Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Former Presidential Candidate calls for Federal Government to pull FSD from public roads ASAP

This site may earn commission on affiliate links.

Active safety features come standard on all Tesla vehicles made after September 2014 for elevated protection at all times. These features are made possible by our Autopilot hardware and software system and include:

  • Automatic Emergency Braking: Detects cars or obstacles that the car may impact and applies the brakes accordingly
----------------------------------------------------------------------
Let's be honest with ourselves here.
How can Tesla unleash this Beta into the hands of 100K drivers who are lead to believe their car can "navigate city streets" when we are also supposed to believe that "Tesla is not saying that FSD will avoid objects in the road".

The larger problem isn't the current state of FSD, but the gap between Musk/Tesla years of bold statements & the current state as it's rolled out to more "testers".

We are supposed to believe its both the most advanced autonomous car and also that it can't be expected to, in the general case, brake/slow/stop/steer around stationary or moving objects of any kind (humans/animals/crash dummies/vehicles/bollards/trains).

Drivers are going to give the car a little too much leash & benefit of the doubt, expecting it to be much more capable than it is.
This is malpractice, especially in light of public statements of wanting 1M beta testers by end of year.

Post all the anecdotes of "it avoided an obstacle on this one specific build of FSD in this one specific test this one time" but this is not how you would grade a human driver. A human driver that handles only 99 out of 100 objects in the road will get kicked off their insurance every year with all the accident claims they'd be making.
The question is whether people have fewer (and less severe) accidents when they are using FSD Beta. It’s possible they do (when measured fairly and accurately). In the end I would think that’s what matters. And of course that must continue to apply to a wider release!

No one knows what the actual situation is or why. There is no data. Of course, accident rates going down when using FSD Beta may not be due to FSD Beta directly, so it all has to be untangled.

There is reason to think the accident rate could go up, and there is reason to think it could go down.
 
The question is whether people have fewer (and less severe) accidents when they are using FSD Beta. It’s possible they do (when measured fairly and accurately). In the end I would think that’s what matters. And of course that must continue to apply to a wider release!

No one knows what the actual situation is or why. There is no data. Of course, accident rates going down when using FSD Beta may not be due to FSD Beta directly, so it all has to be untangled.

There is reason to think the accident rate could go up, and there is reason to think it could go down.
Firstly - only Tesla has this data, and whatever is shared with regulators has been showed to have omissions / aggregations / obfuscations

Secondly - you have to "test in" to FSD beta and therefore in theory, be a very very very safe driver. So any comparisons re: if FSD beta makes driving more or less safe is going to have to be done very mindfully of the population set. That is - you cannot compare FSD beta accidents per mile against all Tesla miles accidents/mile or all drivers accidents/mile. The cleanest comparison would have to use the same cohort of drivers before & after entry to FSD beta. This is like backtesting a trading strategy, you need to remove things like biases like survivorship/look-ahead.

We have no idea if its more or less safe so let's let them keep rolling the dice on public roads is not a great answer.
 
Logically a same-cohort comparison will probably show an increased safety upon FSD start (due to vigilance) and then a decay of safety as the driver lets their guard down.

One bias is likely that drivers most annoyed by / feeling unsafe from FSD will exit the program/stop engaging it.

Another bias is that Tesla pro-actively boots drivers for various deviances or public posting of negative reviews, etc
This is good in the absolute sense for safety on the road.
From a data comparison perspective though, it puts a hard floor in that will keep the average score up.

A scenario where you are constantly cycling through drivers where you only take your best drivers in, and then boot them if they become bad will make for reasonably good looking data..

Further if they are talking 1M drivers by end of year, thats basically ~25% of all Teslas sold.. meaning they are going to start running out of "really really good" drivers and lower their admittance criteria.
 
That’s really my point… That there a bunch of other assistive technologies out there that cause lots of issues.

MySubaru is far from flawless. The eyesight system does a pretty decent job. It will not save your ass from everything…

Picking out Tesla because their system is more advanced than pretty much anybody else is on the road right now seems a little ridiculous.

The tweet about the if a crashed UFO was in the middle of the road the car would stop was interesting .

In that as huge pile of debris in the road is different than a small object (such as a balloon child being dragged across the street low in front of the bumper).

This is really my issue when it comes to comparing things.

Looking at the Subaru system it would slam on the brakes in an attempt to avoid a collision. It would not completely stop in time…

In theory, the Tesla system will do the same thing. It is a much better job with my old model three and the radar…
 
  • Like
Reactions: Terminator857
That’s really my point… That there a bunch of other assistive technologies out there that cause lots of issues.

MySubaru is far from flawless. The eyesight system does a pretty decent job. It will not save your ass from everything…

Picking out Tesla because their system is more advanced than pretty much anybody else is on the road right now seems a little ridiculous.

The tweet about the if a crashed UFO was in the middle of the road the car would stop was interesting .

In that as huge pile of debris in the road is different than a small object (such as a balloon child being dragged across the street low in front of the bumper).

This is really my issue when it comes to comparing things.

Looking at the Subaru system it would slam on the brakes in an attempt to avoid a collision. It would not completely stop in time…

In theory, the Tesla system will do the same thing. It is a much better job with my old model three and the radar…
One difference: I don’t recall Subarus CEO consistently making wild deceptive claims about his self driving tech…year after year.
 
  • Funny
Reactions: GBMaryland
One difference: I don’t recall Subarus CEO consistently making wild deceptive claims about his self driving tech…year after year.
Deceptive claims vs manslaughter and software malfunctioning seem to be two different subjects.
Agree Elon has been very deceptive and the name "full self driving" is very deceptive. I don't agree that software you supervise needs to be perfect. Although I do agree, the more perfect a supervised software becomes, the more dangerous it becomes, because people will naturally tend to go into a lower power state. The data that FSD is full of issues makes it safer.
 
  • Like
Reactions: GBMaryland
Deceptive claims vs manslaughter and software malfunctioning seem to be two different subjects.
Agree Elon has been very deceptive and the name "full self driving" is very deceptive. I don't agree that software you supervise needs to be perfect. Although I do agree, the more perfect a supervised software becomes, the more dangerous it becomes, because people will naturally tend to go into a lower power state. The data that FSD is full of issues makes it safer.
Agreed. But the naming of the feature as "AUTOPILOT" and "FULL self driving" (and I get it..naming it "partial self driving" would be more truthful, but also wouldnt get as much $12k takers), the constant fake and wild and unobtainable proclamations about what it can/will (but didnt..repeatedly) do? Is what gets a LOT of attention..including from NTSHA, DMV's, Consumer Agencies, Federal Government, and even courts in other countries.
 
  • Like
Reactions: Terminator857

And before the usual group starts the typical knee-jerk defense of Elon and attack Nader (vs focusing on the problematic technology):

Nader's book and other advocacy made major waves and is a major contributor to changing the landscape of standard safety equipment in consumer vehicles today. Things like seatbelts and anti-lock brakes are standard equipment now because of his efforts.

So...lest anyone thinks the federal government doesnt pay attention to Nader..historically speaking, that would not be fact.
In many ways I respect Nader and the great work he has done in the past, but the last few years he has certainly been trying to re-live past glories, and not doing too well at it. Has he ever experienced FSD beta? Does he have evidence its more dangerous than human drivers? If so, make a case, if not, stop making yourself look silly, Nader.
 
In recent years, starting in 2019, Tesla has made no such claim AFAIK (and prior to that you could argue that they were only promises). Elon? Sure. He does not typically talk about driver responsibility for their L2 system.
In 2019 and 2020 on Tesla web site there was text about automated city driving coming by end of year. False in both cases, in other words very false advertising of non existing feature.
 
  • Like
Reactions: 2101Guy
In 2019 and 2020 on Tesla web site there was text about automated city driving coming by end of year. False in both cases, in other words very false advertising of non existing feature.
Yeah, sure. Obviously they are way behind schedule.

In my view for FSD Beta testers post 2019 they have basically fully delivered now - people who bought could not reasonably expect more than what is now available and the (poor) level of functionality for the City Streets feature is well within the bounds of what could be interpreted from what was sold.

But we don’t need to get into that discussion; it is boring and has been done many times here. Plenty of reason for people to be upset, for sure.
 
  • Like
Reactions: Terminator857