|
Solkanar512 posted:I brought this up in the previous thread, but what happens if manufacturers end up treating these cars like Android phones where they require updates even after release, but after a year or two stop supporting development? Or maybe the manufacturer goes out of business and there's a massive security issue that comes up, what then? ideally by then manufacturers will get serious about security, and this goes for pretty much everyone involved in the internet of things but ideally since there's so few car producers they should get their poo poo together quicker as an industry. i think if self driving cars coordinating is a reality we can assume at least some decent level of security since it'll be harder to get them to agree on standard protocols for communicating intent and behavior than it will making those communcations secure
|
#
?
Jan 19, 2017 07:06
|
|
|
|
| # ? May 5, 2024 23:57 |
|
|
boner confessor posted:ideally by then manufacturers will get serious about security, and this goes for pretty much everyone involved in the internet of things but ideally since there's so few car producers they should get their poo poo together quicker as an industry. i think if self driving cars coordinating is a reality we can assume at least some decent level of security since it'll be harder to get them to agree on standard protocols for communicating intent and behavior than it will making those communcations secure Yeah okay, let's just ban cars because lol if you think they'll get anywhere near serious about anything besides maybe proofing the lives of rich customers.
|
|
#
?
Jan 19, 2017 10:42
|
|
|
Ardennes posted:There is a genuine fear (with some honest weight behind it) that autonomous cars would only basically backtrack on the progress of the last two decades towards urbanization. Autonomous cars aren't the magic bullet they are cited as but at the same time I think it is fair to see them as a return to the quasi-segregationist ideals of the 1950s where you can sit in leisure in a glass box and not have to deal with the "urban underclass" whizzing by. Obviously, that already in reality happened to the US but the fear is that autonomous cars would led to a return to that period history.
|
|
#
?
Jan 19, 2017 11:39
|
|
|
Cicero posted:If anything I think the effect will be the reverse. Fleets of self-driving taxis will encourage more people to give up owning their own car, and once you don't own your own car, you're going to look at transit more seriously. IIRC this has already happened to some extent with Uber/Lyft/etc. where people who mostly don't need a car but occasionally do used to still have to own their own but can now just call one from an app. This is great if your upper middle class, but there is little reason for Uber/Lyft to drop their prices to any significant degree.
|
|
#
?
Jan 19, 2017 12:04
|
|
|
I think what a lot of people itt aren't recognizing, is that said driving cars aren't a solution to congestion, even if they're been presented as such. People have complained that suburbs are here to stay, because there is just so much of it, but that's ignoring economics. Continuous suburban expansion is not sustainable, it's wasteful of resources, and is going to fall on its face sooner or later. A house is worthless if people cannot afford to live in it, because governments can't afford to provide services to them, because everything is too spread apart. A contraction has to occur. Either you start to deal with it now, or you ignore it, and deal with the consequences. That doesn't mean you ignore what's there, or not try to adapt your solution to what you have available. Of course you do that. What you don't do is pretend the situation we have continues forever. Like I fully expect that you're going to end up with acres and acres of abandoned suburban lots, which then just transforms into a weird kind of shanty town as you get close to hubs, as the eventual end state, if current trends continue. Either you adapt to conditions, or conditions force you to adapt.
|
|
#
?
Jan 19, 2017 12:36
|
|
|
Ardennes posted:This is great if your upper middle class, but there is little reason for Uber/Lyft to drop their prices to any significant degree. Prices will be lower than now because there'll be no driver, and by then electrics will be cheaper (electricity is a lot cheaper than gas and long-run should have fewer maintenance/repair issues). I dunno if they'll be affordable to the working class as a replacement for all trips, but affordable as at least an occasional thing definitely.
|
|
#
?
Jan 19, 2017 12:36
|
|
|
rudatron posted:I think what a lot of people itt aren't recognizing, is that said driving cars aren't a solution to congestion, even if they're been presented as such. People have complained that suburbs are here to stay, because there is just so much of it, but that's ignoring economics. Continuous suburban expansion is not sustainable, it's wasteful of resources, and is going to fall on its face sooner or later. In any case, the development self-driving cars and increasing urbanization/densification of America aren't conflicting things, and I'm not sure why some people are portraying them as conflicting. If anything they can support each other by making it more practical for people to give up personal car ownership.
|
|
#
?
Jan 19, 2017 12:41
|
|
|
American style suburbs are pretty big in America? Which is kind of where a lot of this discussion is focused? I see what you're saying, and i agree in some respects. I have real skepticism it can live up to all the hype, but the tech is good, and it's here to stay. I just don't think that cars may be the most popular or even best use of it. Like have you guys ever seen a brt system? You could mcguyver up something like that on already existing roads, if you're clever about it. Take a road, apply whatever special reflective paints, or other measures your loving computer needs to not gently caress up, block all other traffic from going on them, and have a constant stream of automatic buses, smart enough to recognize when you hail them - boom, there's your mass transit system.
|
|
#
?
Jan 19, 2017 13:05
|
|
|
Owlofcreamcheese posted:This is the derail that started this whole thread but look up "traffic waves", in some cases they are totally unavoidable but are in many cases a result of the way human drivers stack up on stopped or slower moving traffic (getting closer to the car ahead of you if it's slower and farther if it's faster, pulling right up to stopped cars) instead of staying equidistant. Even putting a few car lengths between stopped cars would cure many traffic jams but people refuse to drive like that. They always have and always will pull right up to the stopped car and then when they start moving again they have to pause longer before they can start moving to give the car in front of them more distance instead of stopping with more distance already there. The whole point is that if you switch to just autonomous cars the increased congestion mitigates this effect, and you can't decrease the distance between cars because of inertia making any accident a whole lot worse. Raenir Salazar posted:Some wise words from the King of the Internet This is completely unrealistic for multiple decades to come, and also makes any accident a million times worse because he forgets to factor inertia into his musings. Both physical and societial inertia, btw. Autonomous cars are not a good solution to congestion by themselves. They have some good things about them: they'll at some point lower accidents (the current numbers are cooked and compare distance per accident on almost primarily highways with what drivers do on all roads in the world, that's why they look safe. Also, ignoring problems in China), but easing congestion will not happen for multiple decades. So it is a pie in the sky solution that will not materialise for a long time, approaching the problem from the wrong direction (when all you've got is an autonomous car...). And in the transition period, will be worse for the poor than it is for the rich. Meanwhile, mass transit and infrastructure reshaping is a realistic goal that has already been proven in multiple cities worldwide. It can be done within a realistic timeframe. Claiming it's an untenable goal is ludicrous when engineers have managed to lift an entire city for rebuilding with minimal disruption in less than a decade, all in the 19th century. In more recent times, Japanese have fixed a loving 30x27x15m sinkhole in 48 hours and apologized for being so slow. It can be done, with funding. It doesn't rely on non-existent technology but can use the existing ones (for instance, the lights in my city are controlled by a neural network that learned to ease congestion after the rebuilding). Mass transit is especially good for the poor. And as a bonus, it will benefit autonomous cars too, so you can have your cake and eat it. dex_sda fucked around with this message at 13:12 on Jan 19, 2017 |
|
#
?
Jan 19, 2017 13:10
|
|
|
Cicero posted:Have you never heard of market pressure before? It's not a matter of them going, "yeah dropping our prices sounds like a great idea!", they'll have no choice. Every carmaker that manufactures self-driving cars will be a potential competitor. Are you talking about Uber and Lyft? The way they do business is explicitly to take advantage of market pressure to avoid any competitor to pop up. They're so cheap now thanks to VC precisely to undermine any market pressure that can be applied against them at some point in the future, allowing them to massively raise prices.
|
|
#
?
Jan 19, 2017 13:15
|
|
|
rudatron posted:I think what a lot of people itt aren't recognizing, is that said driving cars aren't a solution to congestion, even if they're been presented as such. Are they presented primarily as a solution to congestion itself or as a solution to the "problem" of congestion. Like no, they are no help in giant long lasting gridlock traffic jams, but day to day congestion is generally bad because it's a burden and an annoyance to drive in. like the fact it makes you get where you are going five minutes late is suboptimal but the reason it was terrible is because you have to drive in it. Like it subjectively makes the experience of driving bad because you are having to make a lot of effort to do it. Like people would rather drive down a straight empty street than a street with a bunch of stop lights even if the total time was the same. But in a self driving car one is the same as the other. You aren't doing anything either way. So congestion as a problem only becomes the amount of time it delays you instead of the way it is now where it's bad even if it's not delaying you significantly because it's just unfun to drive in.
|
|
#
?
Jan 19, 2017 13:22
|
|
|
Cicero posted:For American-style suburbs, yes. But if you look at, say, Europe, you can find suburbs that have a much denser form. And yeah because they're denser cars aren't a required thing, but lots of people still drive. Car usage in the future will probably (hopefully) be reduced, but they're not going to be gone completely. I guess you haven't heard of cartels and price fixing, it is in the best interest of all of those companies to keep prices in a tight band. If anything what would massive increase would be the profit taking as companies simply pocket the money. If people are still willing to pay "around" uber/lift pricing then that is what they are going to charge. Europe and its suburbs are also far far more pedestrian friendly and generally have better public transportation. Self-driving cars are likely just going to make cars more attractive, the problem the car as a vehicle isn't really that great for a livable society. It will certainly more efficient and give people more leisure time but it isn't without some pretty clear negatives that our society probably doesn't need.
|
|
#
?
Jan 19, 2017 13:36
|
|
|
Solkanar512 posted:I brought this up in the previous thread, but what happens if manufacturers end up treating these cars like Android phones where they require updates even after release, but after a year or two stop supporting development? Or maybe the manufacturer goes out of business and there's a massive security issue that comes up, what then? The manufacturer is almost certainly liable for all damage caused by a flaw in the software. Maybe they try to limit it with EULAs but I doubt that would be permitted when it came to physical objects where the software is an intrinsic part of the functionality. As to security issues, that's something that I don't think our legal system has really figured out yet. But auto-makers are going to need to take that seriously and I'm not positive they will - I'd be concerned they slap security on at the end like a lot of companies do that are focused on the hardware, instead of really designing it top to bottom for security. boner confessor posted:ideally by then manufacturers will get serious about security, and this goes for pretty much everyone involved in the internet of things but ideally since there's so few car producers they should get their poo poo together quicker as an industry. i think if self driving cars coordinating is a reality we can assume at least some decent level of security since it'll be harder to get them to agree on standard protocols for communicating intent and behavior than it will making those communcations secure Security is hard, and the more decision makers involved in the chain who aren't experts on the subject (like people who work in software industries where their code is routinely attacked) the more they're likely to overlook, underfund, or underemphasize it. Car companies strike me as having a lot of people who have no earthly idea that's an issue. evilweasel fucked around with this message at 16:20 on Jan 19, 2017 |
|
#
?
Jan 19, 2017 16:18
|
|
|
rudatron posted:I think what a lot of people itt aren't recognizing, is that said driving cars aren't a solution to congestion, even if they're been presented as such. People have complained that suburbs are here to stay, because there is just so much of it, but that's ignoring economics. Continuous suburban expansion is not sustainable, it's wasteful of resources, and is going to fall on its face sooner or later. suburbs are growing in a couple of different ways 1) they're densifying in good ways - well off suburbs recognize that previous autocentric models are unsustainable and that it's better to invest now in making yourself more like a small town than just a mess of subdivisions and strip malls. i can point to two examples (there are more) in the atlanta suburbs, duluth and suwanee. both of them have extensively rebuilt, basically from scratch, their downtown areas duluth, incremental infill development and renovation of the town green:  suwanee, site redevelopment - this used to be a strip mall, now it has retail, townhouses, condos, and single family homes, and government buildings in a brand new downtown:  2) they're densifying in bad ways - since suburban growth is haphazard, there's often plenty of potential for purely residential infill. newer developments tend to be thicker, either high FAR single family homes that dominate a small, constrictive lot as opposed to modest homes on large lots, or one of the uglier forms of suburban development, subdivisions of townhomes 3) they're growing further out - there's still a lot of land to be developed in many american cities, particularly inland sunbelt cities. this is one reason people are a little worried about self driving cars. new development keeps happening on the fringes because suburbs haven't hit a spatial limit yet in many places 4) they're growing poorer - one of the biggest problems in suburbs is concentrated suburban poverty. there's a lot of places that, like you said, would become shantytowns of older, less desirable housing, with no nearby jobs or services, just ugly places to live. these places are becoming home to urban residents being displaced by gentrification and an increasing share of low-income immigrants evilweasel posted:Security is hard, and the more decision makers involved in the chain who aren't experts on the subject (like people who work in software industries where their code is routinely attacked) the more they're likely to overlook, underfund, or underemphasize it. Car companies strike me as having a lot of people who have no earthly idea that's an issue. security is hard, but i dont think it's harder than self-coordination of current position, intent, etc. across multiple vehicle types and makes. and car flocking behavior is a problem for the near future, car security is a potential problem right now - hackers can already tamper remotely with some of the more cutting edge models
|
|
#
?
Jan 19, 2017 17:21
|
|
|
How feasible do you think that actually is, though? Like, you're aware that cars cost money to run, right? And now you've strapped this computer, LIDAR, and all the other bullshit to it, just so your 1 hour commute is somewhat bearable, and you've got that running with the air conditioning because otherwise you'll just die in the heat - all of that, costs money, costs resources. It's so wasteful. You get any kind of contraction, suddenly it all falls apart.
|
|
#
?
Jan 19, 2017 17:38
|
|
|
rudatron posted:How feasible do you think that actually is, though? Like, you're aware that cars cost money to run, right? And now you've strapped this computer, LIDAR, and all the other bullshit to it, just so your 1 hour commute is somewhat bearable, and you've got that running with the air conditioning because otherwise you'll just die in the heat - all of that, costs money, costs resources. It's so wasteful. You get any kind of contraction, suddenly it all falls apart. Every post about everything in the future just assume I included "death is certain, the age of man is ending" or whatever, and if I say avenger infinity is coming out in 2019 I really mean 'avengers infinity is coming out in 2019 if the great and merciful seas have not returned mankind to nothingness and there is a man alive who does not foolishly eat the filmstock trying to claw another day of life on the hell blighted world" or whatever but was trying to save time. Yes, if the age of cars ends the age of automatic cars will also end. That probably isn't in the timeframe of the next 8 years or so.
|
|
#
?
Jan 19, 2017 17:55
|
|
|
MaxxBot posted:How do you propose convincing people to willingly choose less convenient forms of transportation? It literally takes me an hour and a half to bus to work and honestly it's a miracle there's even a route there at all, people are not going to willingly subject themselves to that if they don't have to. This is living in the middle of a large city right by a high frequency bus line and riding at the times when the busses run most frequently. A lot of this is due to underinvestment in transport, in large part due to complaints from car owners. This particularly affects buses. We know how to make buses faster, but the measures necessary to do that are unpopular with drivers, who resent sitting in gridlock while buses cruise by and insist that the infrastructure builds necessary be redirected to the benefit of cars instead of buses. In many cases, bus rapid transit plans have quickly shed features during the design and early implementation phases as they face political backlash from angry drivers who refuse to tolerate anything going faster than them and stubborn officials looking for places to cut the budget. Dedicated bus lanes/busways don't get built or are quickly opened to regular traffic, and traffic light preemption gets turned off. It's not like we can't make public transit faster - it's just that commuters generally prefer that the money be spent on adding road capacity instead (even though traffic engineers know that merely adding capacity doesn't meaningfully reduce traffic). Owlofcreamcheese posted:The idea of triple convergence is that you increase throughput but that doesn't decrease 'congestion", a bigger highway will carry more people but still have everyone driving exactly as slow. But it'll now be twice as many people driving slow. You still have more people getting to their destination in a unit of time, but at the same slow speed. It's an absolute net gain in throughput for the population but the individuals travel at the same rate. No, you're wrong, and triple convergence has absolutely nothing to do with that. Which is strange, because triple convergence is a well-understood phenomenon which has been extensively observed and studied for decades, so the only way you could possibly be wrong about this basic principle is if you don't have even a basic level of respect for the science and study that has gone into transit planning and traffic engineering. The principle behind triple convergence is that some people avoid routes they know will be congested, because congestion is unpleasant and a lot of people don't like it, so congestion is self-limiting. If you decrease congestion on a given route, then the congestion will no longer disincentivize people from taking that route, and therefore the number of people taking that route will rapidly increase until the route is over capacity again.
|
|
#
?
Jan 19, 2017 18:27
|
|
|
Uh, what about the fact that a lot of people avoid congestion because its annoying as gently caress to drive and keep 100% of your attention to drive 100 meters in 30 minutes if it gets bad? I've only ever been a passenger and its only vaguely acceptable to me because I can take a nap. The advantage of self driving cars for making "congestion" better is aside from increased efficiency is the ability to not have to engage with the vehicle. You can watch a movie now, no rush.
|
|
#
?
Jan 19, 2017 19:01
|
|
|
Raenir Salazar posted:Uh, what about the fact that a lot of people avoid congestion because its annoying as gently caress to drive and keep 100% of your attention to drive 100 meters in 30 minutes if it gets bad? I've only ever been a passenger and its only vaguely acceptable to me because I can take a nap what about it? this is why people use alternate routes, time shift their trip to happen before or after peak congestion, or take a different method of transport like a helicopter or jetpack. if you're plowing through anyway because of a failure to determine an alternate path around congestion then that's silly, for one, and is likely to cause you to focus more on the imaginary fantasy of self driving cars than the probable reality Raenir Salazar posted:The advantage of self driving cars for making "congestion" better is aside from increased efficiency is the ability to not have to engage with the vehicle. You can watch a movie now, no rush. this is almost certainly going to be illegal for the next 10-20 years or so, not that it stops people driving manual cars from dicking around with their smartphones like short attention span toddlers the fantasy is that self driving cars will behave like diet pills. i can eat all i want, but not get fat! i can drive all i want, but not have to deal with traffic! the reality is that until self driving cars are robust enough, you will have to pay attention to the road conditions so you can take over in an emergency without frantically pausing your visual novel or whatever vvv it's amazing that oocc can be straight up proven to have no idea what he's talking about re: traffic engineering and yet he just keeps going! the amazing perpetual boy, always talking but never listening. incredible! boner confessor fucked around with this message at 19:16 on Jan 19, 2017 |
|
#
?
Jan 19, 2017 19:13
|
|
|
Main Paineframe posted:No, you're wrong, and triple convergence has absolutely nothing to do with that. Which is strange, because triple convergence is a well-understood phenomenon which has been extensively observed and studied for decades, so the only way you could possibly be wrong about this basic principle is if you don't have even a basic level of respect for the science and study that has gone into transit planning and traffic engineering. The principle behind triple convergence is that some people avoid routes they know will be congested, because congestion is unpleasant and a lot of people don't like it, so congestion is self-limiting. If you decrease congestion on a given route, then the congestion will no longer disincentivize people from taking that route, and therefore the number of people taking that route will rapidly increase until the route is over capacity again. It's saying that if you have a two lane road that is all jammed up and increase it to a 4 lane road that that road will probably also get all jammed up. It's not saying that all roads have identical capacity. a 4 lane road that is slowed to a crawl still can crawl more cars simultaneously than a 2 lane road.
|
|
#
?
Jan 19, 2017 19:13
|
|
|
boner confessor posted:this is almost certainly going to be illegal for the next 10-20 years or so, Oh no, not 10 to 20 years! Will any of us even be alive in the far distant future of 2027?
|
|
#
?
Jan 19, 2017 19:16
|
|
|
Raenir Salazar posted:Uh, what about the fact that a lot of people avoid congestion because its annoying as gently caress to drive and keep 100% of your attention to drive 100 meters in 30 minutes if it gets bad? I've only ever been a passenger and its only vaguely acceptable to me because I can take a nap. If there is no rush to get where you're trying to get, why are you even taking a car?
|
|
#
?
Jan 19, 2017 19:32
|
|
|
boner confessor posted:the reality is that until self driving cars are robust enough, you will have to pay attention to the road conditions so you can take over in an emergency without frantically pausing your visual novel or whatever eh, this will continue to be a theoretical requirement to lower liability for car companies, but as a practical matter it simply isn't a viable solution: relying on people to be attentive enough to what the car is doing such that they can (1) recognize the car is making an error; (2) engage manual control, (3) take the required action correctly all in whatever amount of time is available is just not going to happen. people aren't attentive enough to rare events for this to happen even if they're trying to pay attention - I expect that the only reason it even works well in the testing is that the testers have a pretty good understanding of the flaws of the machine and can recognize when a situation that might exceed what the system is currently capable of is occurring ahead of time, or knows what the typical behavior is enough to know something's not right. I think you're much more likely to have cars able to alert you when their awareness of what is happening is dropping to unacceptably low levels of confidence and you need to take over in the next few seconds, giving enough warning that a distracted driver can recognize what's going on and orient themselves.
|
|
#
?
Jan 19, 2017 19:52
|
|
|
evilweasel posted:eh, this will continue to be a theoretical requirement to lower liability for car companies, but as a practical matter it simply isn't a viable solution: relying on people to be attentive enough to what the car is doing such that they can (1) recognize the car is making an error; (2) engage manual control, (3) take the required action correctly all in whatever amount of time is available is just not going to happen. people aren't attentive enough to rare events for this to happen even if they're trying to pay attention - I expect that the only reason it even works well in the testing is that the testers have a pretty good understanding of the flaws of the machine and can recognize when a situation that might exceed what the system is currently capable of is occurring ahead of time, or knows what the typical behavior is enough to know something's not right. I'm gonna hire a guy from uber to log into my car and react for me any time my car can't figure it out automatically. He can basically sit at a desk all day and wario ware his way through all the one off times a goose is standing in the road or something for the hundred or so cars under his care.
|
|
#
?
Jan 19, 2017 20:02
|
|
|
Good news for all you train-haters in this thread: Trump may zero out federal transit funding! I for one look forward to my budget becoming "pick two: rent, student loan payments, or car insurance/gas payments" after I have to buy a lovely beater car to get to work. 
|
|
#
?
Jan 19, 2017 21:27
|
|
|
Huh, Elon Musk thinks self-driving cars will make traffic worse:quote:“Without tunnels, we will all be in traffic hell forever,” Musk told The Verge via Twitter DM today. “I really do think tunnels are the key to solving urban gridlock. Being stuck in traffic is soul-destroying. Self-driving cars will actually make it worse by making vehicle travel more affordable.”
|
|
#
?
Jan 26, 2017 11:18
|
|
|
Cicero posted:Huh, Elon Musk thinks self-driving cars will make traffic worse: I'll believe that when I can buy a $15,000 Tesla minivan. Until then, Musk needs to keep his mouth shut about affordability.
|
|
#
?
Jan 26, 2017 13:54
|
|
|
That's a pretty bizarre and stupid response to someone who has been pretty upfront about the costs of their cars, and especially the trajectory. An electric car with significant range that was reasonably affordable was an impossibility until right around now. Plus he's obviously referring to the potential of self-driving taxis to lower trip costs, not that the car itself will be significantly cheaper than cars now (although operating costs of electric cars may also make them more affordable compared to gas cars in the long run).
|
|
#
?
Jan 26, 2017 14:30
|
|
|
anonumos posted:I'll believe that when I can buy a $15,000 Tesla minivan. Until then, Musk needs to keep his mouth shut about affordability. The tesla 3 is 35,000 dollars. The average price of a new car bought in the US is 33,560. You can argue that the whole concept of "new car" excludes a large percent of the american poor and middle class but that is a whole other thing, teslas now fall absolutely in the "affordable" category as much as any new car can. No company manufactures 15,000 dollar minivans. Minivans are just old enough you can buy a used one.
|
|
#
?
Jan 26, 2017 14:34
|
|
|
rudatron posted:Like have you guys ever seen a brt system? You could mcguyver up something like that on already existing roads, if you're clever about it. Take a road, apply whatever special reflective paints, or other measures your loving computer needs to not gently caress up, block all other traffic from going on them, and have a constant stream of automatic buses, smart enough to recognize when you hail them - boom, there's your mass transit system. Yea but there will still be some lost cause suburbs. For normal people, at least.
|
|
#
?
Jan 26, 2017 15:29
|
|
|
anonumos posted:http://www.theverge.com/2017/1/25/14391410/elon-musk-tunnels-traffic-tesla-spacex-boring-company by "affordable" i think he means in terms of time and opportunity cost, not money. self driving cars will likely make people deal with longer, uglier commutes, or time shift into causing more congestion at rush hour because instead of having to actually drive in it they can just read a book or whatever
|
|
#
?
Jan 26, 2017 16:39
|
|
|
Self driving cars probably will be more affordable in terms of cost, especially for occasional drivers since you wouldn't have to deal with high fixed costs.
|
|
#
?
Jan 26, 2017 20:36
|
|
|
Cicero posted:Huh, Elon Musk thinks self-driving cars will make traffic worse: What I love is that he thinks tunnels is the answer. Why? Why tunnels? It makes absolutely no sense and can only serve to be an expensive boondoogle that will simply increase congestion. He gets to the right conclusion with automated cars... they won't reduce congestion cuz it might induce demand. Yet, he doesn't connect the dots with expanding road capacity (which is what adding a tunnel would do). And since automated cars won't help with the regular roads, why would they with tunnels?
|
|
#
?
Jan 26, 2017 23:25
|
|
|
In cities space is a limited resource and they need to get better at getting paid for the use of it. If you want to bring your car all the way in you can pay up - or you can park outside the urban core and take a bus the rest of the way. There, we didn't even need to build any tunnels.
|
|
#
?
Jan 27, 2017 03:42
|
|
|
Hey look, actual data on how reliable different self-driving cars currently are: https://www.ft.com/content/77680d24-e8d7-11e6-967b-c88452263daf  at Bosch having to disengage more than once every mile. Although I suppose that probably happened to Google in the early days too.
|
|
#
?
Feb 3, 2017 19:57
|
|
|
|
| # ? May 5, 2024 23:57 |
|
|
Cicero posted:
I think it's less that they tried to do a good job and accidently did a bad job and more that they make one of a kind demo cars largely to show off specific components basically to pitch them to tesla. Like they basically get a barebones skeleton of hardware and software then throw a state of the art lidar then drive it around a little and generate some data with the hopes tesla goes "oh yeah, that lidar is actually way better than the one we were going to use, start supplying us with that". Like they are fine with it just barely working as long as it just barely works in a way where it's obvious their components are superior.
|
|
#
?
Feb 3, 2017 20:41
|
|