Driving to the doc, random truck parked in the road next to a no parking sign, lack of clear lane markings.
Driving to the doc, random truck parked in the road next to a no parking sign, lack of clear lane markings.
Javelin said:Driving to the doc, random truck parked in the road next to a no parking sign, lack of clear lane markings.
In reply to SVreX :
I don't think healthcare will ever become a "right". I can see that some leftists are trying, but it won't work. Not for long anyway.
Healthcare is provided by people. So unless you have a right to somebody else's labor...you don't have a right to healthcare. You may pay somebody for their services, but when you start forcing people to provide services @ some price determined by some arbitrary third party... well, you can see where that's going. It's immoral, and the incentives are all messed up. It's fundamentally unsustainable.
Javelin said:Driving to the doc, random truck parked in the road next to a no parking sign, lack of clear lane markings.
Really? You consider a truck on the side of the road to be a show-stopper?
Yes, there are millions of minor variations to the norm we encounter every day while driving and don't give a second thought. In the beginning there will definitely be situations where the AI gets flummoxed by an encounter the programmers hadn't anticipated. But modern computers can store massive amounts of data and the amount gets larger for less expense every year (month?). It's just data. And it's not like each AI vehicle will need to "learn" what to do. When a few AI systems encounter new situations and learn the response, that information gets uploaded to all of them. In the beginning, I would expect those kind a data updates to happen weekly or even daily.
gencollon said:In reply to SVreX :
Healthcare is provided by people. So unless you have a right to somebody else's labor...you don't have a right to healthcare. You may pay somebody for their services, but when you start forcing people to provide services @ some price determined by some arbitrary third party... well, you can see where that's going. It's immoral, and the incentives are all messed up. It's fundamentally unsustainable.
Not only are you describing health insurance companies to a T, but it sounds like you've never even heard of countries other than America, the last great hanger on of making people choose between medicine, housing, and food.
Javelin said:Driving to the doc, random truck parked in the road next to a no parking sign, lack of clear lane markings.
So, by using your cellphone to actually take pictures while driving, you're not really making a case that humans are smarter or safer than autonomous vehicles.
"Rights" are bullE36 M3 rules humans made up. The ONLY thing anyone is entitled to in life is death. EVERYTHING else is filler. Driving is a privilege, has been from the beginning. It was a rich mans novelty in the beginning, and like everything else in human history, it too will come full circle, and go back to being a rich mans hobby to drive yourself in the future. In fact at the rate other technology is evolving beside the autonmous cars, I'll go out on this limb and say that in 50 years, we won't need to have cars because we won't really have a need for bodies anymore.
In reply to Ian F :
But each one will have to learn on it's own due to the limitations of the lack of network connectivity that we already discussed. Either the cars are connected, and limited to cities where the network exists, or they aren't and each one will need processing power that's currently way too expensive to self-learn.
And yes, random stuff blocking traffic has flumoxed the autonomous cars so far. They are programmed to be careful and law abiding, and crossing a double yellow into incoming traffic is definitely not something they've been programmed to do. Even simple decisions like this are hard from a programming standpoint.
In reply to RevRico :
I was stopped, but fun strawman. I never said anything about rights. I said the road infrastructure, network, and technology isn't ready, nor will it be.
In reply to Ian F :
Sunspots have disrupted satellite communication it's now impossible to communicate with the collective - what now?
In reply to Javelin & Rons :
You can come up with a million reasons why autonomous cars can't work, and I still contend these are just problems to be recognized and overcome. Will there be unforeseen challenges and setbacks? Sure - but each one will provide a chance to learn and thus move the entire process forward.
In the grand scheme of the challenges facing humanity, the autonomous car is nothing.
Javelin said:In reply to RevRico :
I was stopped, but fun strawman. I never said anything about rights. I said the road infrastructure, network, and technology isn't ready, nor will it be.
Uhh... Really?
Here, this is a picture of the "Golden Gate", the strait that connects the San Francisco Bay to the Pacific Ocean. No way you're getting across that without a watercraft, right? These new fangled cars, they'll just have to go around:
Wait a second... Can they do that???
Wow!!!!
Other examples, there is no way the automobile will work, it can't get through the muddy path up to my house, or it can't get through these trees, it can't get over/around these mountains, horses/oxen are better, blah blah blah.
So we can build a cross country (continental!) interstate system, over rivers, lakes, bays, around, over, and through mountains and valleys, taking down trees to do it, but a stopped truck will be the downfall of something... right...
It'll take time, improvements need to be made to roll it out fully, but the fact is that they've already completed millions if not billions of miles.
In reply to Ian F :
Ian I love to give the mick to those who act as evangelists for the next new great thing. I believe that people only look at the goods and not the negatives.
In reply to Rons :
Oh, I can see a lot of negatives. I just don't believe the technical challenges represent a great hurdle, which is the point I've been arguing. Philosophical debates are not really my concern.
Humans are great at making tools. Using those tools, however, often exposes the young age of the species.
In reply to Ian F :
I have a Computer Science degree, I also drive a semi. So I have some insight in to both areas of this.
Software is the most complex item humans make. The AI aspect to make decisions is the main stumbling block. Along with mobile processing power to run it.
Just analyzing my voice for a simple "AI" requires a GPU to eliminate the internet connection. I play with an AI called Mycroft and it is really more of a Chinese room system. And that powerful GPU sucks power and still lags a bit.
Most people think of an AI as a software based person but they are specific to tasks. One programmed and trained to read handwritten text could not easily (or maybe not at all) understand the stereoscopic cameras required for distance and lane position. And vice versa.
I have seen the decline of driver skill and attention span during the 37 years I have been driving. I also know an AI could not do what I do every day. Not at this point in development.
Someday? Possibly. My lifetime? Probably not.
In reply to mtn :
Creating a fully self-aware, grid-free, autonomous car that can truly recognize every situation drivers face within a budget people can afford without technological glitches (hello printer driver that worked fine yesterday!) is about a thousand exponential times more difficult than putting steel, concrete, and asphalt up. Seriously.
Can it be done? Likely. Will it be affordable? Unlikely. Will it be legally mandated to replace self-driven vehicles? No. At least not nationwide. Some city centers or transit corridors sure, but not nationwide, not even lower 48. I'm being realistic, not pessimistic. Hell, I'm personally invested in most of the leaders of the autonomous car technology (Tesla, Ford, Garmin).
Miles traveled mean nothing. Please, read this.
In reply to Javelin :
Why are you hung up on network connectivity?
Every car can download when it has available connectivity. They don’t have to have continuous connectivity.
It isn’t necessary for cars to be independently autonomous, and it isn’t necessary for them to learn a particular problem instantly.
If a car on a remote Alaskan forestry road encounters something new, it can store that info until the next time it has connectivity- even weeks or months. Then share it with other cars.
The power in the existing network is amazing.
I use DropBox every day, and yet I almost never have network connectivity. Because I am in the business of building the buildings that have that connectivity. Doesn’t matter. I go back to a hotel every night, turn on the ball game, and DropBox loads any new info it finds completely invisibly to me.
The connectivity issue seems to me to be one of the easiest to solve.
I think we'll get there soon and easily (easily for the smart people :)
But it will be by degrees. I don't need an autonomous car to drop me off in a specific unmarked alley with a truck in the way...around the corner will be fine.
Places that want you to visit might even have some sort of hardware/software combo to guide your car in quickly....luxury resorts maybe.
There will be places and specific situations where regular vehicles are more common and needed.
They're still driving Yaks with carts along side Soviet trucks in Mongolia, who cares.
What they're going for is 'good enough' for an ever increasing market share... not perfection and not a 'self-aware' car.
So in the forest of all the reasons for both ways, one other thing to consider, WRT banning person driving cars, is market penetration.
There are roughly 270M cars in the US. Of which about 12-15M are permanently taken off the road annually.
Pretend that 1000% cars in 2030 are autonomous- it will 10 years, at a bare minimum to replace half of the US fleet. Half.
Given that the technology will take a good decade to ramp in, the time that half of the US fleet is self driving- picture closer to 2050.
Perhaps at some point, all new cars will have to be self driving in some respects. But the time frame to replace ALL of the cars in the US- to the point where you can ban people driving cars from certain roads, is going to be quite a while. And much of that will be dependent on self driving cars being eligible for the $2050 challenge.
Which is another way of saying that the group that is asking for your money to preserve people driving cars is more likely a scam to take your fearful money than it is reality.
In reply to alfadriver :
Pretty close to the way I see it and I still think that's probably optimistic. And why I believe I won't see it in my lifetime.
In reply to alfadriver :
I completely agree. That’s about how I see a “complete takeover” of autonomous cars. (Quite similar to the “takeover” of ICE cars).
At the opposite end of the spectrum is the point of entry. The starting point. I’d say we are at pre-entry right now. Development.
I think the starting point is the moment they are legalized for use on public roads. I’m gonna guess before 2030. Perhaps as early as 2025. That’s only 6 years from now.
I think we often mix up those two moments when we consider this. The legalization moment and the takeover moment. There is a wide gap between them. Like 50 years or much more.
I generally think of a midway point. Not the entry moment- too many glitches. Not the full takeover moment- I’ll be dead. But rather the moment when they are viable and available and reasonably common.
I expect to see the viability moment in my lifetime. 20 years.
In reply to SVreX :
I would not equate autonomy with ICE's a century ago. Back then, the horse market was not nearly as penetrated as the car market is now- because people also walked. Cities were built around walking to and from places as opposed to having some kind of transportation. So realistically, the "replacement" part of cars was a pretty small segment of mobility. Cars were a very new way of getting around, and they very much shaped how communities formed, especially post WWII.
Also, horses have never been banned from all roads, as is being suggested by this "movement". The only place they are really restricted are freeways, but they are limited access for more than just horses.
So the idea of banning human driven cars along with the long phase in time of this new technology- that just does not make sense. Lest we forget (as I pointed out before) pre Clean Air Act cars are still very legal on the roads. Heck, even "radical" California has not banned pre-'68 cars. So I just don't see pre-autonomy cars being banned.
I agree that there will be autonomous cars within my lifetime. Maybe even before I retire. But banning of human driving cars? Not a chance.
I could imagine human drivers banned from certain road sections - like separated HOV lanes in some areas - but otherwise I agree I don't see human cars ever being entirely banned in the same way I don't see ICE vehicles ever completely going away. Even that HOV lane would likely start out with semi-autonomous cars - those with active (and in the future interlinked?) cruise control systems that would allow the cars to travel in close proximity at higher speeds. Of course, the trick there is how to keep human interlopers from trying to get into the mix and thus screw up the flow. I have no good answer for that.
In reply to alfadriver :
I agree.
I’ve never said I thought human driven cars would be banned.
And you’re right- there is an enormous difference between the turn of the last century and now. The changes are much more complex now, but technology also advances at a MUCH faster pace.
I am recognizing the similarities in the fact that there were massive cultural shifts and changes in how transportation functioned, just like now.
It didn’t happen all at once, but over 50 years NOTHING remained unchanged. The same could be said about the next 50 years.
The interesting thing is Uber is about to do an IPO, which will probably go a bit insane, apparently all based on the fact that it's an excellent base for self-driving cars.
As explored above, even restricted area self-driving cars are likely still a ways away. So, they will still need drivers for most routes, which is the primary problem with Uber now (under payed, pissed off, "contractors", not really "sharing" a ride, unrealistic business model, losing money on every ride...).
You'll need to log in to post.