Tesla FSD drives down railroad tracks, gets hit by train

[Donald Stanfield]

Tesla has always called it "Full Self Driving"
That's a lie.

This is the biggest issue with the whole Tesla thing for me. This is akin to going to a restaurant and ordering a steak that’s been dry aged for 30 days, then when it comes its a piece of meat and a dry aging cabinet and the waiter tells you “well you have to do the dry aging yourself”.

I asked google Gemini and here is the definition of false advertising. I took the liberty of highlighting the relevant information.

Sponsored

 

[SoCal Rob]

It should be called just Supervised Driving.
No Full.
No Self.
Those are lies.

Supervised Driving Assistance

Tesla Advanced Traffic Assistant (Supervised)
I mean, how could Elon NOT go for bragging about his S3XY TATA(S)?

GigaCyberDrive

There are so many options for naming the product which don’t oversell the capabilities. GM and Ford both emphasize the fact that their systems are an extension of cruise control (SuperCruise and BlueCruise, respectively) rather than self driving. Other companies seem to incorporate highway or driving and assist or assistance. Tesla seems intent on overselling the capabilities of the system and then they are shocked, I mean shocked, when someone takes them at their word about FSD.

 

[captainjp]

Again, it's both hilarious and sad that people who have no experience with this think they should express opinions on it. Which are inherently useless because of the complete ignorance expressed right in your post. What's the point? That you as a normal driver have to make many adjustments per second, and the fatigue of this is well proven. You also have one set of eyes that needs to monitor a 360 degree circle. ADAS takes a massive workload off the human, and now the perception and ability to be focused on major threats is much higher. I arrive where I'm going refreshed, not beat down, and during the drive I'm FAR more aware of all the things that could be a problem. Such as seeing a problem a quarter mile ahead, which is easier because I'm not focused on the cars immediately around me, or on simple lane-keeping.

Laughable. Your complacency and smugness will bite you in the ass one day.

FSD is not there to keep you safe. you are naive to think so. It’s there to make the company and its CEO money. That’s why it’s a pay-to-use feature.

Most of the updates that have been made to the system have been in response to incidents and or accidents. Sure you can call it learning, but in actuality this is to limit the company’s liability. FSD disengages when it detects an imminent collision. Again, liability.

You can tout its abilities and advantages it has over humans, but it still doesn’t change my trust in it. No matter what ADAS it is. Your overconfidence in the system will eventually lead to over-reliance

 

[COdogman]

Again, it's both hilarious and sad that people who have no experience with this think they should express opinions on it. Which are inherently useless because of the complete ignorance expressed right in your post. What's the point? That you as a normal driver have to make many adjustments per second, and the fatigue of this is well proven. You also have one set of eyes that needs to monitor a 360 degree circle. ADAS takes a massive workload off the human, and now the perception and ability to be focused on major threats is much higher. I arrive where I'm going refreshed, not beat down, and during the drive I'm FAR more aware of all the things that could be a problem. Such as seeing a problem a quarter mile ahead, which is easier because I'm not focused on the cars immediately around me, or on simple lane-keeping.

What a load of entitled bullshit.

 

[JM.]

FSD is not there to keep you safe.

At no time did I claim it was.

That’s why it’s a pay-to-use feature.

Nowhere in my post did I mention FSD. This is why it's very hard to communicate here when people assume so much and then go attack the straw men they've made up in their own minds. I carefully used neutral language, ADAS, not FSD. I'm talking about not-pay-to-use features.

but it still doesn’t change my trust in it.

Your complacency and smugness will bite you in the ass some day. You could be benefiting from a safety feature and you think you're such a great driver you don't need it.

 

[pamalabama]

As I read it, Snopes said they couldn't confirm the data.

That's a long way from calling the study "ignorant."

Best wishes!

IIHS claims that tesla fatality rate by vehicle is below average. Makes sense for a car with modern safety features and best in class emergency braking.

ISEECARS refuses to respond about how they calculated their data.

When you use the data for tesla mileage provided by lars moravey, the fatality rates from tesla are below average.

ISEECARS analyzed cars from 2018-2022. Model Y was not available for a number of those years.

There is no chance that tesla is the most fatal car brand or even close to it. Which considering tesla only makes "fast" cars is pretty good

 

[KootenayEV]

School zone is a very strict rule, i.e. you go 20mph and not 1 mph over. It takes a lot of parameters and it's relatively not a hard problem compared to fundamentally driving in a safe way.

Besides a school zone is a thing that many people go an entire year without ever driving through

Plus there are different types of school zones. Some flashing, some with a sign that lists specific times. It's not a simple thing to solve and may involve every school zone being manually programmed.

Tesla focuses on the real issues that affect most people

You are a danger to society with that kind of logic. Elon has instilled this kind of thinking in Tesla as well for "FSD". Every single person in my town that is a driver goes through at least one school zone every single day. What a stupid argument.

My newly driving teenager can deal with every permutation of school zone. Until "FSD" can do the same, it should not be legal. I have a 13 year old too who gets on a bus every day with the flashing lights and stop sign that comes out. Every other single road user gets it and respects it. I find it deeply troubling that "FSD" has a problem with such a basic interpretation.

 

[Dark-Fx]

I think this article title needs to get updated. "Train drives down Tesla roadway, gets hit by Tesla"

 

[COdogman]

IIHS claims that tesla fatality rate by vehicle is below average. Makes sense for a car with modern safety features and best in class emergency braking.

ISEECARS refuses to respond about how they calculated their data.

When you use the data for tesla mileage provided by lars moravey, the fatality rates from tesla are below average.

ISEECARS analyzed cars from 2018-2022. Model Y was not available for a number of those years.

There is no chance that tesla is the most fatal car brand or even close to it. Which considering tesla only makes "fast" cars is pretty good

Weird how Tesla gets a pass from you for data integrity - they can combine Autopilot and FSD stats, provide no transparency at all and you are fine with it, but ISEECARS must jump through hoops for you

 

[pamalabama]

Weird how Tesla gets a pass from you for data integrity - they can combine Autopilot and FSD stats, provide no transparency at all and you are fine with it, but ISEECARS must jump through hoops for you

Because ISEECARS does not know how many miles a vehicle has driven. They have no data on this. They have a stupid metric that involves looking at used cars for sale.

If you know anything about 2018-2022 tesla were in high demand. People were selling them for what they paid because there was a tax incentive at the time. They were hard to come by

So you had one camp of people who were buying them to resell. And then another camp who were buying them to drive them.

So therefore the miles per fatality looks ridiculously off because you are only looking at the mileage of used vehicles

 

[JM.]

Man, if nothing else, we can agree that this thread is off the rails?

 

[Dark-Fx]

TLDR Straw man fallacy.

No one expected "fully formed, able to deal with absolutely every edge case". What should be expected is transparency, and as much safety as possible.

There is literally no evidence that FSD/ Autopilot is safer than human drivers. None. Even in your longwinded post you said "already probably"

Remember, there is a world full of humans that didn't sign up to be guinea pigs for Leon's vanity projects.

There's verifiable proof that FSD is safer, in certain situations, than a driver that is not paying attention. Ironically this particular post is about one of the situations where it is not. Because a driver that's not paying attention won't randomly turn down train tracks.

 

[pamalabama]

There's verifiable proof that FSD is safer, in certain situations, than a driver that is not paying attention. Ironically this particular post is about one of the situations where it is not. Because a driver that's not paying attention won't randomly turn down train tracks.

FSD won't turn down train tracks either. This whole post is off.

I suspect the user is trying to blame FSD

If FSD turned down a train track at least it would turn down the right track and not the left. But the right track is the active one

 

[COdogman]

Because ISEECARS does not know how many miles a vehicle has driven. They have no data on this. They have a stupid metric that involves looking at used cars for sale.

If you know anything about 2018-2022 tesla were in high demand. People were selling them for what they paid because there was a tax incentive at the time. They were hard to come by

So you had one camp of people who were buying them to resell. And then another camp who were buying them to drive them.

So therefore the miles per fatality looks ridiculously off because you are only looking at the mileage of used vehicles

Yeah, people only die in new vehicles... that is the standard everyone should use from now on

 

[JM.]

Because a driver that's not paying attention won't randomly turn down train tracks.

It's hilariously easy to prove this wrong.

But I didn't mean to derail your other, valid points.

Sponsored