Posted on 02/09/2016 8:52:56 AM PST by PreciousLiberty
Elon Musk's Tesla recently became the latest big shot company to enter the self-driving car sweepstakes. Mr. Musk recently announced the hiring of software architecture veteran Jim Keller, who previously had played key roles at Apple and AMD, to lead its Autopilot Engineering team. Teslas move follows the recently announced partnership between General Motors and Lyft, in which the automaker is investing $500 million in the ridesharing company as part of a joint venture to develop self-driving cars.
And of course Google, Uber, Mercedes-Benz, Nissan, Apple, Audi, Bosch and Delphi Automotive (the big auto parts manufacturer) all have their own much-hyped development programs for autonomous vehicles in full swing. Forget cures for cancer, climate change or world peace, the media has made it clear that self-driving cars will be the Next Great Step in civilizations drive toward magnificence.
It's time to hit the brakes for a reality check.
Despite how much Uber CEO Travis Kalanick likes to crow about our "driverless future," outside of The Jetsons this one is...not...happening...soon. Besides the remaining technological challenges, the liability and regulatory issues involved in letting a 3,000-pound death machine steer itself with no human at the controls are huge.
(Excerpt) Read more at observer.com ...
Only until the A.I. that drives the car decides it no longer takes you where you want to go, and decides to take you where it or someone else wants you to go.
Sure you want to give up your freedom like that?
it is all about “equalization of opportunity” under the law. (/s)
Personally, I love to drive. I have a performance car with a manual transmission and it's sort of my daily amusement as well as transportation. Others don't feel that way, and more power to them (and more road for me). When it's time for the transition, i.e. when driving isn't fun anymore (my Mom reached that stage when she hit 80 - night driving was just too scary) I'll stop and stick to reminiscing.
Like the time I drove 20 miles in the left lane with my blinker going...
No, I made the point that even with AI stupid behavior like that which I pointed out can cause accidents because it's not possible to pre-program intelligence to react to situations, timing, etc.. that it's not aware of.
Neither can humans, the issue is who is superior since neither option is perfect.
I'll take my adaptability and ability to react over a pre-programmed machine or learning algorithm any day. :-)
The driverless car will also detect the idiot, and will detect his movement and predict the intercept much more quickly and precisely than you are and also will be able to react much, much faster than you can. I'll also add, if it was a driverless car moving the idiot, it wouldn't pull into your right of way either. You're making a great case for getting humans off the wheel!
You're missing the point: only in as much as the AI knows, has been programmed to know or as good as the learning algorithm it's been programmed with is capable of.
That's the rub.
And as I pointed out to another poster above, AI is only "effective" and "practical" for humans to cede decision making authority for us until the AI starts deciding for itself what is in its, or our best interests.
Then, all bets are off.
That's not a road I'm willing to go down (no pun intended, try getting AI to do that! )
And flight is a physical impossibility right?
Never say never :) Every single car maker has autonomous vehicle programs (except Porsche, so they say). They have working vehicles that drive better than most people. We have 30000+ fatalities on our roads each year. Many are due to distracted (or drunk) driving. These vehicles are never distracted. What if the accident rate plummeted?
Today’s software is becoming frighteningly capable. Deep-learning techniques, 3D vision and automated testing allow for trillions of hours of virtual driving with data sets representing all possible input variants (driving scenarios). So yes, they can determine that the vehicle behaves as expected under extremely difficult circumstances. The vehicle can also determine that it must stop and a human take over. The redundancy of the respective systems allows for extremely high levels of safety.
Vehicles will also be connected to infrastructure, with cameras, so tagging them would be a bad idea.
You don’t have to trust me but they’re coming. Everybody knows it. Nobody wants to be left behind. The law is, however, behind the technology. It doesn’t matter if the death rate plummets, it will depend on how the lawsuits influence the industry (when a crash happens). It is going to be difficult to investigate if there was corporate negligence, everyone is spreading responsibility too.
The “roll out” will come behind vehicles that won’t let you crash. So we’ll be getting used to the idea, which is already happening. Lane assist, brake assist, etc. are all becoming common on higher end vehicles and is already being brought down to lower cost vehicles. The technology always becomes cheaper so it’s also inaccurate to say (as some here have) that the cars will cost too much. You could say the same about requiring air bags, anti-lock brakes and tire pressure monitoring (etc.).
Are you comparing basic physics to AI? They're not even close.
We have 30000+ fatalities on our roads each year. Many are due to distracted (or drunk) driving. These vehicles are never distracted. What if the accident rate plummeted?
Or what if there were more accidents because AI cannot account for every (stupid) variable of human behavior behind the wheel?
Your way ONLY WORKS if EVERY SINGLE VEHICLE on the road is replaced with an AI controlled vehicle.
That's NANNY STATE, and I thought Conservatives were against the NANNY STATE!! In short: NO THANK YOU!
There were almost as many gun related deaths in America in 2015 (according to Bloomberg.Com) as there were traffic fatalities.
While we're at it, why don't we just take away everyone's guns too so we can "save lives" while we're at it!
I mean, if we're going to protect stupid people from themselves, we oughtta ban guns, trans-fats, dunkin' donuts, mc donalds, burger king, pizza hut, taco bell, zip lining, parachuting .......
Where do you stop meddling in other people's lives and freedom of choice?!
I would buy a Miata. With manual transmission.
Most of the newer metro rail systems are, for all intents and purposes, driverless. There is an operator in the cab, but they don’t do much except to handle exceptions to normal operation. The computer does everything else, including station timing, opening and closing the doors, and speed between stops.
Big rail is still mostly manual, mostly (as noted) because of unions. Also, because the environment is relatively uncontrolled compared to a city subway system. The driver has to account for weather, diverse “consists”, and the occasional obstacle on the track (tree, moose, drunk railfan etc). I’m sure that if they could escape the unions, heavy rail would be one of the first to automate as much as possible. Either autonomous engines, or remote-piloted drone engines, would work just fine for most of what they do.
That’s kinda gay.
wrong.
we are VERY close to this becoming the norm.
I would trust a computer 10000000000 times more than the average walmart shopper to be driving the car next to me on the road!
I suspect driverless cars will be phased in. Starting with a special lane on the interstates dedicated to them with a higher speed limit. Then gradually the lanes dedicated to driverless will be expanded to include the entire interstate.
Cars driven by humans will be forced to use state highways.
“The braking system will not distinguish between a bag blowing in the wind, a rabbit, a doll, a dog, and a child.”
This isn’t correct with the current systems, and things will only improve from here. I’m guessing the system will brake for anything living if it’s safe to do so, and will probably hit animals if not.
I’m curious to know what the programming is in the case where a child runs out into the road, and there’s no way to avoid him without crashing...
“Or what if there were more accidents because AI cannot account for every (stupid) variable of human behavior behind the wheel?
Your way ONLY WORKS if EVERY SINGLE VEHICLE on the road is replaced with an AI controlled vehicle.”
This is simply not correct. Google cars have already logged hundreds of thousands of miles without being at fault in a single accident. The robotic cars see in all directions at once, have reaction times much faster than humans, and do not get tired, distracted, or drunk.
You’re entitled to your own opinions, but not your own facts. :-)
That’s an irrational comparison and I’ve done nothing to impact choices people can make. Nobody is making you buy one, they’re being developed because people will WANT one. BTW, it doesn’t have to be turned on either. I love driving but there are times I wish I weren’t.
It WOULD be idiotic to hold back on a technology that will save lives because of a perception that it takes away freedom, which it doesn’t. Besides, having a driving license is a privilege, not a right - unlike owning a firearm.
Technology always impacts our lives. So let’s stop advancing technology yes? This is no different, it isn’t about reducing choice it is about increasing choice...and being safer. Whether you choose to let the car drive or not is your choice - but let me have mine too.
I agree. I also think there will be roads that only allow driverless cars because the rules will be different. E.g.no traffic lights at intersections.
Uninsurable, I would presume.
A pitching machine is more consistent than a human pitcher.
All the DWI lawyers will loose their jobs?
The comments and debate here are interesting. I’m an engineer involved in this area (Tier 1 supplier, I’m on the team that will be building the first control module for a semi-automated vehicle going on sale to the public in about 12 months). I’ve been involved with other projects like driver-assist systems (a pre-cursor to autonomous driving) since 2010.
The technology is progressing in steps. It doesn’t have to be perfect because what it is replacing (human drivers) is far from perfect as well. So if it is an improvement over that, it’s a step forward.
The concerns about people losing the skills to manually drive are a concern I share, and the issue of legal culpability will be an interesting one. I expect once we have actual cases that the legal issue will turn out to be simpler than most expect - 99% of the time it will be clear whose at fault, just as it is now.
A concern I’ve been expressing that I haven’t seen much written on is the risk of hackers, or simply extortion artists posing as hackers, taking control of vehicles. The barriers to entry are negligible - anyone with a cell phone will be able to call people and claim they’ve hacked their vehicle and can crash it, extorting money from them. The only real solution is to have bullet-proof cybersecurity in from the start so people know to blow off such calls. I’ve noticed more attention in this area from the OEM’s lately, so hopefully they are taking the issue seriously.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.