Wouldn't we need "smart" roads to make driverless cars operate at the same safety level as, say, airplanes?
The dream of driverless cars is dying
21–30 of 124 posts
Re: The dream of driverless cars is dying
#22Earlier quoted context omitted.
For that to happen does that 2016 Honda Civic have to have a driver in it? That's how you can tell if it's driverless.
You don't have to have ubiquitous Turing-test-passing level 5+ robot vehicles. Lane Keep Assist style things are a good start.
Lane assist is unimpressive, requirements, technology and capability wise. Also, a fourteen year old can do it from the passenger seat.
(Also they don't really work reliably: try using lane assist with roadworks markings and watch your car veer off into the barricades. Or just too-unclear-for-machine—but-blatantly-obvious-for-human markings, like the dead Tesla driver from a few weeks ago encountered. Even a fourteen year old doesn't make that sort of mistake.)
Re: The dream of driverless cars is dying
#23Reminds me of this famously wrong column dismissing the Internet back in 1995: https://www.newsweek.com/clifford-stoll-why-web-wont-be-nirv... > How about electronic publishing? Try reading a book on disc. At best, it's an unpleasant chore: the myopic glow of a clunky computer replaces the friendly pages of a book. And you can't tote that laptop to the beach. Yet Nicholas Negroponte, director of the MIT Media Lab, pr…
To be fair, I find the experience of reading a book or newspaper on at the beach is still inferior to having the dead tree version.
Re: The dream of driverless cars is dying
#24What would a TOS look like for a self driving car?
“Shall hold manufacturer, software developer, etc not liable for any injury or death from the result of using car...”
Re: The dream of driverless cars is dying
#25While I'd agree that driverless cars are not here today, that the hype is over-doing it, and that Tesla has tried to push too fast, there is a huge economic incentive to have driverless cars. If a system can be built to replace long haul truck drivers with even being slightly safer than a human driver, that saves some on insurance and tons on labor, not to mention not having to limit driving time per day. With those kinds of incentives, trucking companies wouldn't blink an eye before going forward with it. People seem so focused on personal vehicles that they ignore that people won't want to give up control at first and don't realize how the change will take place.
Re: The dream of driverless cars is dying
#26We're going to get driverless cars one way or another. Even if the problems are much more intractable than we first gave them credit for people are too clever. Take driving in the snow, for instance. It might be completely impossible to do it safely for the next twenty years by having a computer manage the drive path, but designs might create the ability for cars to form mini-trains where the lead car is driven by a…
Snow seems like one of the easier problems, honestly. Pretty much all the "technical" problems seem pretty straightforward. Waymo already posted an article where they talked about how they can filter falling snow out of their LIDAR input, for example.
It's the social problems that are hard for AI, like how to handle a construction zone that looks like you need to break the law to get around. Those don't always have easily generalizable answers.
Re: The dream of driverless cars is dying
#27We're going to get driverless cars one way or another. Even if the problems are much more intractable than we first gave them credit for people are too clever. Take driving in the snow, for instance. It might be completely impossible to do it safely for the next twenty years by having a computer manage the drive path, but designs might create the ability for cars to form mini-trains where the lead car is driven by a…
Obvious question: why just not take the train? In your example I’d have to depend on other strangers for my private travel, one of the most important things that made me purchase my first car in my 30s, I don’t want that.
At least in the US, transit is unfortunately quite bad on average, and that doesn't look to change anytime soon in most areas.
Re: The dream of driverless cars is dying
#28Reminds me of this famously wrong column dismissing the Internet back in 1995: https://www.newsweek.com/clifford-stoll-why-web-wont-be-nirv... > How about electronic publishing? Try reading a book on disc. At best, it's an unpleasant chore: the myopic glow of a clunky computer replaces the friendly pages of a book. And you can't tote that laptop to the beach. Yet Nicholas Negroponte, director of the MIT Media Lab, pr…
We still don’t have flying cars nor self-conscious robots, things that have been prophetized 100 to 50 years ago. Going back more in time we still don’t have a philosophical stone. The idea is that some technological prophecies are indeed wrong.
The former were pure prophecies with no viable, known methods to achieve them at the time. Self driving cars are in active development, and look to be within spitting distance of full autonomy at least for most driving scenarios. It's possible that last % of situations will be super hard to solve, sure, but it's still a very different thing than someone just going, "oh man it'll be cool when".
Re: The dream of driverless cars is dying
#29The part that I can't figure out is why everyone seems to act like driverless cars will be some kind of all-at-once phase change transition instead of being incremental. It seems almost a certainty that we will gradually ease into the concept. Some cars will be able to do some things autonomously, a little, for some people, some of the time. Eventually the cost/benefit levels will improve, more features will roll out…
Yes, and it's already happening. Automakers have already agreed to implement automatic collision avoidance in all their cars by 2022, IIRC, so there's one example.
Re: The dream of driverless cars is dying
#30We're going to get driverless cars one way or another. Even if the problems are much more intractable than we first gave them credit for people are too clever. Take driving in the snow, for instance. It might be completely impossible to do it safely for the next twenty years by having a computer manage the drive path, but designs might create the ability for cars to form mini-trains where the lead car is driven by a…
Snow seems like one of the easier problems, honestly. Pretty much all the "technical" problems seem pretty straightforward. Waymo already posted an article where they talked about how they can filter falling snow out of their LIDAR input, for example. It's the social problems that are hard for AI, like how to handle a construction zone that looks like you need to break the law to get around. Those don't always have e…
> It's the social problems that are hard for AI, like how to handle a construction zone that looks like you need to break the law to get around.
For these cases, the car could have a manual mode (?)