|
computer parts posted:Anything that requires a PE isn't going to be outsourced (because a *lot* of it is government work which requires citizenship) and there is a very high bar towards getting a PE which lots of people don't bother with if they don't need it. Once the cost is worth it, firms will lobby to lower restrictions like that PE requirement; that's my point. Attorneys must pass a local bar exam, but corporate lobbyists are starting to get exceptions written into state laws to allow legal outsourcing in various ways. It'll happen to engineers too, though it may take slightly longer. Ultimately an Indian engineer *can be* as good as an American one.
|
#
?
Nov 29, 2015 18:57
|
|
|
|
| # ? Jun 6, 2024 11:19 |
|
|
Hieronymous Alloy posted:Once the cost is worth it, firms will lobby to lower restrictions like that PE requirement; that's my point. They'd have to do it for every single state they operate in, because the PE is a state level program. In fact, why haven't they done so already?
|
|
#
?
Nov 29, 2015 19:00
|
|
|
computer parts posted:They'd have to do it for every single state they operate in, because the PE is a state level program. Good question. Maybe institutional inertia, maybe not worth it yet if foreign engineers can't compete on quality yet. But if they can't now they will be able to so compete soon. Just a matter of education and America is no special flower in that regard.
|
|
#
?
Nov 29, 2015 19:08
|
|
|
Hieronymous Alloy posted:Good question. Maybe institutional inertia, maybe not worth it yet if foreign engineers can't compete on quality yet. But if they can't now they will be able to so compete soon. Just a matter of education and America is no special flower in that regard. A matter of education is why China's jets keep falling off their aircraft carriers. It's about as major a thing as any other project you can name, especially given the current state of the system. There's very little evidence that (cheaper) foreign countries will be able to compete with the US, especially in the near time frame.
|
|
#
?
Nov 29, 2015 19:14
|
|
|
Hieronymous Alloy posted:Good question. Maybe institutional inertia, maybe not worth it yet if foreign engineers can't compete on quality yet. But if they can't now they will be able to so compete soon. Just a matter of education and America is no special flower in that regard. Maybe at some point in the future but both countries have massive issues to address before they'll be even close to having an educational system as reliable as the US. Can China/India produce quality engineers? Absolutely. Can they produce them in a certifiable manner? No where even close.
|
|
#
?
Nov 29, 2015 19:26
|
|
|
I think the real problem with automation is the labor shock it creates. Let's take job A, job A won't be automated "for years" it also employs a decent chunk of people. Suddenly Job A is feasible to automate. So within a year or 2 everyone doing job A isn't needed anymore. But for the previous 20 years job A was seen as "safe." That's where we are with every job. No one can seriously claim that computers will never replace X profession. And if the next job A that is automated causes too big a shock, something will have to give, and "undiscovering" a technological advancement has never happened so society will be forced to change, perhaps faster then it ever has before?
|
|
#
?
Nov 29, 2015 19:39
|
|
|
Powercrazy posted:I think the real problem with automation is the labor shock it creates. Let's take job A, job A won't be automated "for years" it also employs a decent chunk of people. Right, and there's a lot of people in A-type jobs who don't care about automation now because it's only lesser (paid) jobs that are currently threatened by automation. So it's understandable that people fear nothing will be done until things reach crisis levels.
|
|
#
?
Nov 29, 2015 21:55
|
|
|
Powercrazy posted:
Also, this doesn't happen in reality. It can take decades to roll out a system even after it's proven to be much better than the existing system simply due to the cost of transition. The diffusion of technology itself is a completely separate process from the invention of technology. computer parts fucked around with this message at 23:02 on Nov 29, 2015 |
|
#
?
Nov 29, 2015 22:24
|
|
|
INH5 posted:We've had the technology to fully automate planes for decades. It's long been a common joke among airline pilots that the standard flight crew will soon be reduced to a man and a dog: the man to feed the dog, and the dog to bite the man's hand when he tries to touch the controls. Yet the vast majority of planes still have pilots, even freight planes and small single engine planes. Single-engine planes are actually the most difficult to automate. They have the least payload capacity to devote to advanced equipment (and since the pilot is almost always a passenger anyway, the pilot is basically "free" in terms of weight). For example the full-fuel useful payload in a 4-seat Piper Archer II is 637 pounds - so avionics, 4 150lb people, and no baggage. Take another chunk off if it's warm out or you're at high altitude (thinner air => less lift). So if your autopilot system is more than 50 lbs or throws the center-of-gravity too far off (either of which is likely if it includes active components like a radar) then it's not really useful there. A 2-seat Cessna 152 is only 352 pounds useful payload, same deal. Also, they often fly into and out of small airports which lack the Instrument Landing Systems you need to guide a plane in without visual guidance from a pilot. Also they tend to have nonstandard features and approaches - for example an airstrip near me has 2 runways but one of them has a displaced threshold because there's huge trees 30 feet from the end of the runway, the other runway is too short to use for takeoffs (and you wouldn't even want to try landing on it), it's sometimes covered with geese (which will gently caress your poo poo up), and only the first one is (sometimes) plowed in winter. It's sort of like the flying equivalent of having a driver who knows that he needs to fade over the centerline a bit because there's a gigantic loving pothole that'll rip off the bottom of your car - we'll eventually be able to automate it but it's definitely tougher than a major airport that can offer optimal conditions. And all the smart systems that could figure out if it's unplowed or covered in geese will add weight. Your conclusion is basically right though, we will basically have pilots/drivers in trucks and airliners to handle autopilot/autodriver failures and the increasingly infrequent chances when a situation is too unexpected/extreme for the autopilot to handle. Especially lately the trend has been towards paying regional airline pilots so little that they need a second job to make ends meet, which lead to pilot fatigue. Also airline pilots spend so much time on autopilot (i.e. so little time actually flying the plane) that they can't really take over in an emergency when needed. So really we're headed for an I, Robot (the movie) type situation where the joke about the dog biting the pilot is actually true - pilots are actually neither needed, nor really competent to fly the plane anymore, and you should really be looking at the guy flying manually with some suspicion. Paul MaudDib fucked around with this message at 23:17 on Nov 29, 2015 |
|
#
?
Nov 29, 2015 22:44
|
|
|
Also planes are actually easier to automate. An autopilot flying a plane at 30,000 ft could malfunction and do almost literally anything in the 10's of seconds it might take the pilots to take over and they'll be able to fix it. There is lots of time and nothing to crash into. Landing and takeoffs are higher stakes but also assisted by ground control. Meanwhile if you're driving in a car at speed you can routinely be about 1 second from potential death due to oncoming traffic or solid objects on the side of the road. There is no margin for error in an automated car because it can kill you before you have a chance to realistically take back over. Worse, the the visual cues needed drive a car on the road are orders of magnitude more complex than most inputs to a plane autopilot (radar processing and other newer systems are complex, but usually for additional detection, not routine navigation/control).
|
|
#
?
Nov 30, 2015 01:33
|
|
|
Paul MaudDib posted:Your conclusion is basically right though, we will basically have pilots/drivers in trucks and airliners to handle autopilot/autodriver failures and the increasingly infrequent chances when a situation is too unexpected/extreme for the autopilot to handle. Right, and this is actually a much more worrying situation than simply killing those jobs outright since the only possible conclusion is a continuation of wage erosion. Nevermind the fact that we're saying that the most socially useful purpose we can find for someone is sitting, alone, behind the wheel of a vehicle that will rarely if ever need their input.
|
|
#
?
Nov 30, 2015 03:49
|
|
|
asdf32 posted:Also planes are actually easier to automate. Yeah, most instrument navigation is along the lines of "set bearing W from VOR C, continue for Y miles, then switch to bearing Z on the outbound leg, and acquire a new VOR at frequency Q". So overall each leg is about as complex as a single turn along a surface street, but a "street" might be 25-80 miles long. GPS navigation is the new norm but you can't depend on it since it can go down during solar flares and so on. You still don't want to get lost, because unlike a car you can run out of fuel and crash, but the reaction times are a lot less precise during routine navigation. I'd imagine it's probably comparable to automatic train control. To fly commercial at all you need to be capable of flying instruments should the need arise, and you can't land on instruments just anywhere. It takes special equipment both on the plane (common) and on the ground (less common). Much like a car, it's a situation where a second or two of error can kill you absolutely dead. However the inputs are relatively straightforward - autoland is pretty much a solved system. ILS systems are relatively common - my local municipal airport in a town of 125k has one - but a lot of airports that small aircraft fly from do not.
|
|
#
?
Nov 30, 2015 04:18
|
|
|
DrSunshine posted:Either way, it seems to me that technological unemployment is becoming a serious issue, one that merits debate and awareness. I am highly suspicious of the conclusions of the liberal and utopian (Singularitarian) authors who've written on this subject -- it doesn't seem likely to me that capitalism can continue to exist if steadily growing numbers of people are being made unemployed. Capitalism as it currently exists cannot continue. The philosophy that everyone must work can continue if the lives of the majority are devalued enough because of their inability to work. In that scenario, the majority of people on the planet would be sterilized so that in a few generations the only humans alive work highly skilled technical positions that robots are incapable of, or jobs that are valued by humans for subjective reasons. I've met someone who is very excited about automation and he hopes that will happen. In reality, there aren't really enough sociopaths around for that to seem likely. A utopian socialist society also seems a long way off. Trying to predict how humans will react as this problem grows is very difficult, but I imagine as the permanently unemployed grow from a large underclass to a majority of the western population, it won't be possible to maintain current attitudes about labour and human value being inextricably linked.
|
|
#
?
Nov 30, 2015 05:35
|
|
|
Weldon Pemberton posted:Capitalism as it currently exists cannot continue. The philosophy that everyone must work can continue if the lives of the majority are devalued enough because of their inability to work. In that scenario, the majority of people on the planet would be sterilized so that in a few generations the only humans alive work highly skilled technical positions that robots are incapable of, or jobs that are valued by humans for subjective reasons. Honestly this creepy obsession with sterilizing poor people is the weirdest loving thing. computer parts posted:Also, this doesn't happen in reality. It can take decades to roll out a system even after it's proven to be much better than the existing system simply due to the cost of transition. The diffusion of technology itself is a completely separate process from the invention of technology.
|
|
#
?
Nov 30, 2015 06:23
|
|
|
computer parts posted:Also, this doesn't happen in reality. It can take decades to roll out a system even after it's proven to be much better than the existing system simply due to the cost of transition. The diffusion of technology itself is a completely separate process from the invention of technology. But it is perceived to happen this way. People cling to "this is a dependable job" for a very long time. Look, nobody know when a ship or boat is going to capsize either. It's a predictable and basically solved problem that follows defined rules. Still catches people by surprise. Nobody pays attention until the automation is actually being rolled out. I mean you have your outliers who see it coming but most people don't. Nobody wants to look at the end of what they do.
|
|
#
?
Nov 30, 2015 07:55
|
|
|
Nessus posted:Honestly this creepy obsession with sterilizing poor people is the weirdest loving thing. My, admittedly, broad opinion: There is a less than zero amount of people who internalize poverty as happening at a genetic, mental, or even spiritual level, even if they don't really realize it. It goes hand-in-hand with behavior that, for a hypothetical, tends to judge people on scales of fluffy, non-tangible bullshit like the concept of karma or soul decay. Poverty is seen as making up the core/soul of the person, not a person in a hosed up situation. On the other hand: some people know this and honestly want other people out of their class to die from some weird idea of a dog-eat-dog personal resource race. There's zillions of reasons why people are like this, Nessus. :/ Edit: Automation edit. They're looking to replace one of my cash registers with a self-service iPad. So, there goes four people they can safely lay off when implemented! They (my district bosses) were elated to tell us this with a region-wide e-mail on Black Friday. We got an awkward email sent again a few hours later out with one of the guys wondering why no one in the district has emailed back. We still haven't replied, we're never going to. We'll believe it when we see it. We're still waiting on three-month old orders for Toilet Paper shipments. Armani fucked around with this message at 08:20 on Nov 30, 2015 |
|
#
?
Nov 30, 2015 08:12
|
|
|
Helsing posted:Automation is a good thing provided we have the right institutions in place to deal with the disruptive social impact. Right now we don't and as a result automation is creating serious political and economic problems, and there's every reason to think that these problems will get worse before they get better. Not just institutions, but I'd argue a change in belief and culture in what "employment" should look like. Maybe it's not the best idea to maximize employment for the sake of being in a bullshit job but rather having a minimum standard of living (mincome?) that allows you to pursue your passions and/or raise a family and other life concerns irrespective of a job market. AKA what this person is saying: LookingGodIntheEye posted:Removing the general populace from pointless labor they don't actually like and allowing them to focus on entirely creative and/or intellectual pursuits could lead to a veritable renaissance. ---- Main Paineframe posted:Technology does not "grow" exponentially, nor does it "grow" in a straight line. To be honest, it's incredibly naive to try to boil down the advancement of all human knowledge and technology to a line on a bar graph. But economists do this all the time?  Teriyaki Koinku fucked around with this message at 19:23 on Nov 30, 2015 |
|
#
?
Nov 30, 2015 17:01
|
|
|
Don't worry, we're going to have another dark age any day now. Get ready to thank ISIS for that, most likely.
|
|
#
?
Nov 30, 2015 21:28
|
|
|
mobby_6kl posted:Don't worry, we're going to have another dark age any day now. I don't know what I like more, the numberless metric of "scientific enlightenment" or the apparent non-existence of all those civilizations that had never even heard of Christianity in that time period. Can you ask your racist grandpa if he has any more cool memes? Maybe something with the minions next time?
|
|
#
?
Dec 1, 2015 03:54
|
|
|
DrSunshine posted:However, I don't think that Russell makes any point that any of us would disagree with. I have yet to see anyone in this thread praise a 8-12 hour workday, 40 hour workweek for the moral virtue that it instills. Most of us, I feel, would enjoy greater leisure time to spend arguing on the internet, or other hobbies. I disagree with some of the specific points he makes, but agree with the gist that our society overvalues work and undervalues leisure. For example, he says "the man who lends his money to a Government is in the same position as the bad men in Shakespeare who hire murderers," when the Government accomplishes much more, such as building roads, than murder people in foreign lands. Despite some specifics, It's a great essay and it altered my perspective. We should take care not to draw too many normative claims from his work. Some read his meaning as leisure is good because it will further the arts and sciences, but leisure holds value regardless. quote:When I suggest that working hours should be reduced to four, I am not meaning to imply that all the remaining time should necessarily be spent in pure frivolity. i mean that four hours' work a day should entitle a man to the necessities and elementary comforts of life, and that the rest of his time should be his to use as he might see fit. After paying your due, you earn your freedom. Freedom to advance the sciences, but freedom to binge on Netflix too. Anyway, I really enjoyed this. Thanks to Hieronymous Alloy for pointing it out. Technological automation should increase leisure time for all, but, as the article in the OP laments, it will probably only increase it for the very few.
|
|
#
?
Dec 1, 2015 15:22
|
|
|
mobby_6kl posted:Don't worry, we're going to have another dark age any day now. We need a :hist99:, Jesus Christ. Everything about this is wrong. Nothing is right.
|
|
#
?
Dec 2, 2015 16:29
|
|
|
Woolie Wool posted:We need a :hist99:, Jesus Christ. Everything about this is wrong. Nothing is right. I'm not sure which graph I like more, Gun Chart, or Dark Ages.
|
|
#
?
Dec 2, 2015 16:38
|
|
|
mobby_6kl posted:Don't worry, we're going to have another dark age any day now. Leading scholars agree 
|
|
#
?
Dec 2, 2015 18:25
|
|
|
steinrokkan posted:Leading scholars agree lmfao yes transhumanism will take us from the 1950s to the 2000s again 
|
|
#
?
Dec 2, 2015 20:27
|
|
|
Nessus posted:Well I think that's the "safe" thing. It "can" take decades. Meanwhile some other position gets automated and the people in it start fleeing around. This puts some wage pressure on your industry - and suddenly perhaps you're the one on the chopping block. Or perhaps your "safe" job starts getting cut because it has fewer customers. Can you post specific and verified examples of this happening in reality?
|
|
#
?
Dec 3, 2015 07:48
|
|
|
I knew of a local business that made/sold paper products. Year after year, there seemed to be less customer demand for printed calendars and graphical design work. Management tried everything, even sourcing cheaper paper and ink from overseas! Alas, nothing worked and the place eventually folded, resulting in many lost salesman and printing tech jobs. Alternate ending: this story takes place in the 18th century and is actually about Benjamin Franklin not being able to sell enough newspapers and bibles.
|
|
#
?
Dec 3, 2015 09:19
|
|
|
Those people saying how driverless cars can do only highways and other simple situations and are still decades away, should watch this TED talk. http://www.ted.com/talks/chris_urmson_how_a_driverless_car_sees_the_road The demonstration starts at 7:45. The car reacts safely to a bunch of stuff like other vehicles blowing through red lights. Also, a woman in a wheel chair chasing a duck.
|
|
#
?
Dec 8, 2015 23:35
|
|
|
Doctor Malaver posted:Those people saying how driverless cars can do only highways and other simple situations and are still decades away, should watch this TED talk. This guy does a great job: 1) Outlining the difference between driver asistance and self-driving (it's huge) 2) Outlining the complexity inherent to driving on real roads (bikers giving hand signals, traffic cops) But the optimism here is misguided in two ways. First, we could have a completely safe self-driving car today and it would still take probably a decade of testing before it got accepted in the market and another decade before it was produced in relevant numbers. The game here isn't just to make this thing, it's to get it deployed. Second, he's naively thinking the end-game is a car that's statistically safer than humans. That's now how this works. Society, for whatever reason accepts high accident rates for human drivers. They're not going to accept those rates from corporate vehicles. Before wide-scale adoption these vehicles are going to have to bend over backwards to prove that they're significantly safer than humans with huge liability consequences if they're not. And any compromises, like no driving in snow or at night in rain won't really be tolerated. Finally I'll add that this guy is right when it comes to driver assistance. Where Tesla and Mercedes are going is largely going to be a dead end. Half-safe pseudo automation isn't going to be acceptable either from a safety or liability point of view.
|
|
#
?
Dec 9, 2015 01:32
|
|
|
I think we all agree that there are different types of driving, and that the solutions for one type don't easily apply in the others. For example, navigating through city streets requires a totally different approach than driving from Phoenix to Vegas, which is also totally different from driving on dirt roads in a snowstorm. Not all of the problems have to be solved at once. If city driving at 35mph is solved, it could result in cars aimed at senior citizens and disabled people. People will like the idea of getting 85 year olds out from behind the wheel. These things aren't gonna be on the freeway. Or maybe they get a handle on interstate driving first. If the weather is good, a shipping company could have human drivers transport the trucks from the warehouse to a depot outside of town, get out, and send the truck 1000 miles down the road unattended to another depot where a second driver finishes the job. Or maybe freeway driving is the focus. A human driver navigates on city streets, but once you hit the on ramp, autopilot kicks in.
|
|
#
?
Dec 9, 2015 14:52
|
|
|
Autonomous cars will initially be a good deal more expensive than "classic" cars so the first widespread adoption will almost certainly be commercial. Making the leap and plopping down and extra 5-10k for a car is not a trivial investment. There's autonomous trucks operating in pit mines and some airports have shuttles running on designated tracks. The next is probably going to be freight and passenger transport on harbor terminals and airports. There will be a substantial incentive for long haul trucking to follow - a truck that can drive 24/7 without taking into account rest periods and driver downtime will be valuable and can easily pay off the extra cost. That's where it will be tested and perfected.
|
|
#
?
Dec 9, 2015 15:15
|
|
|
It seems many people are in agreement that automation in most knowledge-sector jobs and some manual service jobs is inevitable. And a lot of people have advanced the opinion that "something" must be done, but what recommendations are sensible, exactly? I've read mention of something like raising the minimum wage, strengthening unions, and other measures to protect workers in industries. What other policies would help society transition? In The Second Machine Age, MIT economists Brynjolfsson and McAfeee end with a series of policy recommendations which include: increasing education, creating a guaranteed minimum income, creating appropriate taxes, and increasing entrepreneurship. Are these sufficient to mitigate the demand shocks caused by increased unemployment?
|
|
#
?
Dec 9, 2015 18:19
|
|
|
DrSunshine posted:It seems many people are in agreement that automation in most knowledge-sector jobs and some manual service jobs is inevitable. I'm assuming you're using a strange definition of "knowledge sector".
|
|
#
?
Dec 9, 2015 18:24
|
|
|
computer parts posted:I'm assuming you're using a strange definition of "knowledge sector". Well, I guess I mean things like legal discovery, clerical tasks, etc. I guess that's just "white collar"?
|
|
#
?
Dec 9, 2015 19:23
|
|
|
asdf32 posted:Second, he's naively thinking the end-game is a car that's statistically safer than humans. That's now how this works. Society, for whatever reason accepts high accident rates for human drivers. They're not going to accept those rates from corporate vehicles. Before wide-scale adoption these vehicles are going to have to bend over backwards to prove that they're significantly safer than humans with huge liability consequences if they're not. And any compromises, like no driving in snow or at night in rain won't really be tolerated. If by "statistically" you mean "measurably" then you're right and I don't see what's naive about that. If you mean "marginally" then I don't know where you picked that up. Can you point to the place in video where he says that? Dr. Arbitrary posted:I think we all agree that there are different types of driving, and that the solutions for one type don't easily apply in the others. For example, navigating through city streets requires a totally different approach than driving from Phoenix to Vegas, which is also totally different from driving on dirt roads in a snowstorm. I don't agree that types of driving present a meaningful difference to the machine (OK a dirt road is another thing, the AI car probably won't see them as part of navigable system). You attribute human flaws to electronic brain. For me and you when we were learning to drive, a long straight road was easier to navigate than a busy intersection because it gave us fewer things to keep in mind at the same time. A machine that knows how to treat individual elements in traffic won't have problems with many such elements appearing at the same time. The difference between dealing with one truck vs dealing with 2 trucks + 1 scooter + 6 pedestrians + 4 cars + yellow light + night + ice on the road will be negligible. Like a calculator that can calculate 6511.18 x 902.77 as easily as 7+8. Anosmoman posted:Autonomous cars will initially be a good deal more expensive than "classic" cars so the first widespread adoption will almost certainly be commercial. Making the leap and plopping down and extra 5-10k for a car is not a trivial investment. There's autonomous trucks operating in pit mines and some airports have shuttles running on designated tracks. The next is probably going to be freight and passenger transport on harbor terminals and airports. There will be a substantial incentive for long haul trucking to follow - a truck that can drive 24/7 without taking into account rest periods and driver downtime will be valuable and can easily pay off the extra cost. That's where it will be tested and perfected. You are right, except in the part where you assume that automated harbor terminals don't already exist.  https://www.youtube.com/watch?v=WxXZQ7emHC0
|
|
#
?
Dec 10, 2015 10:49
|
|
|
Once doctors and lawyers start being replaced and people realize even education cannot protect you from robots im sure we will start to work it out, unlike labor they have more wealth and pull.
|
|
#
?
Dec 10, 2015 13:21
|
|
|
DrSunshine posted:Well, I guess I mean things like legal discovery, clerical tasks, etc. I guess that's just "white collar"? If anything I'd say most white collar jobs are safe because a lot of them involve interacting with people and machines are very bad at doing that. Unless by "automation" you mean "has a computerized version" ala tech support, in which case I'll point you to companies going back to manned tech support because again people hate interacting with computers.
|
|
#
?
Dec 10, 2015 14:55
|
|
|
computer parts posted:Unless by "automation" you mean "has a computerized version" ala tech support, in which case I'll point you to companies going back to manned tech support because again people hate interacting with computers. Not really. For some tasks it comes down to nothing but convenience and cost - nobody cares that your bank teller has been turned into an ATM or the toll booth operator turned into an automated kiosk. People are increasingly shopping online and are perfectly happy to forego the interaction with salespeople simply because then they don't have to get off the couch. People have computers glued to their hands these days so saying they don't like interacting with them is a little too broad. There's things you probably can't automate. I fully expect to see automated fast food joints because the 30 second interaction with a random stressed out teenager in the drivethrough isn't particularly interesting or exciting - but the waiter at a real restaurant is probably different. People want to be waited on and made to feel special. That interaction is about selling a feeling and comfort - at McDonalds they just want to sell you cheap hamburgers. Tasks that are about accomplishing Thing as cheaply and quickly as possible will be automated as much as possible. Conversely Apple, Nespresso and other stores that sell curated experiences will never be automated because the human connection is the only reason it exists. Incidentally I expect there to be a growing number of brands and by extension stores like that.
|
|
#
?
Dec 10, 2015 15:45
|
|
|
Anosmoman posted:Not really. For some tasks it comes down to nothing but convenience and cost - nobody cares that your bank teller has been turned into an ATM or the toll booth operator turned into an automated kiosk. People are increasingly shopping online and are perfectly happy to forego the interaction with salespeople simply because then they don't have to get off the couch. People have computers glued to their hands these days so saying they don't like interacting with them is a little too broad. Both of those examples rely on relatively simple operations. Anything that is going to require extensive communication is not. As an example - you have an issue, and call up the customer service department. A person is able to get your request, and (usually) at least be able to transfer you to the proper department. With an equivalent automated line, you need to either know where you want to go before hand or have the system list out the options for you to choose. That's not very pleasant. Oh, and on a sidenote it's a very strange definition you're using to be calling a toll booth operator a "white collar" position. computer parts fucked around with this message at 15:57 on Dec 10, 2015 |
|
#
?
Dec 10, 2015 15:55
|
|
|
Anosmoman posted:at McDonalds they just want to sell you cheap hamburgers. Most customers do want to buy cheap hamburgers, but anyone who has worked in retail or fast food or call center can tell you, there's a small, but noticeable market segment that really likes the human interaction. Sure, you can buy a hamburger at a kiosk, forget to press the no-pickles button and then get a hamburger with pickles, but with a human being, you can call them an idiot and loudly exclaim that fast food workers want $15 an hour and can't even figure out how to make a loving hamburger correctly. And when you're dealing with a billing problem on the phone, you can tell the IVR that you hope it gets raped, but it's just not the same as when there's a real live human being on the other end. There's untapped potential here, a business model that can never be automated!
|
|
#
?
Dec 10, 2015 16:53
|
|
|
|
| # ? Jun 6, 2024 11:19 |
|
|
Doctor Malaver posted:I don't agree that types of driving present a meaningful difference to the machine (OK a dirt road is another thing, the AI car probably won't see them as part of navigable system). You attribute human flaws to electronic brain. For me and you when we were learning to drive, a long straight road was easier to navigate than a busy intersection because it gave us fewer things to keep in mind at the same time. A machine that knows how to treat individual elements in traffic won't have problems with many such elements appearing at the same time. The difference between dealing with one truck vs dealing with 2 trucks + 1 scooter + 6 pedestrians + 4 cars + yellow light + night + ice on the road will be negligible. Like a calculator that can calculate 6511.18 x 902.77 as easily as 7+8. "Stop if an object is in your path" is not a particularly impressive programming task, nor is it a representative showing of the potential difficulties involved in self-driving cars. Once you have sensors capable of clearly distinguishing objects in the surrounding space, it's relatively easy. That's why the "snowstorm" was a particularly important part of his post that you glossed over - whether it's a calculator or a self-driving car, it can only do its job properly if it is getting clear and reliable input that the processing unit can make sense of. Try pushing the buttons on the calculator with a beach ball and you might find it becoming a whole lot harder to get it to accurately perform the calculations you want. Likewise, poor weather conditions or unusual road conditions may impact the ability of the sensors to provide clear input to the processing unit, rendering it unable to accurately respond to its surroundings. The inability of a self-driving car to tell what am object actually is can cause serious problems, like slamming on the brakes in 70mph highway traffic because a plastic bag caught by the wind drifted across its path. Anosmoman posted:Not really. For some tasks it comes down to nothing but convenience and cost - nobody cares that your bank teller has been turned into an ATM or the toll booth operator turned into an automated kiosk. People are increasingly shopping online and are perfectly happy to forego the interaction with salespeople simply because then they don't have to get off the couch. People have computers glued to their hands these days so saying they don't like interacting with them is a little too broad. As long as poo poo goes right, people are fine with pushing a button to make things happen with no human interaction. But when things go wrong, people want to talk to a human about it. If you press the button that says "cheeseburger" and a cheeseburger comes out, you're happy. But if you pressed that button and a salad came out, you're going to be demanding the attention of a real human to get that problem fixed - the closer, the better.
|
|
#
?
Dec 10, 2015 21:18
|
|