carmudgeons.com  

Go Back   carmudgeons.com > Automotive Forums > Car Talk

View Poll Results: Would you ride in the robotaxi
Hell Yes 1 8.33%
Hell No 11 91.67%
Voters: 12. You may not vote on this poll

Reply
 
Thread Tools Display Modes
Old 04-25-2019, 07:58 PM   #21
rumatt
Mugwump
 
rumatt's Avatar
 
Join Date: Oct 2003
Carmudgeonly Ride: E46 330i, Chevy Colorado, Tesla Model 3
Location: NY
Posts: 17,475
Quote:
Originally Posted by clyde View Post
You kill someone or don't; the car kills someone or not. 50/50 either way. Got it.

Some risk, but you have no idea how much, how little, how it compares to if you were in charge, but fuck it. Seems fun.
Your attempt to paint anyone who uses autopilot as a reckless, careless individual is failing.

You don't turn it on then hop on the back seat then take a nap. You sit there wjrh your hands in the wheel watching.

Taking your logic to the extreme, anyone who uses old school cruise control is a danger to society. I mean, they might forget to take over control when traffic slows and ram into the back of someone. Or there could be a bug and it could accelerate uncontrollably. How do you know it won't? Did you review the programming with your own eyes?

Have you ever had a beverage while driving? You could have killed someone when looking down for it and then driving one handed. Have you ever driven while tired? Or maybe too many hours straight without a break? Picked up your phone to answer it? Checked a text even though you know you shouldn't?

There are a lot of ways to kill someone while driving. I don't believe that using autopilot is anywhere near the top of the list of dangerous things I've done while driving. It's silly to even discuss it in that category.

If you put on autopilot then read a book? Yeah that's reckless. But I've seen people reading books while driving cars that *don't* have auto pilot as well. It's the book, not auto pilot.
rumatt is offline   Reply With Quote
Old 04-25-2019, 08:49 PM   #22
clyde
Chief title editor
 
clyde's Avatar
 
Join Date: Oct 2003
Posts: 26,599
Quote:
Originally Posted by rumatt View Post
Your attempt to paint anyone who uses autopilot as a reckless, careless individual is failing.

[snip an amazing amount of false equivalency trying to prove multiple wrongs make a right]


Quote:
Originally Posted by clyde View Post
If I am missing something persuasive that these systems are as safe as humans driving, please show me. I would really like to stop and move on.
Quote:
Originally Posted by rumatt
I don't believe that using autopilot is anywhere near the top of the list of dangerous things I've done while driving.
Why don't you believe that? Why do you believe that it's safe enough to use? I get that it's fun. I don't dispute that. Again, I'm not even suggesting that it's unsafe.

I am just completely unaware of any reasonable reasons to believe that it is safe (or at least using it is roughly as safe as a regular person driving manually) and asking all of you what makes you believe that it's safe to use.

Quote:
Originally Posted by rumatt View Post
Quote:
Originally Posted by clyde View Post
Something convinced you that it's reasonable to turn over your driving responsibilities to your car
It seemed fun.

I also work in AI and love seeing examples of it running action.
Do you see the disconnect between my question and your answer?

Musk and Tesla have been deceptive and misleading about capabilities and safety in hawking Autpilot and then hidden behind the word "beta" when caught. This also doesn't mean that it's not safe, but it doesn't give any reason to believe that it is. Given the man and the company's repeated difficulties in being truthful about...damn near everything, why on earth would you risk your lives and the lives of those around you based on their claims without any kind of independent verification? Because it's fun?

sigh

https://www.thedrive.com/tech/26455/...study-unmasked

https://www.bloomberg.com/news/artic...dog-groups-say

https://abcnews.go.com/Business/tesl...ry?id=61312214

https://www.bloomberg.com/news/artic...outing-model-3

https://www.forbes.com/sites/alanohn.../#63417b5a675a

https://www.reddit.com/r/teslamotors...cing_crap_has/

https://bgr.com/2018/07/11/tesla-mod...safe-to-drive/

https://www.reddit.com/r/EnoughMuskS...into_stanford/
__________________
OH NOES!!!!!1 MY CAR HAS T3H UND3R5T33R5555!!!!!!1oneone!!!!11

Team WTF?!
What are you gonna do?
clyde is offline   Reply With Quote
Old 04-25-2019, 08:58 PM   #23
rumatt
Mugwump
 
rumatt's Avatar
 
Join Date: Oct 2003
Carmudgeonly Ride: E46 330i, Chevy Colorado, Tesla Model 3
Location: NY
Posts: 17,475
Quote:
Originally Posted by clyde View Post
Why do you believe that it's safe enough to use?
Why do you believe cruise control is safe to use?

Because you're still "driving" and you can override it any time you don't like what it is doing. What about this isn't clear and/or convincing?

It's really not anywhere near as complicated as you're trying to make it.

Quote:
snip an amazing amount of false equivalency trying to prove multiple wrongs make a right
Nope.

Was shooting down your implication that anything that could theoretically reduce safety a small amount (or an amount that you cant quantify precisely) is somehow reckless.

Quote:
why on earth would you risk your lives and the lives of those around you
Why do you risk the lives of everyone around you by using cruise control? Who told you it was safe?
Please answer seriously so I can copy paste the answer back to you regarding autopilot.

Last edited by rumatt; 04-25-2019 at 09:16 PM.
rumatt is offline   Reply With Quote
Old 04-25-2019, 10:00 PM   #24
clyde
Chief title editor
 
clyde's Avatar
 
Join Date: Oct 2003
Posts: 26,599
Quote:
Originally Posted by rumatt View Post
Why do you believe cruise control is safe to use?

Because you're still "driving" and you can override it any time you don't like what it is doing. What about this isn't clear and/or convincing?

It's really not anywhere near as complicated as you're trying to make it.


Nope.

Was shooting down your implication that anything that could theoretically reduce safety a small amount (or an amount that you cant quantify precisely) is somehow reckless.


Why do you risk the lives of everyone around you by using cruise control? Who told you it was safe?
Please answer seriously so I can copy paste the answer back to you regarding autopilot.
I asked a question. You don't answer and then demand I answer your counter questions? Nice.

I don't understand what's so "complicated" about telling me what makes you believe it's safe enough to use. I really don't.
__________________
OH NOES!!!!!1 MY CAR HAS T3H UND3R5T33R5555!!!!!!1oneone!!!!11

Team WTF?!
What are you gonna do?
clyde is offline   Reply With Quote
Old 04-26-2019, 07:09 AM   #25
ff
.
 
Join Date: Mar 2004
Posts: 13,514
Quote:
Originally Posted by clyde View Post
it pisses the fuck out of me that others choose to risk my well being, the well being of my family, and the well being of those I care about without our consent.
Me too.

ff is offline   Reply With Quote
Old 04-26-2019, 07:30 AM   #26
JST
195
 
JST's Avatar
 
Join Date: Oct 2003
Posts: 24,593
I mean that guy isn’t driving because his engine is off.
JST is offline   Reply With Quote
Old 04-26-2019, 07:43 AM   #27
JST
195
 
JST's Avatar
 
Join Date: Oct 2003
Posts: 24,593
The question Clyde is posing reflects a pretty fundamental issue at the heart of various legal systems.

There are basically two ways of dealing with new things. Either you prohibit them until people can show they are safe, or you let people do them until they are shown to be dangerous, and require compensation be paid to anyone that is injured.

The American legal system (generally) works on the second principle. Other, code-based systems are (generally) closer to the first. The problem with the first is that it’s hard to prove a negative, so innovation can be harder in a system like that.

Even in America, there are exceptions, of course, for things where we feel like innovation and experimentation are just too risky to proceed without approval. Airplane type approval and drugs are two big ones.

But cars have generally not fallen into that category. The govt sets a minimum safety standard for some things, but beyond that you’re free to innovate. If you fuck up and kill people, they can sue you.

I don’t wholly trust autopilot. I don’t like it and I don’t use it. But the answer to Clyde’s question about how we as a society let people use it lies in the tort system. If it’s defective, if it causes crashes, people will sue and the problem will get fixed. The threat of that (hopefully) means that Tesla has done its due diligence to make sure the system works.
JST is offline   Reply With Quote
Old 04-26-2019, 12:01 PM   #28
robg
Carmudgeon
 
Join Date: Jun 2004
Posts: 3,224
Would you ride in Elon's robotaxi

Quote:
Originally Posted by JST View Post
The question Clyde is posing reflects a pretty fundamental issue at the heart of various legal systems.

There are basically two ways of dealing with new things. Either you prohibit them until people can show they are safe, or you let people do them until they are shown to be dangerous, and require compensation be paid to anyone that is injured.

The American legal system (generally) works on the second principle. Other, code-based systems are (generally) closer to the first. The problem with the first is that it’s hard to prove a negative, so innovation can be harder in a system like that.

Even in America, there are exceptions, of course, for things where we feel like innovation and experimentation are just too risky to proceed without approval. Airplane type approval and drugs are two big ones.

But cars have generally not fallen into that category. The govt sets a minimum safety standard for some things, but beyond that you’re free to innovate. If you fuck up and kill people, they can sue you.

I don’t wholly trust autopilot. I don’t like it and I don’t use it. But the answer to Clyde’s question about how we as a society let people use it lies in the tort system. If it’s defective, if it causes crashes, people will sue and the problem will get fixed. The threat of that (hopefully) means that Tesla has done its due diligence to make sure the system works.


Well said. There was a recent article in the times about this too (about the 737 max but more broadly about how the US differs in its approach to regulation from Europe:
https://www.nytimes.com/2019/03/15/o...on-boeing.html

Autonomous vehicles (and AI in general) is definitely leading us into uncharted territory.

Humans are also really bad at judging risk.
robg is online now   Reply With Quote
Old 04-26-2019, 12:55 PM   #29
JST
195
 
JST's Avatar
 
Join Date: Oct 2003
Posts: 24,593
Quote:
Originally Posted by robg View Post
Well said. There was a recent article in the times about this too (about the 737 max but more broadly about how the US differs in its approach to regulation from Europe:
https://www.nytimes.com/2019/03/15/o...on-boeing.html

Autonomous vehicles (and AI in general) is definitely leading us into uncharted territory.

Humans are also really bad at judging risk.
Yes, that's an interesting article, though I think the point sits somewhat awkwardly in the 737 Max discussion, since aircraft type approval is one of the best examples of "ask questions, shoot later" that exists in the American legal system.

The precautionary principle is one reason why the EU has been so aggressive in banning the use of GMOs, despite the complete lack of any evidence of harm. The problem with it is precisely the one the NYT identifies--without data, if you have to prove a negative you're really vulnerable to magical thinking. "What if" becomes a roadblock that you can't get around.
JST is offline   Reply With Quote
Old 04-26-2019, 01:46 PM   #30
robg
Carmudgeon
 
Join Date: Jun 2004
Posts: 3,224
Would you ride in Elon's robotaxi

Quote:
Originally Posted by JST View Post
Yes, that's an interesting article, though I think the point sits somewhat awkwardly in the 737 Max discussion, since aircraft type approval is one of the best examples of "ask questions, shoot later" that exists in the American legal system.



The precautionary principle is one reason why the EU has been so aggressive in banning the use of GMOs, despite the complete lack of any evidence of harm. The problem with it is precisely the one the NYT identifies--without data, if you have to prove a negative you're really vulnerable to magical thinking. "What if" becomes a roadblock that you can't get around.


For sure. Defining strong regulations upfront sounds good in theory but is very difficult in practice when it comes to new technology.

The 737 max issue is an interesting example, though, in that the problem involved software and it sounds like the FAA feels increasingly out of their depth when it comes to evaluating the safety of new software features (leading to the “self certification”). Unintended consequences arising from software are going to be an increasing problem in general. That being said, I do wonder wtf Boeing was thinking when it implemented that software with no sanity check/ redundancy. That’s a basic principle that’s been part of aerospace engineering for a long time.

I believe there are already regulations in the US that require redundancy when it comes to autonomous vehicles atkeast.
robg is online now   Reply With Quote
Reply

Bookmarks

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -4. The time now is 01:42 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Forums © 2003-2008, 'Mudgeon Enterprises - Site hosting by AYN & Associates, LLC