Hands Free!

[CrazyOne]

Like everything else statistical data, FSD safety is measured as a number of events, disengagements, interventions, etc. per a given data sampling. In this case FSD is compared to the same data size, e.g. perhaps 100k miles driven against the same results with human driving.
A terrible, bad experience with FSD does not mean FSD is bad or worse than human driving anymore than somebody saying that his great experience with FSD means that it is indeed perfect.

Statistical data shows that FSD driving has fewer safety incidents compared to human driving per a given number of miles driven.

Another similar point. ICE cars catch fire at a much higher rate than EVs. That's not because there are more ICE cars than EVs. It does mean that given a sampling of, say 100k ICE and 100k EVs, the former sampling of ICE cars will experience far more fires than the 100k pool of EVs. Same approach for FSD.

I will answer the question, since you skipped it. Source is Elmo. Enough said. Range lies are fully out. Autopilot/FSD is getting there.

I can talk a lot, without much facts but this is a real source.

"We are able to look at insurance claims data. We have been able to look at vehicles with and without these (systems) and determine there is no reduction in claims as a result of these more advanced systems," IIHS President David Harkey told Reuters.

https://www.reuters.com/business/au...arks-automated-driving-technology-2024-03-12/

Sponsored

 

[therealcmj]

Statistical data shows that FSD driving has fewer safety incidents compared to human driving per a given number of miles driven.

This is a bold claim. Yes, Tesla says this. But the stat as provided by Tesla is very misleading since you're not comparing apples to apples. Specifically, you're comparing FSD which only works in some situations to humans who drive in all situations. What's more (unless something has changed since I read up on this) Tesla has been unwilling to release raw data for others to analyze. Including to the NHTSA. And they've been under investigation by the NHTSA over FSD's safety and their claims about it. And weirdly DOGE just ripped through the NHTSA - disproportionately firing those assessing the safety and risks of automated driving. Well, maybe not so weirdly.

https://www.reuters.com/business/au...over-full-self-driving-collisions-2024-10-18/

https://arstechnica.com/cars/2025/04/car-safety-experts-at-nhtsa-which-regulates-tesla-axed-by-doge/

 

[bigsky]

I will answer the question, since you skipped it. Source is Elmo. Enough said. Range lies are fully out. Autopilot/FSD is getting there.

I can talk a lot, without much facts but this is a real source.

"We are able to look at insurance claims data. We have been able to look at vehicles with and without these (systems) and determine there is no reduction in claims as a result of these more advanced systems," IIHS President David Harkey told Reuters.

https://www.reuters.com/business/au...arks-automated-driving-technology-2024-03-12/

Ah, I see that you are making conclusions about FSD according to a given person, not on the merits. I don't care for [person]; ergo, FSD is a pos. That's not the way it works.

Say what you will, quote whom you will, the data says it all: with Tesla FSD:

• Average crashes on non-highway sections of road: 0.31 times per million miles (compared to the average American's 1.53 times per million miles).
• Airbag-deployed crashes: About every 3.2 million miles (about five times safer than the US average of 600,000 miles per police-reported crash).

 

[bigsky]

This is a bold claim. Yes, Tesla says this. But the stat as provided by Tesla is very misleading since you're not comparing apples to apples. Specifically, you're comparing FSD which only works in some situations to humans who drive in all situations. What's more (unless something has changed since I read up on this) Tesla has been unwilling to release raw data for others to analyze. Including to the NHTSA. And they've been under investigation by the NHTSA over FSD's safety and their claims about it. And weirdly DOGE just ripped through the NHTSA - disproportionately firing those assessing the safety and risks of automated driving. Well, maybe not so weirdly.

https://www.reuters.com/business/au...over-full-self-driving-collisions-2024-10-18/

https://arstechnica.com/cars/2025/04/car-safety-experts-at-nhtsa-which-regulates-tesla-axed-by-doge/

I am not comparing apples to apples, all right. Indeed, Rivian ADAS and Tesla FSD is not apples-to-apples. Not even close.
DO you own a Tesla, have you ever owned one with FSD?
FSD only works in some situations? Really?
I think you meant to say Rivian ADAS only works in some situations.
Tesla FSD works on the vast majority of situations; Tesla FSD will take me door-to-door on just about every single solitary road in America, rural, highway, etc. Rivian ADAS works, whatever that means, only on some little mapped roads. That's it. And the latter is just a glorified cruise control at this point.

FSD is safer than human driving, full stop. It is not perfect. Indeed, it never may be, but a computer shall always be better than a human at doing things. That is settled.

 

[VSG]

Indeed, Rivian ADAS and Tesla FSD is not apples-to-apples.

Then why do you keep comparing them? Rivian is not even trying to do FSD, yet you are constantly disparaging Rivian for being

just a glorified cruise control at this point.

But that's exactly what it was intended to be. It's called Highway Assist because ... it's intended to assist you on highways. And that's what it does.

Tesla on the other hand calls their system "Full Self-Driving" and even today their implementation doesn't live up to their own hype. Regardless of how good it is, it doesn't fulfill their promise.

 

[Glogic]

Musk has real things to concern himself with.

FSD on my Tesla will pull out of my driveway, go down the road, stop at entrance gate waiting for it to open, goes right on through when it does, and takes me to my destination using any and all roads even back roads with no lane markings, all the while I am sitting in the car not knowing what to do with my hands.

Translation: Rivian ADAS is light years behind Tesla FSD even though the latter still is far from pefect.

I had a Tesla and fsd late last year kept doing crazy stuff making me take control back. I ended up with no trust except in very specific scenarios. Well marked lanes and no intersections was pretty much it. I do miss being able to use it off major highways. Even on major highways it will do things that made me disable it. Glad it is working out well for you though!

 

[CrazyOne]

Ah, I see that you are making conclusions about FSD according to a given person, not on the merits. I don't care for [person]; ergo, FSD is a pos. That's not the way it works.

Say what you will, quote whom you will, the data says it all: with Tesla FSD:

• Average crashes on non-highway sections of road: 0.31 times per million miles (compared to the average American's 1.53 times per million miles).
• Airbag-deployed crashes: About every 3.2 million miles (about five times safer than the US average of 600,000 miles per police-reported crash).

#1 thing to do while looking at data is to exclude vehicles without emergency braking. Avg Americans number is useless. You do you.

 

[bigsky]

Then why do you keep comparing them? Rivian is not even trying to do FSD, yet you are constantly disparaging Rivian for being

But that's exactly what it was intended to be. It's called Highway Assist because ... it's intended to assist you on highways. And that's what it does.

Tesla on the other hand calls their system "Full Self-Driving" and even today their implementation doesn't live up to their own hype. Regardless of how good it is, it doesn't fulfill their promise.

Somebody else on here had the audacity to suggest that the Tesla CEO "should be worried" about Rivian ADAS; ergo, this thread started. I merely pointed out that such a statement is blatantly absurd as it indeed compares Rivian against Tesla.

Tesla FSD is not what it was promised to be. Absolutely. I just think that it is a big pile of bs for anybody to therefore suggest that it is no good.

 

[bigsky]

#1 thing to do while looking at data is to exclude vehicles without emergency braking. Avg Americans number is useless. You do you.

Well, average measurements are pretty standard. For decades.

Sponsored