Problems with the concept of "Auto Pilot" - Page 3 - Jaguar I-Pace EV400 Forum
 17Likes
Reply
 
LinkBack Thread Tools
post #21 of 32 (permalink) Old 05-18-2019, 04:26 PM
Senior Member
 
Mister Dave's Avatar
 
Join Date: Oct 2018
Location: Illinois
Posts: 521
Quote:
Originally Posted by phil View Post
I have trouble with it too, but it's gonna happen. Really, it has already happened - it was 1997 when IBM's Deep Blue 'anticipated the human behavior' of Garry Kasparov, world champion of chess, and checkmated him. Google's DeepMind appears to be leaps and bounds beyond Deep Blue's cababilities, only 22 years later. Granted, driving (moreso than chess) is the sort of activity where humans tend to do better than computers. But really, it's just a matter of degree, and a matter of time.
A fair point and I'm okay with being wrong, but....

Quote:
Originally Posted by phil View Post
And when you get right down to it, how unpredictable is human behavior anyway? Perhaps less than you think...
Chess has very well defined parameters. I've beaten computers at chess many times, but never played Big Blue (obviously). The trick is to play and unbalanced game. The computer "AI" hasn't been programmed to deal with that.

Humans are very unbalanced in their logic, myself included.
Mister Dave is offline  
post #22 of 32 (permalink) Old 05-18-2019, 10:21 PM
Senior Member
 
McRat's Avatar
 
Join Date: May 2018
Location: Norco, California, USA
Posts: 1,352
Trivia: The DARPA $2 Million Urban Challenge was in 2007 with 6 cars completing the course which included traffic control and other cars. This was 60 miles of driving city streets without error.

My bet is lawyers will crush the technology, or at least delay it a decade after it's viable.

Jaguar i-Pace FE Photon Red 20" wheels, "Leaper"
Two Chevrolet Volts in service
24.2 kW x 480v 3ph solar array self-installed.
McRat is offline  
post #23 of 32 (permalink) Old 05-19-2019, 12:04 AM
Senior Member
 
jsimon7777's Avatar
 
Join Date: Jun 2018
Location: Castro Valley, CA
Posts: 773
Quote:
Originally Posted by phil View Post
I have trouble with it too, but it's gonna happen. Really, it has already happened - it was 1997 when IBM's Deep Blue 'anticipated the human behavior' of Garry Kasparov, world champion of chess, and checkmated him. Google's DeepMind appears to be leaps and bounds beyond Deep Blue's cababilities, only 22 years later. Granted, driving (moreso than chess) is the sort of activity where humans tend to do better than computers. But really, it's just a matter of degree, and a matter of time. And when you get right down to it, how unpredictable is human behavior anyway? Perhaps less than you think...
You watch the new version of Westworld?
jsimon7777 is online now  
Sponsored Links
Advertisement
 
post #24 of 32 (permalink) Old 05-21-2019, 12:44 AM
Senior Member
 
Join Date: Nov 2018
Location: Scottsdale, AZ
Posts: 380
This is why we can't have nice things. The law needs to recognize the point that the autopilots - are likely to SAVE more lives than they lose. Why is there an expectation for it to be "foolproof"? It just needs to be better than the (totally fallible) human operator MOST of the time. I think that the company is liable, if there's a known problem that they do not immediately acknowledge and move to correct and improve the system. But, they shouldn't be liable for huge damages just because a self-driving system isn't 100% successful - provided that they can show a lower accident rate than human drivers per mile. In such cases, the "damages" should be covered by the owner's insurance to reasonably spread the risk. If we don't make the laws work that way, we'll never get serious automation that does improve the world.

Last edited by WattAJag; 05-21-2019 at 12:46 AM.
WattAJag is online now  
post #25 of 32 (permalink) Old 05-21-2019, 10:33 AM
Senior Member
 
Join Date: Jan 2019
Location: SF Bay Area
Posts: 138
Garage
Quote:
Originally Posted by WattAJag View Post
This is why we can't have nice things. The law needs to recognize the point that the autopilots - are likely to SAVE more lives than they lose. Why is there an expectation for it to be "foolproof"? It just needs to be better than the (totally fallible) human operator MOST of the time. I think that the company is liable, if there's a known problem that they do not immediately acknowledge and move to correct and improve the system. But, they shouldn't be liable for huge damages just because a self-driving system isn't 100% successful - provided that they can show a lower accident rate than human drivers per mile. In such cases, the "damages" should be covered by the owner's insurance to reasonably spread the risk. If we don't make the laws work that way, we'll never get serious automation that does improve the world.
But why should the owner have to have insurance at all for when autopilot is operational? In at-fault states, the driver's insurance is the one that has to cover an accident, not the owner's. So if I lend my car to a friend, and that friend hits another car, their insurance has to take the hit, not mine. If the country went to no-fault insurance that would be a different story. But the transition period for that will suck.
epirali likes this.
CyberSlug is online now  
post #26 of 32 (permalink) Old 05-21-2019, 12:41 PM
Senior Member
 
Mister Dave's Avatar
 
Join Date: Oct 2018
Location: Illinois
Posts: 521
We have serious automation that does improve the world. I helped to build some of it. Why does it have to be on the roadway?

Does anyone seriously think the robocar will be remotely affordable for any but the most wealthy of us? Then what's the point?

Taxis. If it isn't earning cash 24/7 it isn't going to sell.

Speaking of 'ride sharing', Mary Barra is pulling her Maven service from selected cities.

General Motors Pulls Maven Out Of Eight Different US Cities
Mister Dave is offline  
post #27 of 32 (permalink) Old 05-21-2019, 12:55 PM Thread Starter
Senior Member
 
Join Date: Oct 2018
Posts: 823
I think at some point if the entire auto system was controlled by a centralized system with no human drivers we could move a lot more people on existing roadways (both at faster speeds, and god much better flow efficiency as everyone in US seems to have forgotten how to drive). But that is an ideal, and I do not believe we have well developed enough technologies yet to do it.

If every car fully reported their location, and there were NO UNKOWN OBSTACLES allowed on the roadways, then this would work. Until then...
epirali is online now  
post #28 of 32 (permalink) Old 05-21-2019, 03:33 PM
Senior Member
 
Join Date: Jan 2018
Location: San Jose, CA
Posts: 125
Not sure if this was posted, but the EU has regulators with actual balls and essentially AP will be at a standstill until Tesla has a fully functioning autonomous system--which, is to say, never IMO

https://www.engadget.com/2019/05/17/...opilot-europe/
Mister Dave and epirali like this.

2017 Jaguar F-Pace S
Ordered: 2019 Jaguar I-Pace HSE First Edition Corris Grey
Pseudonym is offline  
post #29 of 32 (permalink) Old 05-21-2019, 11:18 PM
Senior Member
 
Mister Dave's Avatar
 
Join Date: Oct 2018
Location: Illinois
Posts: 521
Quote:
Originally Posted by epirali View Post
I think at some point if the entire auto system was controlled by a centralized system with no human drivers we could move a lot more people on existing roadways (both at faster speeds, and god much better flow efficiency as everyone in US seems to have forgotten how to drive). But that is an ideal, and I do not believe we have well developed enough technologies yet to do it.

If every car fully reported their location, and there were NO UNKOWN OBSTACLES allowed on the roadways, then this would work. Until then...
Now imagine you live in Montana.........
Mister Dave is offline  
post #30 of 32 (permalink) Old 05-22-2019, 01:49 PM Thread Starter
Senior Member
 
Join Date: Oct 2018
Posts: 823
Quote:
Originally Posted by Mister Dave View Post
Now imagine you live in Montana.........
Do I have to? It is very pretty but I REQUIRE high population density to survive...

But this is obvious but interesting:

Consumer Reports: Latest Autopilot “far less competent than a human”

https://arstechnica.com/cars/2019/05...-than-a-human/
epirali is online now  
Reply

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page



Similar Threads
Thread Thread Starter Forum Replies Last Post
Jaguar I-Pace Debuts as an All-Electric Crossover Concept That Actually Looks Cool Administrator 2018+ Jaguar I-Pace EV 400 News 1 03-06-2018 09:45 AM
Jaguar I-Pace Took Home 2017's Most Significant Concept Vehicle Award Jaguar I-Pace News 2018+ Jaguar I-Pace EV 400 News 4 08-09-2017 10:59 AM
Most Significant Concept Vehicle of 2017 PinkJaguar 2018+ Jaguar I-Pace EV 400 General Discussion 0 08-01-2017 12:46 PM

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
Trackbacks are On
Pingbacks are On
Refbacks are On

 
For the best viewing experience please update your browser to Google Chrome