Important to note, this was a 2014 Model S. Most 2014s did not have any autopilot (just standard cruise control), and the ones that did used an off the shelf module from a third party company.
Vehicles are expected to last on the road for more than 9 years. Today, the average age of a vehicle on the road is a little over 12 years[1] - average.
It's a significant problem if vehicle's critical safety software/parts fail yet are still on the road. The future utopia of self driving vehicles can never become a reality if slightly older models kill people because they are no longer supported or updated.
The article says the vehicle was likely using Autopilot, which is a Tesla installed feature - not an unapproved 3rd party component. What led you to say that?
It's not an unapproved third party part. I can understand why you would think that given GP's misleading characterisation of a vendor-supplied OEM part as third party, but Teslas of this era came equipped with MobileEye technology as selected, specified and installed by Tesla.
It’s still a nonstandard modification to the original vehicle. It’s not that the car wore out or didn’t last. It was never designed to have FSD in the first place. The categorization of this as a reliability problem is not accurate.
Yes completely agree. But, this car never even pretended to advertise driving itself. Most likely, it advertised plain, vanilla "cruise control." At best, it also advertised lane keeping and radar cruise control.
It doesn’t matter what gadgets or gizmos are advertised on the car. You’re still responsible for driving the car. Sitting there waiting for the car to do something while it drives full speed into a stationary object is not the car’s fault.
As long as the car still had functioning brakes and steering, this is on the driver.
Your eyes don’t lie. If you’re paying attention like you’re supposed to, you have plenty of time to see any obstacles and take manual control if the car doesn’t slow down like it’s supposed to.
People are so desperate to blame literally anything except the person directly responsible.
Do you always specify when a part is made by a third party, or just when you're trying to insulate a trillion dollar corporation from liability? Because most parts in most cars are manufactured by "third party" vendors.
I mean it's SFGATE, which is more of a blog than proper journalism (they used to be associated with the SF Chronicle, but the two split a few years ago). The AP story [1] that they based their post on strikes a more professional tone:
> Messages were left Wednesday seeking comment from Tesla, which has disbanded its public relations department.
It's purely factual, just not formal, and your allegation of 'blind hatred' seems hyperbolic. It's odd to me how the tech community celebrates iconoclasm and disruption when things are going well but retreats into formalities when they're not.
It seems more unbecoming to me to keep marketing and selling a system called 'autopilot' without responding to the very reasonable criticism that it has a rather dubious safety record - 'According to AP News, “at least 15 Teslas have crashed into emergency vehicles nationwide while using the system.”' I don't have fine-grained accident statistics for emergency vehicles year by year, but that seems disproportionately high for a somewhat niche brand.
Pretty much. Also this paragraph borders on deliberate misinformation: "The car involved in the fire truck crash was a 2014 Tesla Model S. Recently, Tesla, which is owned by controversial Twitter CEO Elon Musk, recalled over 360,000 cars because of issues with the “Full Self-Driving” feature. The Model S was one of the models recalled."
No 2014 Model S's were sold with FSD. The product didn't exist yet. It's not subject to that recall notice, so the final sentence is just outright incorrect.
At this point I just give up on the ability of society to deal rationally with anything involving "Tesla". This was a decade-old car with a lanekeeping solution along the lines of what normal cars have today. Absolutely no one ever billed it as an autonomous vehicle. And, like normal cars today, sometimes they end up in fatal crashes. But no one cares if a Lexus on cruise control hits someone. That's not news.
Doesn't matter. My car drives me around anyway. Haters gonna hate, as it were.
Did you miss the news last year where Elon bought the company? Its not a public company anymore and is indeed owned by Musk. It has been delisted from the NYSE and shareholders were paid out at $54.20 a share.
> Absolutely no one ever billed it as an autonomous vehicle
The CEO of Mobileye thought very differently, and walked away from a Tesla partnership after Musk's rhetoric killed someone in 2016[1].
Mobileye CEO Amnon Shashua claiming that Tesla "was pushing the envelope in terms of safety" and that Autopilot is a "driver assistance system" and not a "driverless system"
Musk has been calling it Autopilot since day one, and promising features and safety levels that are completely unobtainable even today.
The fact that you are citing what Mobileye said Tesla said[1] instead of what Tesla said seems unconvincing. Link to marketing material or documentation or... anything (even a Musk tweet, I guess) that bills a 2014-era Tesla as an autonomous vehicle? You know that doesn't exist.
But it doesn't matter, because putting "Tesla" in the headlines drives people absolutely bananas. We'll literally engage in screaming matches on the internet about whether it's fair to blame brand new autonomy software for an accident involving a nine-year-old vehicle running a tech stack from a different company. Because Tesla.
[1] After a very public split where Mobileye lost their biggest customer, they still haven't really recovered and at this point are way behind Tesla, Waymo and Cruise.
You're arguing a straw man. Nowhere does it say "Full Autonomous Level 5" or whatever you are expecting - nor does it have to. They literally named it Autopilot and told people it could drive down the freeway without user input.
The facts are - Tesla installed this system, called it Autopilot, told the buyers it does magical things, and people have died and probably will continue to die.
I understand some people badly want Tesla to succeed no matter what. But this is silly... they chose not to call it a Driver Assist (like all the other car manufacturers) on purpose. Now Tesla reaps what it sowed...
> They literally named it Autopilot and told people it could drive down the freeway without user input.
And again I ask for a contemporary cite relevant to the car in question where Tesla (or Mobileye) claimed it would drive autonomously. They didn't make that claim, and you know it. So... I guess your whole argument rests on the name[1] of the product[2] and not the technology? If they called it "BananaDrive"
you wouldn't be upset? That seems doubtful.
Look: this is a 100% vanilla distracted driver accident under cruise control, and you know it. It's not the first. It won't be the last. It happens probably every day in other makes of cars with very similar systems, and you don't care. You only care because of the name of the company that made the car.
[1] Not to put a fine a point on it, but Chrysler was selling cars with "Auto-Pilot" in the 50's. People didn't freak out about those.
[2] Again, to repeat: not the product being sold today or at any time over the last seven years, with which you keep trying to conflate.
1. Some 2014 model S were retrofitted and subject to the recent recall. The news article simply said some model S were subject to a recent FSD recall. It did not say whether the one in the recent fatal accident was
2. Musk as recently as 2021 was bragging that “autopilot” specifically (not FSD) reduces crash rates by a “factor of 10”
This is completely 100% untrue
The stats Elon continues to tout currently about safety are a complete joke, and not normalized by road type, driver age, or condition
It’s like telling someone drug A given to cohort 1 had no side effects, while drug B given to cohort B caused tons of bad effects, and leaving out the fact that cohort A is 20 year old men and cohort B is 95 year old men
You’re not a victim. Please allow people to have normal discussions that are had about all cars.
> 1. Some 2014 model S were retrofitted and subject to the recent recall.
No, they weren't. That generation of cars can get a HW3 upgrade[1], but that's only offered to cars that purchased FSD. No FSD was installed on a 2014 vehicle, ergo no 2014 S's have HW3. Is it possible someone got an aftermarket part installed? Sure. Is it likely? Be real. It is correct for a journalist (or you) to say that this car was subject to the FSD recall? Absolutely not.
> 2. Musk as recently as 2021 was bragging that “autopilot” specifically (not FSD) reduces crash rates by a “factor of 10” [...] This is completely 100% untrue
They literally release a safety report with data to back that kind of thing up (honestly not sure which Musk statement you're talking about, you didn't cite, but the numbers are indeed in the order of magnitude range). Yelling about it in an HN comment from a 1-comment green account doesn't make it "100% untrue".
[1] Edit: post-some-research, even that's not necessarily true. Apparently early S's lacked the camera mounting points and can't be upgraded for mechanical reasons. Exactly where the cutover happened isn't clear.
Agreed. They just released their V11 FSD revision and it's amazing. Yet instead of seeing a post about it here people are talking about a 2014 Model S crash.
Don't sleep on FSD. I've had it for a year and a half and it's improved dramatically in that time.
Could you please stop posting unsubstantive comments and flamebait? You've unfortunately been doing it repeatedly. It's not what this site is for, and destroys what it is for, so we end up banning accounts that post this way. We don't really care about your views on the underlying topics but it matters a lot not to set the threads on fire and/or pour fuel on fires once started.
> Could you please stop posting unsubstantive comments and flamebait?
Define unsubstantive comments and flamebait. Is this not unsubstantive flamebait?
"What propaganda? That it would be nice if Britain wasn't being beaten in the supercomputing industry by a totalitarian dictatorship and warmongering dictatorship? As a Brit, I think it's just common sense rather than propaganda."
That's the definition of flamebait and propaganda. But it wasn't flagged. Why do you allow unsubstantive comments and flamebait while flagging every substantive comment that responds to and opposes flamebait?
This comment I replied to dehumanized arabs by calling them extremists and barbarians.
"I don't know much about Israel or the politics, but just want to say I deeply admire the country. A shining light of democracy, technological progress, innovation, and prosperity in the midst of many countries ruled by dictators and extremist barbarians more concerned with infighting and grandstanding than the progress of their people."
Who did you flag? You flagged me and let this bigot alone. It was only after days of complaints that you finally flagged him. Did you give him a warning? Of course not. You wrote "As for anti-Semitic comments, I've personally banned plenty of accounts for posting like that." You do realize that arabs are semites. How come you haven't banned boeingUH60 for his antisemitism? Or is antisemitism directed against arabs okay?
> We don't really care about your views on the underlying topics
No you do. Anyone who reads my flagged comments and the comments I replied to can see it.
You can say all kinds of nasty shit about russians, chinese, arabs, indians, etc but that's not flamebait. It's only flamebait when you defend them. So I'll keep fighting flamebait til you learn to do your job better. In the meantime I share your concern about unsubstantive comments and flamebait and will do my best to deal with them accordingly.
Edit:
You forgot to flag another of boeingUH60's flamebait in the same thread. You couldn't be bothered to read the thread to flag his responses?
I understand the feeling! but I need to explain that we don't see all the comments or read all the threads—not even close. In fact not even 10%. There's far too much for any human to read it all and we're pretty bare-bones over here. So when you see a post that ought to have been moderated but hasn't been, the likely explanation is that we didn't see it. You can help by flagging it or emailing us at hn@ycombinator.com.
That said, other people's behavior doesn't change the moderation response to your own account. It's just like when the cops pull you (or me, at least) over for speeding—it's always annoying because you know that other people were speeding much worse. Why didn't they get a ticket instead? The answer, in both cases, is simple if unsatisfying: it's just random. We need you to follow the rules regardless of what other accounts are doing. If they're breaking the rules repeatedly, they'll eventually get "pulled over" and dealt with too.
In case it helps, I have scolded and banned countless accounts for posting "nasty shit about russians, chinese, arabs, indians, etc" - that's totally not cool regardless of which population someone has a problem with. If you're trying to stand up for minority populations on HN, we're on your side in that respect. You can see a bit of the long history that I have with this dynamic here: https://news.ycombinator.com/chinamod. It's the same for other populations.
But you simply can't do it by flouting the rules. I know it's almost impossible not to be angry when encountering the double standards and one-sidedness that circulate around these issues, but if you try to defend against it with snark, flamewar, fulmination, and so on, it only makes things worse, and you put us in a position where we're forced to moderate you regardless of how right you are. It may not be fair, but it's inevitable.
I've written many past explanations about this in case any of it is helpful:
Are we publishing a story every-time someone gets in an accident with adaptive cruise control? It would be hard to buy a car in 2023 that didn't have adaptive cruise control with lane assist, which is all autopilot is.
Yes, that's the problem. Autopilot was named Autopilot which is just ACC. Full self driving is level 2 self driving. FSD should be renamed to Autopilot and Autopilot just take on another name or just ACC.
Just because one does not know the definition of Autopilot, doesn't make on right. If you look at the definition you will se that "Autopilots do not replace human operators."
> Just because one does not know the definition of Autopilot, doesn't make on right. If you look at the definition you will se that "Autopilots do not replace human operators."
If someone needs to know the definition of a feature in order to dispel myths based on the name, surely you can admit it's a bad/misleading name?
It'd get very onerous for the consumer to keep track of all the deceiving practices that every business could throw at them if we go down that path.
You are basically stating that it's the consumer's job to know when a marketing tactic is deceiving them, that simply doesn't work in a society... Or why do you think there are consumer protection laws?
What is your definition of autopilot? Are you okay with your pilot taking a nap on your next flight? Adaptive cruise control with lane assist is a nearly perfect analogy for what autopilot does on an aircraft, which is the origin of the word.
Does your adaptive cruise control drive into fire trucks that are partially blocking a freeway? I'm pretty sure mine would not- but I don't plan to test it (I never enable it).
Yea probably. I think people overestimate what these cars can do. I wouldn't want to test it, but after being into this stuff for a while it probably would. Longitudinal control is radar based in mainstream ACC, but I know for a fact that it would probably brake too late going 70mph with something sitting in the road . Open pilot's new longitudinal is vision based but it's probably not trained on a fire truck in the middle of a highway.
> but it's probably not trained on a fire truck in the middle of a highway
This is such a staggeringly, damning statement to be made about the state of self driving technology in 2023 - yet it's significance will be completely lost on so many people.
This strikes me as exactly the problem of these ML/AI based vision approaches.
There doesn't seem to be any higher level self protection layer on top of it to handle the general cases of the 1000s of things that have not been specifically trained yet/ever.
The current state of Tesla AP/FSD willfully ignores the saying "Things that never happened before happen all the time". If you have ever lived somewhere like NYC you will have seen things happen on the road that can never be trained.
I remember the example of FSD trying to turn left across the path of an oncoming trolley car. What is incredible was that the UI had rendered the trolley, so it knew a large object was moving towards it and yet decided to turn. Similar responses from Tesla booster of "well it was never trained on trolleys before, and now it is".
What is it going to do when a car goes the wrong way down the road? What is it going to do when someone is cruising in reverse half a block to get a parking spot? We can think of untrained scenarios all day. "Big moving object closing on my trajectory" seems like a general case being missed.
I guess the way I think about it is, if person is driving 70mph with ACC enabled and crash into a clearly visible roadway obstruction (firetrucks are highly visible even from a distance), that's totally on the driver who enabled the feature. It's not the job of that feature to prevent that kind of accident.
If other statements on this thread are correct, then this Tesla likely did not have autopilot and certainly not FSD- it had a pre-autopilot ACC system which led to a tiff between the vendor and tesla (possibly because tesla was being irresponsible while deploying it). So it's really hard to say this is a case of tesla cars automatic driving being systemically faulty. I do believe that to be the case, even for FSD, but it's entirely speculation.
I don't use the adaptive cruise control on my vehicle that has it, but I've read that many to most vehicles with adaptive cruise control will ignore stationary objects while travelling above a threshold speed. Radar, as implemented for adaptive cruise control, doesn't provide height information, so a big fire truck looks similar to a big information sign mounted above the roadway, and cruise control won't stop for either.
Tesla vehicles accelerate much faster than my vehicle too, which would give me more time to react if a vehicle ahead of me left my lane, exposing a stationary emergency vehicle (or other hazard), and the ACC acts to resume the set speed.
Play extraordinarily suggestive claims games, win extraordinarily suggestive claims prizes. Unfortunately that’s a much more forgiving consequence than the traditional equation.
We aren't talking about "adaptive cruise control". We're talking about "Full Self-Driving".
Your entire point is predicated on mislabeling Elon's fraudulent service. Yes, every time Elon's "Full Self-Driving" technology gets into a horrific accident and exposes the depths of the lies he peddles, it will become news.
If Tesla labeled the technology "adaptive cruise control", 90% of these stories would go away. No one cares about any other companies cruise control. They only care about "10X Tesla // autonomous taxi // Full Self Driving" future cars getting into accidents.
> investigators already suspect the crash took place when the driver of the Tesla was employing an automated driving system
Contradicts the headline, which implies that NHTSA told AP that they had specific reason to believe Autopilot was engaged.
Tesla Autopilot is 10x safer than driving without it [0]. We keep seeing speculative clickbait headlines because "Tesla" gets clicks. We also keep seeing confirmation a few months later that, oh yeah... turns out the driver was drunk and pressing the accelerator, and Autopilot wasn't engaged. Telemetry clears these up easily.
Also the "recall" (OTA software update) they mention at the end was for FSD, not Autopilot, and it was unrelated: they asked Tesla to eliminate human-like rolling stop behavior and to stop at yellow lights.
> Tesla Autopilot is 10x safer than driving without it [0].
No, Autopilot only works in relatively safe conditions where accident rates are already lowest (highway driving). A fair comparison would be accidents per mile for Autopilot vs. Accidents per mile with human drivers in similar conditions, where Autopilot is almost certainly not significantly better, and that's why Tesla doesn't share that data.
> Contradicts the headline, which implies that NHTSA told AP that they had specific reason to believe Autopilot was engaged
No it doesn't unless both the original headline and HN updated in the past 3 minutes.
Exactly - theres huge selection bias at play, plus Tesla selecting what numbers to release to prove how much safer they are.
If you are using AP as intended, then you are using it on the easiest places to drive - limited access highways in good weather.
All of the accidents my family & I have been in over the course of our lives have been in places you are not supposed to be using AP.
Crossing traffic, left hand turns, mall parking lots, etc are where the majority of accidents happen.
There is also arguably selection bias in the driver profiles.
Young males (16~20ish) and older people (75+) have significantly higher accident rates but are less likely to be driving a Tesla, on AP.
I don’t think many people who tried full self driving believe the “better than a human driver” thing. The ridiculous idea seems to be mostly in fan circles, not among Tesla owners.
Tesla themselves say that the tech is meant to assist the driver, and not be FSD in the true sense of the words. And even with assistance, it only makes driving somewhat safer, and somewhat more dangerous in other areas.
I wouldn't want to use FSD in the city with my kids in the car, let's say it this way. It's an amazing technology to my SWE brain. And from a business perspective, it disrupted the automotive industry. We should appreciate it for what it is, not try to make it what it is not. And full self driving it is not. Reasonable self driving, maybe it is. It would be more accurate to call it full driving assist.
I'm sorry, but this reads like it was written by Tesla PR. As the other child points out, this is conveniently making rather specious/unproven assertions.
There is also this analysis [1] which normalises crash data.
Although Level 2 vehicles were claimed to have a 43% lower crash rate than Level 1 vehicles, their improvement was only 10% after controlling for different rates of freeway driving.
Direct comparison with general public driving was impossible due to unclear crash severity thresholds in the manufacturer’s reports, but analysis showed that controlling for driver age would increase reported crash rates by 11%.
Ok, and how many drivers have crashed into emergency vehicles when being driven by humans? Without that number, at a minimum, the article conveys very little meaningful information.
Also, anyone who has ever used an actual aircraft autopilot knows that they are actually very similar to what Tesla has. They relieve you of making constant control inputs need near constant supervision. Is the marketing deceptive, perhaps, but its not incorrect.
You are absolutely right. When I look at the ridiculous numbers of preventable vehicle deaths in our country, caused almost entirely by carelessness, I do wish that getting a driver's license was much much harder and that we actually bothered to enforce traffic laws consistently.
84 comments
[ 3.3 ms ] story [ 128 ms ] threadiirc Mobileye right?
It's a significant problem if vehicle's critical safety software/parts fail yet are still on the road. The future utopia of self driving vehicles can never become a reality if slightly older models kill people because they are no longer supported or updated.
[1] https://hedgescompany.com/blog/2022/02/how-old-are-cars/#:~:....
Where is the disconnect happening here? Not driving full-speed into stationary objects strikes me as "Driver Assist 101" stuff...
As long as the car still had functioning brakes and steering, this is on the driver.
People are so desperate to blame literally anything except the person directly responsible.
I think it's relevant because back then Tesla did not advertise "FSD", whatever that means.
If you look at my comment history, you can see I routinely criticize their practices around FSD, not insulate them from liability.
Like the 3d view in many cars is just a Bosch module.
That's some salt.
Apparently, blind hatred for someone can excuse an awful lot of illicit behavior...
> Messages were left Wednesday seeking comment from Tesla, which has disbanded its public relations department.
[1] https://apnews.com/article/tesla-firetruck-autopilot-investi...
It seems more unbecoming to me to keep marketing and selling a system called 'autopilot' without responding to the very reasonable criticism that it has a rather dubious safety record - 'According to AP News, “at least 15 Teslas have crashed into emergency vehicles nationwide while using the system.”' I don't have fine-grained accident statistics for emergency vehicles year by year, but that seems disproportionately high for a somewhat niche brand.
No 2014 Model S's were sold with FSD. The product didn't exist yet. It's not subject to that recall notice, so the final sentence is just outright incorrect.
At this point I just give up on the ability of society to deal rationally with anything involving "Tesla". This was a decade-old car with a lanekeeping solution along the lines of what normal cars have today. Absolutely no one ever billed it as an autonomous vehicle. And, like normal cars today, sometimes they end up in fatal crashes. But no one cares if a Lexus on cruise control hits someone. That's not news.
Doesn't matter. My car drives me around anyway. Haters gonna hate, as it were.
this reporter didn't know it's a public company, not owned by Musk?
The CEO of Mobileye thought very differently, and walked away from a Tesla partnership after Musk's rhetoric killed someone in 2016[1].
Mobileye CEO Amnon Shashua claiming that Tesla "was pushing the envelope in terms of safety" and that Autopilot is a "driver assistance system" and not a "driverless system"
Musk has been calling it Autopilot since day one, and promising features and safety levels that are completely unobtainable even today.
[1] https://en.wikipedia.org/wiki/Mobileye#Tesla
But it doesn't matter, because putting "Tesla" in the headlines drives people absolutely bananas. We'll literally engage in screaming matches on the internet about whether it's fair to blame brand new autonomy software for an accident involving a nine-year-old vehicle running a tech stack from a different company. Because Tesla.
[1] After a very public split where Mobileye lost their biggest customer, they still haven't really recovered and at this point are way behind Tesla, Waymo and Cruise.
The facts are - Tesla installed this system, called it Autopilot, told the buyers it does magical things, and people have died and probably will continue to die.
I understand some people badly want Tesla to succeed no matter what. But this is silly... they chose not to call it a Driver Assist (like all the other car manufacturers) on purpose. Now Tesla reaps what it sowed...
And again I ask for a contemporary cite relevant to the car in question where Tesla (or Mobileye) claimed it would drive autonomously. They didn't make that claim, and you know it. So... I guess your whole argument rests on the name[1] of the product[2] and not the technology? If they called it "BananaDrive" you wouldn't be upset? That seems doubtful.
Look: this is a 100% vanilla distracted driver accident under cruise control, and you know it. It's not the first. It won't be the last. It happens probably every day in other makes of cars with very similar systems, and you don't care. You only care because of the name of the company that made the car.
[1] Not to put a fine a point on it, but Chrysler was selling cars with "Auto-Pilot" in the 50's. People didn't freak out about those.
[2] Again, to repeat: not the product being sold today or at any time over the last seven years, with which you keep trying to conflate.
1. Some 2014 model S were retrofitted and subject to the recent recall. The news article simply said some model S were subject to a recent FSD recall. It did not say whether the one in the recent fatal accident was
2. Musk as recently as 2021 was bragging that “autopilot” specifically (not FSD) reduces crash rates by a “factor of 10”
This is completely 100% untrue
The stats Elon continues to tout currently about safety are a complete joke, and not normalized by road type, driver age, or condition
It’s like telling someone drug A given to cohort 1 had no side effects, while drug B given to cohort B caused tons of bad effects, and leaving out the fact that cohort A is 20 year old men and cohort B is 95 year old men
You’re not a victim. Please allow people to have normal discussions that are had about all cars.
No, they weren't. That generation of cars can get a HW3 upgrade[1], but that's only offered to cars that purchased FSD. No FSD was installed on a 2014 vehicle, ergo no 2014 S's have HW3. Is it possible someone got an aftermarket part installed? Sure. Is it likely? Be real. It is correct for a journalist (or you) to say that this car was subject to the FSD recall? Absolutely not.
> 2. Musk as recently as 2021 was bragging that “autopilot” specifically (not FSD) reduces crash rates by a “factor of 10” [...] This is completely 100% untrue
They literally release a safety report with data to back that kind of thing up (honestly not sure which Musk statement you're talking about, you didn't cite, but the numbers are indeed in the order of magnitude range). Yelling about it in an HN comment from a 1-comment green account doesn't make it "100% untrue".
[1] Edit: post-some-research, even that's not necessarily true. Apparently early S's lacked the camera mounting points and can't be upgraded for mechanical reasons. Exactly where the cutover happened isn't clear.
Also, I love my car driving me around. It's the reason I bought it and it is amazing.
Don't sleep on FSD. I've had it for a year and a half and it's improved dramatically in that time.
If you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
Define unsubstantive comments and flamebait. Is this not unsubstantive flamebait?
"What propaganda? That it would be nice if Britain wasn't being beaten in the supercomputing industry by a totalitarian dictatorship and warmongering dictatorship? As a Brit, I think it's just common sense rather than propaganda."
https://news.ycombinator.com/item?id=35067967
That's the definition of flamebait and propaganda. But it wasn't flagged. Why do you allow unsubstantive comments and flamebait while flagging every substantive comment that responds to and opposes flamebait?
This comment I replied to dehumanized arabs by calling them extremists and barbarians.
"I don't know much about Israel or the politics, but just want to say I deeply admire the country. A shining light of democracy, technological progress, innovation, and prosperity in the midst of many countries ruled by dictators and extremist barbarians more concerned with infighting and grandstanding than the progress of their people."
https://news.ycombinator.com/item?id=34997427
Who did you flag? You flagged me and let this bigot alone. It was only after days of complaints that you finally flagged him. Did you give him a warning? Of course not. You wrote "As for anti-Semitic comments, I've personally banned plenty of accounts for posting like that." You do realize that arabs are semites. How come you haven't banned boeingUH60 for his antisemitism? Or is antisemitism directed against arabs okay?
> We don't really care about your views on the underlying topics
No you do. Anyone who reads my flagged comments and the comments I replied to can see it.
You can say all kinds of nasty shit about russians, chinese, arabs, indians, etc but that's not flamebait. It's only flamebait when you defend them. So I'll keep fighting flamebait til you learn to do your job better. In the meantime I share your concern about unsubstantive comments and flamebait and will do my best to deal with them accordingly.
Edit:
You forgot to flag another of boeingUH60's flamebait in the same thread. You couldn't be bothered to read the thread to flag his responses?
https://news.ycombinator.com/item?id=35003295
That said, other people's behavior doesn't change the moderation response to your own account. It's just like when the cops pull you (or me, at least) over for speeding—it's always annoying because you know that other people were speeding much worse. Why didn't they get a ticket instead? The answer, in both cases, is simple if unsatisfying: it's just random. We need you to follow the rules regardless of what other accounts are doing. If they're breaking the rules repeatedly, they'll eventually get "pulled over" and dealt with too.
In case it helps, I have scolded and banned countless accounts for posting "nasty shit about russians, chinese, arabs, indians, etc" - that's totally not cool regardless of which population someone has a problem with. If you're trying to stand up for minority populations on HN, we're on your side in that respect. You can see a bit of the long history that I have with this dynamic here: https://news.ycombinator.com/chinamod. It's the same for other populations.
But you simply can't do it by flouting the rules. I know it's almost impossible not to be angry when encountering the double standards and one-sidedness that circulate around these issues, but if you try to defend against it with snark, flamewar, fulmination, and so on, it only makes things worse, and you put us in a position where we're forced to moderate you regardless of how right you are. It may not be fair, but it's inevitable.
I've written many past explanations about this in case any of it is helpful:
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...
If I had to bet, I’d bet on Teslas having better numbers. A guy intentionally drove off a cliff in a murder-suicide attempt and failed, after all.
Pretty sure there is government agency that tracks these data…
Autopilot results in 10x fewer accidents than a car without it.
If that’s all autopilot is, why not market it as adaptive cruise control? Search in your heart, I believe you know the answer.
https://www.businessinsider.com/elon-musk-tesla-worth-basica...
If someone needs to know the definition of a feature in order to dispel myths based on the name, surely you can admit it's a bad/misleading name?
One warning though: my product is not safe for toddlers.
You are basically stating that it's the consumer's job to know when a marketing tactic is deceiving them, that simply doesn't work in a society... Or why do you think there are consumer protection laws?
This is such a staggeringly, damning statement to be made about the state of self driving technology in 2023 - yet it's significance will be completely lost on so many people.
There doesn't seem to be any higher level self protection layer on top of it to handle the general cases of the 1000s of things that have not been specifically trained yet/ever.
The current state of Tesla AP/FSD willfully ignores the saying "Things that never happened before happen all the time". If you have ever lived somewhere like NYC you will have seen things happen on the road that can never be trained.
I remember the example of FSD trying to turn left across the path of an oncoming trolley car. What is incredible was that the UI had rendered the trolley, so it knew a large object was moving towards it and yet decided to turn. Similar responses from Tesla booster of "well it was never trained on trolleys before, and now it is".
What is it going to do when a car goes the wrong way down the road? What is it going to do when someone is cruising in reverse half a block to get a parking spot? We can think of untrained scenarios all day. "Big moving object closing on my trajectory" seems like a general case being missed.
If other statements on this thread are correct, then this Tesla likely did not have autopilot and certainly not FSD- it had a pre-autopilot ACC system which led to a tiff between the vendor and tesla (possibly because tesla was being irresponsible while deploying it). So it's really hard to say this is a case of tesla cars automatic driving being systemically faulty. I do believe that to be the case, even for FSD, but it's entirely speculation.
Tesla vehicles accelerate much faster than my vehicle too, which would give me more time to react if a vehicle ahead of me left my lane, exposing a stationary emergency vehicle (or other hazard), and the ACC acts to resume the set speed.
Your entire point is predicated on mislabeling Elon's fraudulent service. Yes, every time Elon's "Full Self-Driving" technology gets into a horrific accident and exposes the depths of the lies he peddles, it will become news.
If Tesla labeled the technology "adaptive cruise control", 90% of these stories would go away. No one cares about any other companies cruise control. They only care about "10X Tesla // autonomous taxi // Full Self Driving" future cars getting into accidents.
Contradicts the headline, which implies that NHTSA told AP that they had specific reason to believe Autopilot was engaged.
Tesla Autopilot is 10x safer than driving without it [0]. We keep seeing speculative clickbait headlines because "Tesla" gets clicks. We also keep seeing confirmation a few months later that, oh yeah... turns out the driver was drunk and pressing the accelerator, and Autopilot wasn't engaged. Telemetry clears these up easily.
Also the "recall" (OTA software update) they mention at the end was for FSD, not Autopilot, and it was unrelated: they asked Tesla to eliminate human-like rolling stop behavior and to stop at yellow lights.
[0] https://www.tesla.com/VehicleSafetyReport
No, Autopilot only works in relatively safe conditions where accident rates are already lowest (highway driving). A fair comparison would be accidents per mile for Autopilot vs. Accidents per mile with human drivers in similar conditions, where Autopilot is almost certainly not significantly better, and that's why Tesla doesn't share that data.
> Contradicts the headline, which implies that NHTSA told AP that they had specific reason to believe Autopilot was engaged
No it doesn't unless both the original headline and HN updated in the past 3 minutes.
If you are using AP as intended, then you are using it on the easiest places to drive - limited access highways in good weather.
All of the accidents my family & I have been in over the course of our lives have been in places you are not supposed to be using AP.
Crossing traffic, left hand turns, mall parking lots, etc are where the majority of accidents happen.
There is also arguably selection bias in the driver profiles. Young males (16~20ish) and older people (75+) have significantly higher accident rates but are less likely to be driving a Tesla, on AP.
Tesla themselves say that the tech is meant to assist the driver, and not be FSD in the true sense of the words. And even with assistance, it only makes driving somewhat safer, and somewhat more dangerous in other areas.
I wouldn't want to use FSD in the city with my kids in the car, let's say it this way. It's an amazing technology to my SWE brain. And from a business perspective, it disrupted the automotive industry. We should appreciate it for what it is, not try to make it what it is not. And full self driving it is not. Reasonable self driving, maybe it is. It would be more accurate to call it full driving assist.
Although Level 2 vehicles were claimed to have a 43% lower crash rate than Level 1 vehicles, their improvement was only 10% after controlling for different rates of freeway driving.
Direct comparison with general public driving was impossible due to unclear crash severity thresholds in the manufacturer’s reports, but analysis showed that controlling for driver age would increase reported crash rates by 11%.
[1] https://www.tandfonline.com/doi/abs/10.1080/19439962.2023.21...
Also, anyone who has ever used an actual aircraft autopilot knows that they are actually very similar to what Tesla has. They relieve you of making constant control inputs need near constant supervision. Is the marketing deceptive, perhaps, but its not incorrect.