Self-Driving Vehicles
10-25-18, 03:28 PM
#811
There is a wide range of topics like this.
Should the owner fill out a survey that helps refine the vehicles "morals"? Not a proposal, just thought exercise.
Will all these cars go the speed limit. Should owners be able to tell it to go faster? By how much?
Should the vehicle be assertive or cooperative at zipper merges. Does that change based on the model of vehicle that wants to merge?
Going to be exciting.
Should the owner fill out a survey that helps refine the vehicles "morals"? Not a proposal, just thought exercise.
Will all these cars go the speed limit. Should owners be able to tell it to go faster? By how much?
Should the vehicle be assertive or cooperative at zipper merges. Does that change based on the model of vehicle that wants to merge?
Going to be exciting.
10-25-18, 07:44 PM
#812
Yes. I do understand that the car, if programmed this way, would favor the younger person over the older person. I know its not going to missile towards the elder folk. The problem is - if there is a situation where the AI car needs to hit someone, it ill-moral way to think that its "okay" to hit the elder person rather than the younger person. At the end of the day - these are someone's kids, parents, or grand parents.
Since you are the driver, you should hit a pillar and save other people first because you are the one liable behind the wheel.
There is someone in the comment section of the article that made a good point: I believe, the person said and I loosely quote, "the AI self-driving car should protect the people outside of the car before inside of the car because the people inside of the car agreed to letting the car drive them and agree to the risk of the terms of a self-driving car while the people outside of the car never agreed to such terms."
Since you are the driver, you should hit a pillar and save other people first because you are the one liable behind the wheel.
There is someone in the comment section of the article that made a good point: I believe, the person said and I loosely quote, "the AI self-driving car should protect the people outside of the car before inside of the car because the people inside of the car agreed to letting the car drive them and agree to the risk of the terms of a self-driving car while the people outside of the car never agreed to such terms."
How about this scenario where there is no AI/computer involved. The car you're driving car is out of control, and you only have 3 options -- there is no options like flying into the sky or sacrificing yourself: you can turn right to hit a young person, turn left to hit an elderly, or do nothing and go straight to hit both. What would you do?
10-25-18, 09:50 PM
#813
Lexus Test Driver
I believe taking the driver out instead of anyone outside is also a consideration for the AI. The ethical question though is when the AI doesn't have that choice.
How about this scenario where there is no AI/computer involved. The car you're driving car is out of control, and you only have 3 options -- there is no options like flying into the sky or sacrificing yourself: you can turn right to hit a young person, turn left to hit an elderly, or do nothing and go straight to hit both. What would you do?
How about this scenario where there is no AI/computer involved. The car you're driving car is out of control, and you only have 3 options -- there is no options like flying into the sky or sacrificing yourself: you can turn right to hit a young person, turn left to hit an elderly, or do nothing and go straight to hit both. What would you do?
I believe its morally flawed to program an AI to say "in a situation with no other options - hit the elderly person."
12-01-18, 10:44 AM
#814
Self-driving cars
Good item in today's WSJ by Holman Jenkins regarding the hype over self-driving cars. Get a copy of the paper (or suscribe to get an online version) to see the whole thing. I've always been highly skeptical of the practicality of the concept, and think this article lays out the basis for that view.
I've excerpted parts of the article here:
------------------------------
This column got interested in self-driving cars back in 2008, for two reasons. One was the success of the Pentagon’s “urban challenge,” won by a team from Carnegie Mellon that created a self-driving Chevy Tahoe and navigated an obstacle course while obeying California traffic laws...
Self-driving cars became interesting again for a whole new reason during the Obama years. Suddenly the media was filled with giddy prognostications that such cars were just around the corner when they clearly weren’t...
Audi, at the 2013 Las Vegas Consumer Electronics Show, unveiled a self-driving vehicle, supposedly soon to be available to the public, which would handle highway driving until it didn’t, at which point a passenger would be expected to take over within seconds. Elon Musk seemingly promised every year that a completely capable self-driving car was just a year away...
Toyota, at the same time, was routinely ignored for saying the new technology would compensate for a driver’s errors long before it was ready to accommodate his desire to be doing something else. Now most sane people agree...
It was clear even then what was really going on. Public and press were head over heels for autonomous vehicles. An irresistible, almost mandatory marketing opportunity was born. Google, for the tens of millions it has spent on self-driving runabouts, reaped billions in publicity of the kind that can’t be bought with any advertising budget...
Uber had a perfectly good thing going with ride-sharing, but felt the need to boost its gee-whiz quotient by becoming an autonomous-vehicle pioneer. Apple managed to create a still-vibrating buzz for itself in early 2015 simply by placing a few ads for automotive engineers and leaking word that it was working on a car...
Traditional car companies like Ford and General Motors , meanwhile, scrambled to brand themselves with autonomous-mobility credibility in hopes of obtaining a sliver of the stock-market favor accruing to Tesla, Apple and Google...
Lately, with signals from its industry sources, the press has finally decided to acknowledge that the autonomous car isn’t just around the corner. It can’t handle inclement weather. It can’t reliably tell a plastic bag blowing across the road from a child on a bicycle and won’t be able to soon...
Axios, that breathless compendium, now admits after launching its own autonomous vehicle newsletter that true self-driving cars are still “10, 20, 50 years” away...
Toyota was right. For the foreseeable future, autonomous features will mainly serve to stop us from screwing up. And yet what’s being cooked up today may prove more transformative in the long run than even the hype-mongers predicted...
Take the machine vision, 3-D mapping and ubiquitous low-latency broadband networks needed to make driverless cars possible. These technologies will also make many trips superfluous...
The signs are already visible. On average, each of us drives less per year than we did in 2004. More Americans work at home, watch Netflix instead of venturing to the movies, and rely on Peapod and Amazon to save them trips to the grocer. Like many technological forecasts, these visions may be slightly off-kilter from the future that actually unfolds...
------------------------------------
I DO think that we may very well see a partial implementation where parts of the Interstate system are upgraded to a high standard and upon which cars conforming to that standard can be totally autonomous while on such a highway. But that standard will be set very high because no loss of life will ever be tolerated, and that's asking a lot.
I've excerpted parts of the article here:
------------------------------
This column got interested in self-driving cars back in 2008, for two reasons. One was the success of the Pentagon’s “urban challenge,” won by a team from Carnegie Mellon that created a self-driving Chevy Tahoe and navigated an obstacle course while obeying California traffic laws...
Self-driving cars became interesting again for a whole new reason during the Obama years. Suddenly the media was filled with giddy prognostications that such cars were just around the corner when they clearly weren’t...
Audi, at the 2013 Las Vegas Consumer Electronics Show, unveiled a self-driving vehicle, supposedly soon to be available to the public, which would handle highway driving until it didn’t, at which point a passenger would be expected to take over within seconds. Elon Musk seemingly promised every year that a completely capable self-driving car was just a year away...
Toyota, at the same time, was routinely ignored for saying the new technology would compensate for a driver’s errors long before it was ready to accommodate his desire to be doing something else. Now most sane people agree...
It was clear even then what was really going on. Public and press were head over heels for autonomous vehicles. An irresistible, almost mandatory marketing opportunity was born. Google, for the tens of millions it has spent on self-driving runabouts, reaped billions in publicity of the kind that can’t be bought with any advertising budget...
Uber had a perfectly good thing going with ride-sharing, but felt the need to boost its gee-whiz quotient by becoming an autonomous-vehicle pioneer. Apple managed to create a still-vibrating buzz for itself in early 2015 simply by placing a few ads for automotive engineers and leaking word that it was working on a car...
Traditional car companies like Ford and General Motors , meanwhile, scrambled to brand themselves with autonomous-mobility credibility in hopes of obtaining a sliver of the stock-market favor accruing to Tesla, Apple and Google...
Lately, with signals from its industry sources, the press has finally decided to acknowledge that the autonomous car isn’t just around the corner. It can’t handle inclement weather. It can’t reliably tell a plastic bag blowing across the road from a child on a bicycle and won’t be able to soon...
Axios, that breathless compendium, now admits after launching its own autonomous vehicle newsletter that true self-driving cars are still “10, 20, 50 years” away...
Toyota was right. For the foreseeable future, autonomous features will mainly serve to stop us from screwing up. And yet what’s being cooked up today may prove more transformative in the long run than even the hype-mongers predicted...
Take the machine vision, 3-D mapping and ubiquitous low-latency broadband networks needed to make driverless cars possible. These technologies will also make many trips superfluous...
The signs are already visible. On average, each of us drives less per year than we did in 2004. More Americans work at home, watch Netflix instead of venturing to the movies, and rely on Peapod and Amazon to save them trips to the grocer. Like many technological forecasts, these visions may be slightly off-kilter from the future that actually unfolds...
------------------------------------
I DO think that we may very well see a partial implementation where parts of the Interstate system are upgraded to a high standard and upon which cars conforming to that standard can be totally autonomous while on such a highway. But that standard will be set very high because no loss of life will ever be tolerated, and that's asking a lot.
Last edited by riredale; 12-01-18 at 10:49 AM.
12-01-18, 12:15 PM
#815
All good points, and i mostly agree. But we have to start somewhere. Windows 3.1 was horrific compared to windows 10, but the pc gui had to start somewhere.
I really like how GM is doing their super cruise - taking the long boring driving on interstates off of our hands. I don't mind driving around town, but having the car drive 20 of my 30 miles on highway is a very nice thing to have.
I really like how GM is doing their super cruise - taking the long boring driving on interstates off of our hands. I don't mind driving around town, but having the car drive 20 of my 30 miles on highway is a very nice thing to have.
12-01-18, 02:21 PM
#816
The self driving cars are coming only a matter of time many here are against it and I understand but the tech is here and will get better. Will only be a matter of time. If every car is aware of the other car(s) then collisions will drastically drop. AI on the whole is somewhat scary when you look at it. AI logic will determine that we (humans) need to be protected from ourselves.
12-01-18, 09:16 PM
#817
Self driving cars are not going to be a thing for the public. The world is simply too complicated and the human mind is too complicated to replicate.
The best you can do is Super Cruise, which is autonomous on the highway only, or have Autonomous vehicles on private streets.
The best you can do is Super Cruise, which is autonomous on the highway only, or have Autonomous vehicles on private streets.
12-02-18, 03:06 AM
#818
Lexus Fanatic
Self driving cars are not going to be a thing for the public. The world is simply too complicated and the human mind is too complicated to replicate.
The best you can do is Super Cruise, which is autonomous on the highway only, or have Autonomous vehicles on private streets.
The best you can do is Super Cruise, which is autonomous on the highway only, or have Autonomous vehicles on private streets.
12-02-18, 03:19 AM
#819
Can you elaborate a bit more on this part?
12-02-18, 04:24 AM
#820
It is complex indeed but does that mean they should just stop where they are now no, they will continue to try and improve that's what innovators do. I go to Silicon Valley quarterly some of the tech I'm seeing out there is amazing especially on the AI side of the house.
12-02-18, 10:22 AM
#821
Lexus Fanatic
Well, I'm not an electrical engineer, but part of it is obvious, even to those without an electronics degree. There are innumerable different possibilities a vehicle can actually encounter on the road that are not programmed (and cannot be programmed) into the data base. Speed limits are often (temporarily) lowered for construction, utility crews, or road maintenance. Cones or other obstacles can temporarily block off certain lanes. Police and School-Safety Officers often halt and direct traffic in the morning and afternoon as private schools start and let out, with parents trying to get in and out to drop off or pick up their kids. Long funeral-processions of vehicles may have the right away over normal red/green traffic lights or stop signs. A street may have a "DO NOT ENTER" sign or barricade at certain periods of the day, which restricts or forbids access....this is very common on neighborhood streets that are sometimes used for cut-through traffic. Any number of signs could restrict motion in a private or mall parking lots. Military bases often have very strict rules about who can come in the gates without proper documentation or passes. I could go on and on, but the list is endless.
Last edited by mmarshall; 12-02-18 at 04:00 PM.
12-02-18, 01:35 PM
#822
Pole Position
Sorry, I can't possibly disagree more with people who say how "complex" the world and human mind is... You are aware that AI is successfully being used to read through medical images and can diagnose abnormalities and cancer as accurate as the top tier of radiologists, correct? Said another way, computers are already equally as accurate as doctors who have spent decades in school, in training and actual practice and are in the top tier of their trade for accuracy. Teaching a computer how to operate a vehicle is no more difficult than teaching a computer how to become a physician. Radiologists are training the computers to make better diagnoses, just as human drivers and engineers are training the cars to make the right driving decisions.
Having speed limits temporarily lowered for construction is incredibly easy to overcome - the car will see speed limit signs with cameras and adjust accordingly. On top of that, cars will be connected into a live cloud that will get live-updated information on road closures, traffic, etc. (think of Waze) - so the car will already know and make adjustments.
Having speed limits temporarily lowered for construction is incredibly easy to overcome - the car will see speed limit signs with cameras and adjust accordingly. On top of that, cars will be connected into a live cloud that will get live-updated information on road closures, traffic, etc. (think of Waze) - so the car will already know and make adjustments.
12-02-18, 01:49 PM
#823
Sorry, I can't possibly disagree more with people who say how "complex" the world and human mind is... You are aware that AI is successfully being used to read through medical images and can diagnose abnormalities and cancer as accurate as the top tier of radiologists, correct? Said another way, computers are already equally as accurate as doctors who have spent decades in school, in training and actual practice and are in the top tier of their trade for accuracy. Teaching a computer how to operate a vehicle is no more difficult than teaching a computer how to become a physician. Radiologists are training the computers to make better diagnoses, just as human drivers and engineers are training the cars to make the right driving decisions.
Having speed limits temporarily lowered for construction is incredibly easy to overcome - the car will see speed limit signs with cameras and adjust accordingly. On top of that, cars will be connected into a live cloud that will get live-updated information on road closures, traffic, etc. (think of Waze) - so the car will already know and make adjustments.
Having speed limits temporarily lowered for construction is incredibly easy to overcome - the car will see speed limit signs with cameras and adjust accordingly. On top of that, cars will be connected into a live cloud that will get live-updated information on road closures, traffic, etc. (think of Waze) - so the car will already know and make adjustments.
12-02-18, 03:23 PM
#824
Marshall, I agree with all of those things. The sum of those problems is exactly why full autonomous will never happen for the larger public. Im glad you and I get along because you see things objectively, as things should be seen.
And im just loling at some peoples' knowledge on AI. Yall do know that the best AI systems in place are in video games, right? And they can't even get that right. And these developers are some of the brightest folks you can hire. To add insult to injury, video game AI have been in development for the past 20 years. Autonomous vehicles? 6ish.
And im just loling at some peoples' knowledge on AI. Yall do know that the best AI systems in place are in video games, right? And they can't even get that right. And these developers are some of the brightest folks you can hire. To add insult to injury, video game AI have been in development for the past 20 years. Autonomous vehicles? 6ish.
Last edited by theory816; 12-02-18 at 03:36 PM.
12-02-18, 04:29 PM
#825
Lexus Fanatic
The sum of those problems is exactly why full autonomous will never happen for the larger public. Im glad you and I get along because you see things objectively, as things should be seen.
And im just loling at some peoples' knowledge on AI. Yall do know that the best AI systems in place are in video games, right? And they can't even get that right. And these developers are some of the brightest folks you can hire. To add insult to injury, video game AI have been in development for the past 20 years. Autonomous vehicles? 6ish.
And im just loling at some peoples' knowledge on AI. Yall do know that the best AI systems in place are in video games, right? And they can't even get that right. And these developers are some of the brightest folks you can hire. To add insult to injury, video game AI have been in development for the past 20 years. Autonomous vehicles? 6ish.
BTW, Och and several others also agree with you.
Last edited by mmarshall; 12-02-18 at 04:39 PM.