Lofty promises for autonomous cars unfulfilled
The first driverless cars were supposed to be deployed on the roads of American cities in 2019, but just a few days before the end of the year, the lofty promises of car manufacturers and Silicon Valley remain far from becoming reality.
Recent accidents, such as those involving Tesla cars equipped with Autopilot, a driver assistance software, have shown that "the technology is not ready," said Dan Albert, critic and author of the book "Are We There Yet?" on the history of the American automobile.
He questioned the optimistic sales pitch that autonomous cars would help reduce road deaths—40,000 every year in the United States, mostly due to human error—because these vehicles themselves have caused deaths.
As a result, self-driving maneuvers in the technology-laden vehicles are limited to parking, braking, starting or driving in a parking lot.
[...] "Automation may be used in areas such as closed campuses, where speeds are low and there is little or no interaction with other vehicles, pedestrians or cyclists or inclement weather," said Sam Abuelsamid, engineer and expert at Navigant Research.
The big problem is "perception": the software's ability to process data sent by the motion sensors to detect other vehicles, pedestrians, animals, cyclists or other objects, and then predict their likely actions and adapt accordingly, he said.
And that part is key, said Avideh Zakhor, engineering and computer science professor at the University of California-Berkeley.
"The perception part is not solved yet. The most advanced publicly available is 80-85 percent (reliable). That means that 15 percent of the time, it's going to hit objects and kill and destroy them," she said.
(Score: 3, Interesting) by The Mighty Buzzard on Monday December 16 2019, @04:30AM (2 children)
Is anyone actually surprised by this? Lofty promises pretty much always go unfulfilled.
My rights don't end where your fear begins.
(Score: 2) by c0lo on Monday December 16 2019, @04:59AM
Not me.
But I know many soylenters will be at least disappointed that human drivers, as imperfect as they may be, turned out to be still better at assessing and reacting to driving in real world conditions than neural network classifiers.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 3, Insightful) by driverless on Monday December 16 2019, @08:50AM
It's not just that in general, in the specific case of cars the manufacturers quickly grabbed all the low-hanging fruit, which was what led to amazing-seeming progress in the first couple of years when driving-assist went from "too risky to touch" to "let's do it anyway since Tesla seem to have gotten away with it". What's left now that the easy stuff is gone are the essentially insurmountable hard problems that are going to take years of chipping away at just to ameliorate slightly.
(Score: 2, Informative) by Anonymous Coward on Monday December 16 2019, @04:35AM
Dug back for some interesting self-driving car stories, Google found 300+ SN links with site:search so this has been a fairly popular topic. No particular selection criteria except that I looked for older hits:
Should a Robot Sacrifice Your Life to Save Two? https://soylentnews.org/article.pl?sid=14/05/13/1653250 [soylentnews.org] Tuesday May 13 2014, @06:49PM from the ethical-quandaries dept.
Uber Gives $5.5M to Make Amends for Stealing Carnegie Mellon's Self-Driving Car Scientists https://soylentnews.org/article.pl?sid=15/09/15/144233 [soylentnews.org] Wednesday September 16 2015, @12:57AM from the can-you-recruit-some-more-please? dept.
Self-Driving Cars Don't Need to be Perfect https://soylentnews.org/article.pl?sid=16/01/04/0138204 [soylentnews.org] Monday January 04 2016, @02:25AM from the how-long-could-they-last-in-Boston? dept.
Save the Driver or Save the Crowd? - The Moral Dilemma of Driverless Cars https://soylentnews.org/article.pl?sid=16/06/28/2114231 [soylentnews.org] Wednesday June 29 2016, @02:32AM from the always-be-ready-to-stop dept.
Michigan Senate Considering Ditching the Requirement of a Human Driver in Self-Driving Tests https://soylentnews.org/article.pl?sid=16/09/08/0144205 [soylentnews.org] Thursday September 08 2016, @07:21AM from the don't-forget-the-kill-switch dept.
(Score: 1, Interesting) by Anonymous Coward on Monday December 16 2019, @04:45AM
Maybe it's just the big players that are over-promising. Smaller startup operating in small areas, simple areas, office areas and real estate developments, but still public places is promising driverless cars in 2020:
https://arstechnica.com/cars/2019/10/how-self-driving-shuttles-could-enable-car-free-living-in-the-suburbs/ [arstechnica.com]
I think May Mobility may do it about the same time.
Deep Learning isn't solving all the world's problems like promises suggested, and it's not only cars where the promises are falling through. What has the newest, compute-intensive artificial intelligence really accomplished? It seems like just more (albeit _better_) classification. Granted, in the self-driving car arena, it can help classify stop-lights and cars and pedestrians, but it won't realistically develop the rules, obedience to the law, braking behaviors, or avoid throwing passengers from one side of a vehicle to the next on a turn.
These are systematic principles created by people, not by AI that solves all the world's problems. They're developed by iteration, testing, further iteration, learning from example, and so on. I'd suppose that Google, Apple, Cruise, Toyota and others have thrown enough deep-learning at this problem that if it hasn't been solved already it simply won't be solved by AI. (Really, what is there to develop if the AI gives you what-to-do at the next step? why isn't this done already?) Like everything, like the development of the LCD, like the development of computer processors, none of AI nor anything else will allow anyone to skip from step 1 to step 967.
It will all be incremental. Solve a small problem, then solve a slightly bigger one.
Even IBM's Deep Blue was meant to replace doctors (Watson). Diagnose cancer. Find new drugs. Largely, it hasn't. Even back in the 1990's, software was supposed to make treatment plans. Better your care. Allow everyone to live longer, healthier. Maybe, in fact, it's the press who is over-representing everything, or else heavily leaning on the most extreme offering of any spokesperson out there.
Lets assume: How long, with computers helping in design, development, and manufacturing, did it take to get CPU processors from 5um to 7nm? (How many process generations were skipped/advanced because of Deep Learning?) Then, how long might it take driverless cars to get from the light pedestrian areas to car-filled urban downtown? How long do _you_ think it will take for self driving cars to permeate society?
(Score: 2) by takyon on Monday December 16 2019, @04:47AM (13 children)
Incomplete partially driverless technology, criticized from the start for not being ready, is the most visible implementation and dragging down everyone else.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 2) by c0lo on Monday December 16 2019, @05:06AM (12 children)
Easy, tak, that's close to "blame X for burning down the idea of true communism" type of argument.
Better to take your time and first be sure the autonomous driving is actually achievable. 'Case, see? Been already... how many year and $zillion poured into such projects with little to show so far?
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by takyon on Monday December 16 2019, @05:22AM (3 children)
https://www.abc.net.au/news/2018-10-27/driverless-cars-promise-a-victoria-without-crashes/10428760 [abc.net.au]
It's not just the technology. Bureaucracy and roos have to be dealt with.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 3, Funny) by c0lo on Monday December 16 2019, @05:44AM
>blockquote>It's not just the technology. Bureaucracy and roos have to be dealt with.
Heh, you don't feel like listening to the lessons of the history, do you?
(grin)
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 3, Funny) by Anonymous Coward on Monday December 16 2019, @06:55AM (1 child)
Kangaroos have user stories too ..
https://www.erepublik.com/en/article/964834 [erepublik.com]
(Score: 2) by c0lo on Monday December 16 2019, @07:49AM
Yeap - that's Ozland.
The kangaroos will kill you with missiles, if the weather, geography, flora and the rest of fauna doesn't get you first.
The fact that kangaroos never get to fire missiles speaks volumes about the efficiency of the first lines of defense.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 0) by Anonymous Coward on Monday December 16 2019, @05:25AM (7 children)
Easy c0lo, that's close to "if we haven't done it yet, we never will" type of argument. There can be all sorts of technological advances or outright paradigm shifts that can enable many things. Otherwise, people would have given up on rockets to the moon, quantum computers, airplanes, etc. I mean how many years and $zillion were poured into such, until they finally worked and just improved from there?
(Score: 2) by c0lo on Monday December 16 2019, @05:39AM (6 children)
Close, but not exact. In any case, my assertion** is closer to reality than the position of "Blame Tesla for the current perception of over-promise and under-deliver"
---
** "More difficult that initial estimated" => "don't hold your breath, cause if you do, for sure it won't happen in your life-time. Otherwise, it may or may not happen in your life time, but that's because the difficulty of the problem, not because of the lack of trying or because some scapegoat".
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 0) by Anonymous Coward on Monday December 16 2019, @07:46AM (3 children)
That isn't my position, but mistaking me for the OP is understandable. Regardless, I also doubt I'll see it in my lifetime, but then again I am sort of old. I do think that my children will see limited use of self-driving cars. Things like follow truck convoys on controlled-access roadways to get around driving restrictions. Driving on uncontrolled city streets is probably a pipe dream even in my grandkids time, barring some sort of paradigm or other unforeseen change.
(Score: 2) by c0lo on Monday December 16 2019, @07:59AM (2 children)
If you look closer, I never attributed to you the position of the OP.
I just made my position clearer (I hope) - not Tesla but the problem's difficulty is to blame; given that, noone should expect a full/proper solution soon
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1, Informative) by Anonymous Coward on Monday December 16 2019, @10:29PM (1 child)
You are right. The denotation definitely didn't but the connotation did. At least for me. But it might also have something to with the fact my screen reader starts to sound sarcastic after awhile due to its flat tone.
(Score: 2) by c0lo on Monday December 16 2019, @10:55PM
If that's not quip material, I don't know what else is.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 4, Interesting) by MostCynical on Monday December 16 2019, @11:15AM (1 child)
Mr Musk is, primarily, as salesman.
Do all salesmen lie? Black swan territory for most of us - as I doubt any would have met any that don't lie (capabilities, speed, delivery date, availability, configurability, integration, cost... pick all of the list)
However, that one salesman has managed to make electric vehicles an available, viable (for the consumer) option for private transport is quite remarkable.
That he convinced his customers to also be beta testers for new features is even more. Many (most?) of the Tesla crashes seem to be from over-confidence (or just stupidity) on the part of the person in the driver's seat
If people didn't buy what salesmen sell, we'd probably all be subsistence farmers using rocks and blunt sticks to defend ourselves.
Hype wins - especially when something like the P100D S and X types exist, and have dragged all the world's manufacturers along with them.
"I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
(Score: 2) by c0lo on Monday December 16 2019, @02:34PM
I think you misspelled "engineers" (Thomas Watt and Richard Trevithick) and "chemists" (Fritz Haber and Carl Bosch) as the ones responsible to industrial and, respectively, agricultural revolutions.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by PiMuNu on Monday December 16 2019, @11:59AM (2 children)
It's a book promotion article (nicely timed for xmas!) so I don't like it, though I agree with the sentiment of TFS.
It is obvious that this technology will never work, for anyone who knows about multiparameter optimisation/etc. It's a shame that they overpromised so much, because the technology does have potential - but it needs infrastructure to support it i.e. signage that broadcasts its intent, cars that broadcast their location to nearby cars, etc.
It is extremely challenging to 1. identify that an object is a concrete block (and not e.g. a funny reflection in a puddle) and 2. decide on an appropriate evasion maneuver in real time. Far easier to read a signal that says "the road goes this way" and follow the signal.
(Score: 2) by HiThere on Monday December 16 2019, @07:14PM (1 child)
That's clearly not true, as human drivers offer an existence proof of a counter example. OTOH, that might well be the easiest way to do it. But I'm not sure. Depending on communication offered by others might open the door to various exploits that would be worse than implementing everything locally.
Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
(Score: 2) by PiMuNu on Wednesday December 18 2019, @11:15AM
> human drivers offer an existence proof of a counter example
This requires that human brains are equivalent to modern computers. I don't think this is true to any approximation.