You realize this is similar to how radical groups get formed, right? ISIS 'gives' education to the poor and desperate in the nation and, oh, it turns out they actually start to understand the world in the way ISIS wants them to and it starts to makes sense that they should strap a bomb to their body. Not that I am suggesting Facebook is a terrorist group but they are a group with their own interests, thus making it a dangerous enough power (to disseminate propaganda no matter how subtle) that should be closely monitored. A massive group of people understanding the world according to Facebook is not good for anyone.
Exactly.. People make a choice how they drive, so if AI isn't satisfying that choice people won't adopt it. But satisfy the choice and they will be more unsafe than the human in the first place.
I can't wait to get caught behind a double-wide tractor driving 30mph down the shoulder/lane and have the AI too afraid/unable to pass in the oncoming lane.
Actually the number of wrong turns down a one way that do not cause an accident are not relevant, because in every one of those cases a human has dealt with it in the correct way once realizing the mistake (and thus not caused an accident). The fact that the AI must be shut down once such a serious error is made is basically an admittance that there is no faith in the AI to deal with the error and drive out of it safely so you have to count any such mistake made by AI as an accident causing event. All this goes to show is that AI is currently pitiful compared to humans. This whole conversation is beside the point, since there is apparently no AI at all and the car is just blindly following a virtual line drawn on a map. With such serious and glaring single points of failure in the automated car's ability to navigate, any conversation even entertaining the idea that these cars are as safe as humans seems laughable.
How many of these Uber cars are on the road? Twenty? And one million human drivers. One must only stand on a one way road for ten minutes or so and see whether anyone goes the wrong way in that ten minutes to determine whether humans are doing a better job or not.
I've noticed that there are a lot of people on Slashdot who read too much science fiction. Instead of fixing local problems, people get wrapped up in tech porn. It's almost like they escape problems here by thinking of grandiose plans in space and imagine that they will solve anything. If your dog covers your back yard in shit, you clean it up, not buy a new house.
Bingo. As I keep saying, autonomous cars are only fair for up to and including the average driver. The problem is there is no average driver, and half the drivers are above average. Many people who would never get in an accident otherwise are playing Russian roulette with autonomy. Also, even the worst driver always has the option to slow down.
I always wonder what will happen when the city decides to turn a two way street into a one way street.. will they be obligated to coordinate every mapping service out there to the exact second they change the usage of the street? Surely there will be days delay before everything gets updated and it will be unavoidable.
Slashdot Mirror
You realize this is similar to how radical groups get formed, right? ISIS 'gives' education to the poor and desperate in the nation and, oh, it turns out they actually start to understand the world in the way ISIS wants them to and it starts to makes sense that they should strap a bomb to their body. Not that I am suggesting Facebook is a terrorist group but they are a group with their own interests, thus making it a dangerous enough power (to disseminate propaganda no matter how subtle) that should be closely monitored. A massive group of people understanding the world according to Facebook is not good for anyone.
Or at least have it decided/overseen by an independent board with members chosen in a fair way.
Exactly.. People make a choice how they drive, so if AI isn't satisfying that choice people won't adopt it. But satisfy the choice and they will be more unsafe than the human in the first place.
My house contains the gate to the other side. Send me $100,000,000 and wait two weeks and I'll let you see it.
Sennheiser doesn't seem to make bluetooth earbuds? I can't sleep in full headphones.
I can't wait to get caught behind a double-wide tractor driving 30mph down the shoulder/lane and have the AI too afraid/unable to pass in the oncoming lane.
Should we all welcome our autonomous vehicle overlords now?
As soon you figure out how to buy every American a $100K vehicle and also have automation drive safely on its own.
The fact of the matter is, Tesla depends on their customers being shit scared of the technology or they wouldn't log the 'safety' numbers they do.
If all humans drove that speed, they would easily be safer than a google car.
I keep my headphones on while I sleep, and expect them to not only last the night but still be fresh in the morning.
But if it were your choice to drive manually and live or die by machine.. what would it be?
I can't wait for millions of lost/broken lightning port dongles and millions of wireless headphones with dead batteries to be added to the mix.
Kevin Costner may be fine with people thinking it was Mel Gibson.
If that were the case I would have said 'as much as anyone going to Mars'.
Actually the number of wrong turns down a one way that do not cause an accident are not relevant, because in every one of those cases a human has dealt with it in the correct way once realizing the mistake (and thus not caused an accident). The fact that the AI must be shut down once such a serious error is made is basically an admittance that there is no faith in the AI to deal with the error and drive out of it safely so you have to count any such mistake made by AI as an accident causing event. All this goes to show is that AI is currently pitiful compared to humans. This whole conversation is beside the point, since there is apparently no AI at all and the car is just blindly following a virtual line drawn on a map. With such serious and glaring single points of failure in the automated car's ability to navigate, any conversation even entertaining the idea that these cars are as safe as humans seems laughable.
You still have to frame it in terms of the total number of drivers. 760 is a lot out of 1000 drivers but irrelevant out of millions.
The fact is, hundreds of thousands of people are navigating one way streets correctly every day.
A lot more than anyone going to Mars.
How many of these Uber cars are on the road? Twenty? And one million human drivers. One must only stand on a one way road for ten minutes or so and see whether anyone goes the wrong way in that ten minutes to determine whether humans are doing a better job or not.
I've noticed that there are a lot of people on Slashdot who read too much science fiction. Instead of fixing local problems, people get wrapped up in tech porn. It's almost like they escape problems here by thinking of grandiose plans in space and imagine that they will solve anything. If your dog covers your back yard in shit, you clean it up, not buy a new house.
So then they aren't ready for public use. They should still be in test towns where they won't hurt anyone.
Bingo. As I keep saying, autonomous cars are only fair for up to and including the average driver. The problem is there is no average driver, and half the drivers are above average. Many people who would never get in an accident otherwise are playing Russian roulette with autonomy. Also, even the worst driver always has the option to slow down.
If they're on public roads it's not a test. Real people may die.
I always wonder what will happen when the city decides to turn a two way street into a one way street.. will they be obligated to coordinate every mapping service out there to the exact second they change the usage of the street? Surely there will be days delay before everything gets updated and it will be unavoidable.
So if there was an accident at a stop sign you would blame the driver that did stop?