to acknowledge FSD can fully drive in many cases, not enough for me to be comfortable enough maybe, but comparing to rat poison seems like a stretch.
> "Full Self-Driving (Supervised)"
Water (Dry)
https://en.wikipedia.org/wiki/Dry_water
Of course dry-water is actual water. Full self-driving isn't.
"I thought previous lawsuits would've forced Tesla to call it "Advanced Driver Assist" or something"
Securities fraud
Problem: need evidence of intent, e.g., Musk knew software would fail
Example
LaMontagne v. Tesla
Claims
https://dn721902.ca.archive.org/0/items/gov.uscourts.cand.40...
Order
https://dn721902.ca.archive.org/0/items/gov.uscourts.cand.40...
Appeal
https://dn721902.ca.archive.org/0/items/gov.uscourts.cand.40...
False advertising
In progress, currently Tesla is appealing class certification
Example
In Re Tesla Advanced Driver Assistance Systems Litigation
Claims
https://dn721905.ca.archive.org/0/items/gov.uscourts.cand.40...
Tesla arguments
https://ia800708.us.archive.org/3/items/gov.uscourts.ca9.f97...
As long as Tesla doesn't insure their cars (whenever in self-driving mode) completely for the price of their FSD, it's not FSD, just scammy marketing by a chronic charlatan.
Here is a video from the youtube channel Better Biomed, from the scene, apparently it was his house: https://www.youtube.com/watch?v=6L_8x-Y0jqg
And doorbell footage of the crash.
https://www.youtube.com/watch?v=HR_Y_yf84m8
Holy shit, it looks like it was going at full highway speeds (at least) in a residential neighborhood. Doesn't look like it attempts to stop or steer away at any point. This seems like the outcome one would expect from a passed-out driver and a stuck accelerator. I'm honestly amazed that the driver survived.
Yeah, would be interesting to see what the cause is. These things do happen from time to time with cars
Sometimes, it just happens. People explode. Natural causes.
It looks like a culdesac? Why would it be driving that fast into a culdesac?
The telemetry will tell the story, not the clickbait news headline based on preliminary information. I seriously doubt FSD would drive into a house.
> telemetry will tell the story
Idk, the death-toll gap between Tesla and Waymo seems to tell a story enough.
There are very few (if any?) known deaths caused by FSD accidents.
Note: it’s important to distinguish between Autopilot and FSD here. “Autopilot” is Tesla’s old assisted driving stack that comes free in most vehicles and has had no significant updates in years. “FSD” is an entirely different software stack that only works with newer vehicles and that Tesla charges $$ for. It’s much more advanced and IMO a lot safer.
This article never mentions FSD, only Autopilot.
> There are very few (if any?) known deaths caused by FSD accidents
It’s tough to say given many data sources are aggregated. For what it’s worth, my parents’ car is a Tesla with FSD and I’ve stopped it from, off the top of my head, racing into a red-lit intersection, running over a small dog and running into a closing garage door.
I still use it. It mostly works. But I’m vigilantly monitoring it in a way that isn’t supported by Tesla’s marketing (which frequently shows drivers engaging it hands off).
The garage door I can certainly believe. Its AI brain just won’t be trained to look for that sort of thing.
I’m surprised about the red lights and animals, however: ours seems very cautious around any kind of live animal on the road, even braking and manoeuvring to avoid birds on the road. It’s not so good at avoiding the corpses of already dead ones, however (bump!).
> The garage door I can certainly believe. Its AI brain just won’t be trained to look for that sort of thing.
Why not? That’s likely to come up at minimum thousands of times per day, and likely vastly more as the system improves.
I get the impression that it kind of comprehends the world as a horizontal plane. It’s focused on objects on the surface and isn’t looking for hazards coming from above. Could that be improved? Sure! Is it a priority for them? Maybe not…
At minimum it needs to pay attention to railroad crossing bars and bridges.
They show a 2D representation to the driver because that’s good enough for drivers, but I wouldn’t assume that represents how the system operates internally at every stage. Even navigating requires the concept of bridges crossing roads without intersecting them.
It is indeed known to have issues at railway crossings:
https://news.ycombinator.com/item?id=47783427
AFAIK, there's no issue with bridges that cross roads, however. It will just ignore a road that's above or below the one you're travelling on that doesn't intersect with yours. Just like a human driver would do.
> It is indeed known to have issues at railway crossings:
At scale you can “have issues” and still work 99% of the time. If it’s completely incapable of handling railroad crossings that’s relatively easy to verify.
> AFAIK, there's no issue with bridges that cross roads, however. It will just ignore a road that's above or below the one you're travelling on that doesn't intersect with yours. Just like a human driver would do.
As a user of the system that’s worked it all out it’s not an issue for you, but getting that behavior from raw sensor data is non trivial.
> I’m surprised about the red lights and animals
Both only happened once. But they were shocking when they did. (To be fair, my Subaru tried to push me into oncoming traffic because it was avoiding a “collision” from the guy in the other lane turning weirdly. Turned at collision-avoidance feature off.)
> “Autopilot” is Tesla’s old assisted driving stack that comes free in most vehicles and has had no significant updates in years. “FSD” is an entirely different software stack that only works with newer vehicles and that Tesla charges $$ for. It’s much more advanced and IMO a lot safer.
I'm not sure the point you're making here - that just sounds like "Tesla couldn't care less about updating their software and it's still "not good". People it hits are still dead. "Oh well, it's not like Tesla had updated the software, so you can't blame them".
Oh, I absolutely agree. Ideally Telsa would dump the old Autopilot stack and put all cars (that have the hardware to support it) on a free version of FSD, nerfed so that it's roughly feature-equivalent to Autopilot.
That way everyone would get the safety advantages of FSD, including the really important stuff like better driver attention monitoring. Unfortunately, Tesla keeps Autopilot bad because it forces more people to pay for FSD, and in this case they seem to care more about profits than safety.
> Unfortunately, Tesla keeps Autopilot bad because it forces more people to pay for FSD, and in this case they seem to care more about profits than safety.
Not saying you're wrong, but I think "it forces people to pay for FSD" is a poor decision if true.
More likely it creates market confusion (as we see in this thread) about exactly which of the two products that Tesla advertises as things which drive your car for you is the more dangerous one and which is "the good one" (though from their taxis, I think the scare quotes are still justified even then).
I'm also not even sure if they do care about profits (or safety), given Musk's made about as much from selling Tesla shares as the company has made in lifetime profit, before getting shareholder approval for a deal to give him more shares. (Conditionally give, yes, but the nature of it all suggests share price is more important than profit or safety, and if you don't care about such things that makes it much easier for them to sell a million robots or whatever).
That the reporting is so tortured sort of shows this works. It’s irrelevant in a broad or legal context. But for the narrow audience of his fans, it carries merit.
When you say this, do you have in mind that "his fans" are the customers, or the shareholders?
I ask because he has made about as much from the latter as Tesla has made from the former.
You are on Hacker News, you don't think we know what FSD is?
The article only mentions Autopilot, but half the comments here are talking about FSD.
Makes sense
It's relevant because it arguably affects the driver's beliefs and behaviour, whether they are using FSD or just Autopilot
It's a separate upgrade but it's reasonable to assume the marketing may affect drivers using only Autopilot
If HN commenters are confusing the two then this only strengthens that argument
To me, having safety features that cost nothing for the manufacturer to enable sold as an option is a gross practice.
See: The software that enabled a secondary AOA sensor on the Boeing 737Max.
"There are very few (if any?) known deaths caused by FSD accidents"
https://theintercept.com/2023/01/10/tesla-crash-footage-auto...
No deaths in this one
Note the comment about Google's Waymo ceasing to use the term "full self-driving" after this accident
Why would they do that
Let the reader decide
does it makes sense to compare the two given that waymos are driven in a limited set of circumstances and most of the times below speeds that could kill you on a crash?
If Tesla was bold enough to intentionally auto disengage seconds before an accident (to avoid liability) then can they be trusted to maintain telemetry throughout dicey circumstances?
That's not what they do, and any crash happening within 5 seconds of fsd disengagement is considered to be using fsd.
Apparently they do at least intentionally disengage, often with little time to react. When they could instead start braking and alert the driver.
https://futurism.com/tesla-nhtsa-autopilot-report
I also don't trust the motives of a company that names something "Full Self Driving" knowing it's not fully self driving. Never mind their shenanigans around avoiding or disregarding regulations and reporting requirements.
Citation please.
That's how it is now, and (like most things safety-related) Tesla had to be dragged kicking and screaming into it. Just like the initial attention/distracted monitor only required your hand on the steering wheel once every fifteen minutes.
They also considered AEB-activation to mean that meant that FSD wasn't active, even if AEB only kicked in because of FSD's decisions.
Auto disengaging seconds before an acciden't isn't a Tesla-specific thing, it's part of the standard for any driving automation SAE level 3 or lower (https://www.sae.org/news/blog/sae-levels-driving-automation-...). The vehicle will return driving over to the human any time it's unable to determine what to do.
Are there any US federal or state laws stipulating some sort of black box style recording of data for accident investigation?
If not then I suspect a Tesla will turn out to be quite surprisingly forgetful about what it was up to in a road traffic collision.
RTC is a UK term that took over from RTA (road traffic accident) - it describes what happened rather than heading off into the weeds as to cause.
I don’t know about laws, but Teslas automatically record and save everything in the event of a collision. My brother was hit by another car while driving a Tesla a few years ago and it was very easy to retrieve video from the Tesla’s cameras and show who was at fault.
The reddit post from the grandchild says "autopilot", which is really just lane keeping and adaptive cruise control.
I believe autopilot would totally run into a house. It doesn't respect stop signs or red lights. If the house is at a T intersection the autopilot might try to drive right through it. I agree about FSD though
Adaptive cruise control certainly shouldn't run into a house, for the same reason it should detect a stationary (or slow) object in front of you and gradually slow to a stop before it.
And _we_ will never see the telemetry.
Maybe NTSB will.
Certainly Tesla will. And that will inform how their PR team responds to this collision.
Maybe some plaintiffs will if they can manage to subpoena the data from Tesla in some hypothetical future court case.
The reason I will never buy a Tesla is because it is one of the most advanced surveillance systems against the driver, but there is no one empowered to inspect the car / company (comprehensively, not superficially).
> I seriously doubt FSD would drive into a house.
Strawman. You seem to be insinuating that FSD intended to aim for a house. Usually the chain of events would start with something like “FSD was engaged on the road and in the intended lane” then “FSD lost track of the boundary of the lane” or maybe “FSD identified an obstruction in the lane so it maneuvered out of lane”.
Software with bugs in it? Preposterous!
Teslas do glithcy speed limits readings often, from GPS drift (ie. it thinks the car is on the highway) or may be deceived by some sign with numbers on the street.
When that happens and depending on Autopilot [1] settings, the car will shoot out to reach the new limit. If the driver is distracted enough not to notice or react in time, the car could crash. GPS inaccuracy does sound plausible given Teslas slow down if an abrupt change of direction is near (ie roundabout) per GPS and map data.
The house sits at a corner, not a culthesac.
1. It's not clear the driver allegedly said "autopilot" or "FSD"
Previously: A Florida jury found that flaws in Tesla's self-driving software were partly to blame for a crash that killed a 22-year-old woman in 2019 and severely injured her boyfriend.
The jury verdict requires Tesla to pay $243 million in punitive and compensatory damages to the parents of the woman and to her boyfriend.
https://www.reuters.com/world/us-judge-upholds-243-million-v...
Many Tesla owners I know are 100% convinced Assited Driving is 100% safe and reliable and can not go wrong.
Ok, 3 things.
1) I’ve put enough kms on FSD - it’s taken me across Australia a few times, probably 10,000kms in total - to know that it isn’t going to drive into a house.
2) Even if FSD is enabled, there’s loads of things you can do to create an accident like press the brakes or accelerator pedals, which doesn’t necessarily disengage FSD right away, so let’s just wait for the telemetry to get released.
3) Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
Sad for all involved.
Edit: my experience is HW4 by the way.
> 1) I’ve put enough kms on FSD to know that it isn’t going to drive into a house.
Is it possible FSD on this vehicle was a different version? Can't FSD change from one drive to the next, based on software updates or even external conditions?
Perhaps you drove in a different region with differing conditions?
> 3) Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
Why is it called "Full Self Driving" if the person behind the wheel must control (or even just monitor) the speed?
That’s exactly the thing. We’re on FSD 13 and America is on FSD 14, we’re always about a year behind. Yours is significantly better than ours (being RHD market), but ours still wouldn’t drive through a house.
You’re extremely confident about this just based on your experience of it not happening to you. It’s good that it hasn’t driven you into a house but that doesn’t mean it can’t fail in a way that does drive someone else into a house.
Well, I’ve seen it very confidently and correctly navigate rural unmarked roads, and drive around rocks and logs across the road and things like this. I find it hard to believe it would drive through a huge stationary object like a house, which I imagine is definitely in the training data.
I guess I’m not saying it’s impossible, but I think at this stage it’s improbable.
The computer does not see things like you see things ;)
The fact that you are saying "but houses are in the training data!" demonstrates your naivete.
Do you think there's some magic going on here?
I believe how good your FSD is also depends on how new your Tesla is because older ones are underpowered and maybe use an underpowered vision model
> Perhaps you drove in a different region with differing conditions?
Agreed.
He's driven it "across Australia". Which means long straight highways for hundreds and hundreds of miles at a time (then recharge, then do long straight highway for hundreds and hundreds of miles), and often fairly lightly trafficked:
https://www.google.com/maps/dir/Perth,+Western+Australia,+Au...
Like... absolutely shocking that FSD performs decently on well maintained, largely straight highway in generally clear desert conditions. Surely this is indicative of its behavior everywhere else! /s
Haha. Driving through Sydney is way worse than Los Angeles. We don’t all live out in the sticks with snakes in our boots.
I would agree, and know well, though I lived in Melbourne for two decades. But still, "driving across Australia" doesn't involve a lot of time spent in the Sydney (or Melbourne) CBDs.
“Wait for the telemetry”. Like that case where it took years and reverse engineering to pull the footage that Tesla claimed did not exist?
Surely, we can trust Tesla will be providing all relevant information to the authorities without delay.
https://www.washingtonpost.com/technology/2025/08/29/tesla-a...
"I've driven drunk enough times to know I'd never cause an accident"
> Regardless of who was controlling it, why did this guy let his car jump the kerb and go through a house? Why was he going fast enough?
There are decades worth of man-machines UX research to prove this: the more you lean on automated systems to perform a manual task, like specific vehicle operations, the more your reaction time and relefexes for that specific motor skill will suffer.
Non-level 4 driver assistance tech should only be used for helping prevent accidents and not pretend to be actual full self driving
"I coded like 500 line of C, I know for a fact segfaults cannot happen"
That's how you sound
"I haven't encountered a rare edge case, therefor the rare edge case must not exist"
Is there a Polymarket bet for this being Autopilot vs DWI? These types of accidents happen all the time, alcohol involved in many (most?).
Why do news outlets so often use the phrase "high rate of speed"? Speed isn't a discrete event, it can't really have a rate, unless it is a rate of change, in which case they would be referring to acceleration.
rate = distance / time
Yeah that's cringey, "high speed" is enough.
Sounds like a junior reporter trying to hit a word count.
There's absolutely no way this was on autopilot. A) autopilot would not drive this fast in a nieghborhood and B) Autopilot would not drive into a house like this...
Whether if it is either autopilot or Fools Self Driving mode, if this driver did not have their hands on the wheel then he is absolutely in deep trouble.
Tesla themselves got into trouble after previous crashes and are finally telling their drivers to keep their hands on the wheel at all times, and toned down their false advertising.
Hands on wheel doesn’t absolve the FSD from stopping the car before hitting the wall/object. And the person. This was completely avoidable.
The latest FSD does not attempt to check if your hands are on the wheel at all.
Does it do alternate checks for human awareness, like gaze detection?
Yes. But if you place screens strategically, the camera won't be able to see that you use them.
Whatever they do or don't say in the fine print, they and their CEO clearly communicate to every Tesla owner I've met that you do not have to keep your hands on the wheel. tesla.com/fsd has a splash page showing someone driving the car hands-free. I do agree that the driver should also be held responsible.
I can't imagine the lawyers going after him when they can go after the world's only trillionaire.
Something is terribly broken in the software release process that constantly allows a worse user experience (including autonomous) as Tesla 'matures' its operations. I'm driving a Tesla, that recently added a 'warning' over a center-display showing the three rear-pointing cameras.
1. The cameras only show while operating in reverse;
2. The warning entirely obscures 30-50% of the view in one or more cameras;
3. The warning tells you that there is dirt or debris on the camera.
So, you are warned, that your vision, via the cameras could be better -- by deliberately worsening the view.
Genius.
Made my day lol. Brilliant.
Cameras get dirty and it is unreasonable to think that they wouldn't. This would force my hand to clean the cameras. Why is this a bad thing?
Used to work in the hardware industry. Friend worked for a company who was a component manufacturer about the mindset of Tesla - which affirmed to him why he'd never buy one:
"Hey, we sent you over the new firmware for the component, check it out." (The test suite for this component takes approximately 36 hours to execute.)
Three hours later:
"This is working so much better, thanks a lot!"
"???"
"Oh, we just flashed a car we have here and took it out for a drive."
"?!?"
Oof.
I wonder if they will implement dead zones for self-driving. I was using a Lime scooter in a new city and when I entered a university campus, the scooter slowed to a crawl.
I don’t understand why a person would need FSD in a suburb street.
If FSD proves to be demonstrably safer in such a scenario then it would make sense to use it there.
All of the optimism surrounding this industry hinges on "ifs" that never materialize.
It's shocking. I can't understand why people repeatedly fall for it.
My nightmare has become reality. Software engineers now rule the world.
Life and death policies being made by people who last week you were arguing about whether using Kubernetes is worth it for a throwaway project, or whether tabs or spaces are more appropriate.
I just wonder how bad it will have to get before they decide we need to be reigned in. Going by how the 20th Century went, if it's anything like that we're in for a rough time before we get there--like, to a place where software engineers have to act like, well, engineers.
First people need to realize just how much software engineers have become more influential in their day to day lives.
They do that on bikes as well. At first, I was very confused about what was happening.
> don’t understand why a person would need FSD in a suburb street
My parents have a Tesla. It’s convenient. I engage my Subaru’s lane-keeping in suburbs, too, to reduce driver fatigue.
The Subaru behaves predictably. The Tesla is mostly more capable, but does something dumbfucked and dangerous every few dozen trips.
> to reduce driver fatigue
Fatigue is not the only, or even the biggest, risk to driving safely. Being distracted is frequently one of the biggest risks.
It’s also a convenient way to commit to the speed limit.
It's just not a serious technology. Street-by-street speed limit data exists, and where it doesn't, there are laws across the US as to what speed limits are based on area. Tons of map data to tell you where is residential and commercial, a street and a highway, an on-ramp, and so on.
But FSD doesn't abide by speed limits, and Waymo does, and it is truly self-driving.
So, it's all bullshit. Since day one, it's never been a real attempt at autonomous, legal, safe driving.
There used to be UI options for "by how much can FSD exceed the posted speed limit" and "should FSD actually come to a stop at a stop sign".
What is "autopilot mode"? Tesla doesn't use that term.
Someone needs to go to jail for this.
Either the driver or someone at Tesla (their pick, who cares).
This cannot go unpunished.
Machines can never be held accountable.
It seems your standards are pretty arbitrary.
Personally I want to know that a provable crime was actually committed before calling for someone to be jailed.
Also jailing basically doesn’t solve this problem?
Whenever the victim happens to be you or your loved ones, I'd love to see you maintaining that character you seem to portray.
And this is why justice is done by the state and not individual families?
What will happen to the liable person? Nothing.
Crazy I saw one of the victim's grandchildren post on Reddit yesterday and thought it should make the news.
The grandchildren's post: https://www.reddit.com/r/legaladvice/comments/1uardtu/driver...
The title is misleading and shifts blame from the driver to the machine.
Ultimately the driver is responsible.
Edit: For the folks who seem to think that this is marketed as unsupervised self driving, from Teslas own website it states
“Currently enabled features require active driver supervision and do not make the vehicle autonomous.”
https://www.tesla.com/fsd
Looks accurate to the content to me.
If it is Full-Self Driving, the machine is the driver. The company providing is ultimately responsible.
The tell is the the name. Full Self Driving.
> Ultimately the driver is responsible
Recent lawsuits [1] seem to suggest both are. The driver committed manslaughter and should go to jail. The company sold a dangerous product that killed someone and should pay massive damages.
[1] https://electrek.co/2026/04/16/tesla-facing-up-to-14-billion...
It's not even a recent thing. Before the turn of the century, jurors holding Cessna liable for incidents cause by pilot error was a common enough occurrence that it basically destroyed the light airplane industry through skyrocketing liability insurance prices.
If full self driving was active then the driver was Tesla.
You might not think so, but there are many people that believe Tesla marketing that FSD is better than paying attention. I struggle convincing my parents that they shouldn't drive if they are likely to fall asleep behind the wheel and it's helpful to categorize an FSD crash separately from a human driver crash.
Being responsible and being at fault are two different things.
But the car was fully self driving right? Fully, as completely driving by itself? If the person in the front seat is driving, the feature isn't really "full self driving" now is it?
"The name is a lie, but we have a disclaimer" is not generally considered adequate.