Page 1 of 3 123 LastLast
Results 1 to 30 of 62

Thread: When will they come for the truckers?

  1. #1

    Default When will they come for the truckers?

    Fully autonomous vehicles are coming, soon. In the US, there are over 4 million professional drivers (people who drive trucks, buses, taxis etc for a living). Most of them seem to be truck-drivers (figures vary but one commonly quoted statistic is 3.5 million). If I'm not mistaken, most have at best a highschool-level education. Many are getting on in years.

    The nature of freight-trucking, esp. long haul, is such that many freight companies will find it economical to quickly phase out human drivers, whether they be employees or independent drivers. Average salary for a professional truck-driver in the us is > $60k (plus benefits?). The Tesla Semi, which will be ready for at least semi-autonomous driving from the start, is expected to cost $150-200k. A fully autonomous Tesla Semi isn't likely to cost much more when that capability finally arrives.

    When do you think fully autonomous vehicles will begin to provide significant competition for professional drivers?

    How quickly do you think professional drivers will be replaced?

    What will be the social consequences of this revolution?
    "One day, we shall die. All the other days, we shall live."

  2. #2
    In about 4-5 years I think serious competition will occur.

    I think it will start very slow as liability issues get settled out. A lot of companies are risk adverse.

    Social consequences? Not many.

  3. #3
    4-5 years seems optimistic to me, I think we're about a decade away but I expect it will definitely happen within our lifetime.

    Humans are the cause of more errors and crashes than machines so I expect it will happen ultimately but both the liability and confidence issues are big ones. There's also an issue that the drivers sometimes aren't just drivers but also eg pack and unpack their vehicle - something a driverless vehicle may struggle with. A business I used to work with got overnight deliveries, the drivers has keys to the building and the alarm code and would park up overnight, unpack the delivery into the fridge and then lock up again and continue their route - no vehicle would ever be expected to do that.

    Unions will also be an issue for many places - the technology for automated trains has existed for many decades (and trains are a lot simpler running as they do on tracks) yet despite that we still have Tube drivers on massive salaries. The Dockland Light Rail was built as a new thing driverless from the start but the Tube has never been switched over to being driverless as the drivers would go on strike to prevent it and it isn't possible to just switch overnight. New businesses and routes may find it easier to be driverless from the beginning than existing ones do - which in the private rather than public sector may result in some existing businesses going out of the business and being replaced by more economical competitors.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  4. #4
    People working on fully autonomous vehicles tend to underplay the potential for imminent disruption, offering frequent assurances that drivers will be needed for decades to come.

    Most analysts appear to believe the most significant obstacles will be regulatory and social. On the regulatory side, a transition is already underway, at least for semi-autonomous vehicles, and I don't believe regulation permitting fully-autonomous vehicles will be very far behind, at least not in the EU or the UK. In the US, with its fragmented legal system--and with trucking being the most important job in many states--it will likely take longer time. Wrt social obstacles, given the amount of protest that accompanied the rise of Uber, I think it's reasonable to expect a great deal of intense protests when the transition to fully-autonomous trucks begins. Trucking isn't unionized to the same extent many other sectors are, but I think there may be a push toward unionization in the next few years.

    The intervening period will, I believe, be relatively calm. During that time, semi-autonomous trucks will make life considerably more pleasant for drivers, both employed and independent ones. They'll be able to relax more, get more rest, do paperwork etc. Owners of electric trucks will make large savings on fuel. Of course, companies will likely respond to the arrival of electric trucks by cutting compensation to independent drivers.

    Given the successful early tests with long-haul platooning involving large caravans of semi-autonomous trucks, I think we'll see a rapid transition to driverless trucks under those circumstances, so long as the lead truck has a driver. Fully automated trucks are already in use outside of public roads, eg. mines and the like (see Rio Tinto for example).

    There are a number of technological problems left to solve, eg. dealing with a large variety of types of cargo and loading configurations, varied driving environments (going from highway to city driving for example, which may be more difficult with a huge heavily loaded truck than it is with cars) etc. For large logistics companies, I think the issue of loading and unloading will be quickly solved, considering how far companies like Amazon have already come with their experiments in warehouse automation.

    Ethical problems will be more difficult to address. When it comes to liability, a human driver may get a pass where a robot or its owner will be held to the highest standards with no room for error. A human driver may make a split-second decision to solve an immediate ethical dilemma (if it even registers as such) and humans will recognize the difficulty of acting appropriately in such a situation, whereas a robot making the same decision will be condemned.

    At the same time, I don't think the cost aspect of this liability problem is likely to be a deterrent, because of the immense savings from going driverless. I think conservative estimates of savings to industry are at around $150-200bn. On top of that, you have savings to society at large. In 2014, close to four thousand people died in trucking-related accidents in the US, with another 111 thousand injured. In the vast majority of these cases, driver error seems to have been either directly responsible or a significant contributor to the accident. Even if you only automate the trucks, these numbers may be reduced to the point where liability issues would not be a major deterrent.

    The savings and productivity gains are so immense that I think there will be a much greater impetus for changing regulations etc. The comparison to train-drivers, I think, is misguided, because of the magnitude of the expected savings.

    I think the social aspects should be given some serious thought. With millions of poorly-educated full-time professional drivers at risk of losing their livelihoods, we should expect social repercussions. Many of these will find it difficult to find new jobs, and those who do may see substantial losses in income. I would expect another wave of populism, with truckers becoming the new and more influential successors of the coal-miners and oil-rig workers. Higher proportion of trucking jobs correlated with a larger margin of victory for Trump, and unless the economies of those trucking-dependent states change very soon, I'd expect similar outcomes in future elections.

    Politics aside, there are other important social concerns, such as the prospect of hundreds of thousands of people ending up in poverty, or the high likelihood of increased suicide rates.
    "One day, we shall die. All the other days, we shall live."

  5. #5
    I saw the linked article in SciAm this morning. It's specific to AI art and music and only casually mentions transportation implications of AI, but really it's all the same thing. We don't even have strong AI yet and we're already facing automation of work beyond what one might have thought possible two decades ago.

    The issue is really about the Future of Work. The core of our civilization, every major culture, is all about work. What you do, how you do it, how much you earn for it is core to your identity. We work to live and live to work. Who doesn't work, why, and how they should be treated is a huge socio-political issue. What the hell happens when AI can easily do pretty much anything that people get paid to do today - and faster, with better quality, and virtually no labor cost?

    I remember mid-20th scifi that predicted robots doing all the physical labor for people and as a result humanity enjoying lives of leisure and wealth. No more poverty or blue collar drudgery. Ridiculous. The truth is when automation replaces workers, companies and governments pat them on the back and basically say "sucks to be you, find another job." Given the coming AI revolution, facing the real possibility of most workers being replaced with automation, how will our civilization deal with this economically? How will individuals deal with it psychologically -- if self-worth is measured in large part by the ability to provide for yourself with work, what happens when no work you do is valuable anymore?


    https://www.scientificamerican.com/a...ai-really-art/
    The Rules
    Copper- behave toward others to elicit treatment you would like (the manipulative rule)
    Gold- treat others how you would like them to treat you (the self regard rule)
    Platinum - treat others the way they would like to be treated (the PC rule)

  6. #6
    Quote Originally Posted by EyeKhan View Post
    I saw the linked article in SciAm this morning. It's specific to AI art and music and only casually mentions transportation implications of AI, but really it's all the same thing. We don't even have strong AI yet and we're already facing automation of work beyond what one might have thought possible two decades ago.

    The issue is really about the Future of Work. The core of our civilization, every major culture, is all about work. What you do, how you do it, how much you earn for it is core to your identity. We work to live and live to work. Who doesn't work, why, and how they should be treated is a huge socio-political issue. What the hell happens when AI can easily do pretty much anything that people get paid to do today - and faster, with better quality, and virtually no labor cost?

    I remember mid-20th scifi that predicted robots doing all the physical labor for people and as a result humanity enjoying lives of leisure and wealth. No more poverty or blue collar drudgery. Ridiculous. The truth is when automation replaces workers, companies and governments pat them on the back and basically say "sucks to be you, find another job." Given the coming AI revolution, facing the real possibility of most workers being replaced with automation, how will our civilization deal with this economically? How will individuals deal with it psychologically -- if self-worth is measured in large part by the ability to provide for yourself with work, what happens when no work you do is valuable anymore?


    https://www.scientificamerican.com/a...ai-really-art/
    You find new work. Same as we've been doing for three centuries of progress now.

    For an era where people are supposedly being put out of work at an unprecedented rate due to technology those nations with liberalised labour markets are keeping people in employment. The UK has at the moment despite AI, Brexit and everything else our highest ever employment rate. Never before in our entire history has a higher proportion of people been employed than at the moment.

    Similarly in the USA the unemployment rate is at one of its lowest rates ever at just 4.1%

    Even in the Eurozone unemployment is coming down and is "just" 8.7% which is low for them. Even France has got its unemployment rate below 10% and now stands at "just" 9.8% which is remarkably low for France in recent years.

    What drives people into unemployment isn't technology, it is socialism.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  7. #7
    Quote Originally Posted by RandBlade View Post
    You find new work. Same as we've been doing for three centuries of progress now.
    I see your conservativism reflex reacted to my characterization in "sucks to be you...". But if you read my post carefully, you'll understand it isn't about today, or the conditions workers are facing today. Find new work has always been a solution, but what happens when virtually all work is automated? Sounds impossibly ridiculous, I know, so I'll understand if you don't want to talk about that. But that's the topic.
    The Rules
    Copper- behave toward others to elicit treatment you would like (the manipulative rule)
    Gold- treat others how you would like them to treat you (the self regard rule)
    Platinum - treat others the way they would like to be treated (the PC rule)

  8. #8
    Quote Originally Posted by EyeKhan View Post
    I see your conservativism reflex reacted to my characterization in "sucks to be you...". But if you read my post carefully, you'll understand it isn't about today, or the conditions workers are facing today. Find new work has always been a solution, but what happens when virtually all work is automated? Sounds impossibly ridiculous, I know, so I'll understand if you don't want to talk about that. But that's the topic.
    That won't happen in our lifetime. I believe it is a valid concern in the far flung future but not in the short term. I'd wager to guess that the labor force participation rate won't drop below 50% in our lifetime despite many jobs today being automated. Just like improvements in farming led to whole new branches of employment we'll see the same here.

  9. #9
    Quote Originally Posted by EyeKhan View Post
    I see your conservativism reflex reacted to my characterization in "sucks to be you...". But if you read my post carefully, you'll understand it isn't about today, or the conditions workers are facing today. Find new work has always been a solution, but what happens when virtually all work is automated? Sounds impossibly ridiculous, I know, so I'll understand if you don't want to talk about that. But that's the topic.
    Virtually all work automated? It's a looooong time before the office-based work I do can be automated. When it can, artificial intelligence will be at or beyond current human thought-pattern levels. And once we're at that level, well, we'll be at some Banks-esque Culture type of semi-utopian society where no one has to work, everything that needs doing is done by artificial automatons of some sort, society is not monetary-based, and we're free to pursue fulfillment or hedonism or spiritualism or whatever without the shackles of needing to work to live. And I think that would be great.

    And mentioning the work I do, I have team-members and at the PhD-level working with Google scientists on their Machine-Learning product, developing it for the bank to train it in understanding patterns of payment behaviour to aid in the detection of financial crime. This is cutting-edge stuff, it's software capable of learning, but it's still pretty damn basic in the context of what we're discussing here. AI is centuries away from taking over the work of the middle-classes. We're only just reaching maturity on automation taking over the repetitive work of the working classes, where automative-intelligence at Step 1, 2, 3, Repeat is all that's needed.

    ~

    And on the subject of driving - we're not 5 or 10 years but many decades away from fully-automated cars being commonplace, and I'm not sure it will reach ubiquity in our lifetime. It only really becomes viable and safe when human drivers - and their inherent unpredictability - are removed entirely from the roads. That's a long way off.
    I watched a recent test on the UK roads of Tesla's Model X, which has the intelligence capable of staying in lanes on motorways, and of moving lane automatically should a slower car be encountered ahead - it detects the movement and speeds of all surrounding vehicles and has the intelligence to drive within that fluid and dynamic environment. Yet even in a brief test, it nearly caused an accident because a car approaching in a different lane accelerated suddenly, and the Tesla pulled out in front of it. Human unpredictability will always flummox and undermine even the best-trained AIs.

    There is inertia too. Driving is so central to our society, so much a part of human life on this planet currently, that changing that will be a sloooow process, even if the technology is capable and safe enough to remove the squishy unpredictable pink thing from the equation entirely. I for one don't want to stop driving.
    Last edited by Timbuk2; 02-15-2018 at 06:52 AM.
    Quote Originally Posted by Steely Glint View Post
    It's actually the original French billion, which is bi-million, which is a million to the power of 2. We adopted the word, and then they changed it, presumably as revenge for Crecy and Agincourt, and then the treasonous Americans adopted the new French usage and spread it all over the world. And now we have to use it.

    And that's Why I'm Voting Leave.

  10. #10
    I don't disagree, though it is probably a safe bet that some kinds of work will be automated faster than others, and faster than people think.

    My job is safe because much of it is based in lunacy that my craggy old corporation can't get out of because it's magnificently dysfunctional and entrenched bureaucracy won't allow it. And no, it isn't lost on me that this state of affairs might well spell the death of the company altogether. If a competitor of ours found a way to leverage something like AI to eliminate the jungle of chaos I live in every day in their company, mine would follow suit or die a sad, pitiful death. And I'm thinking that's going to be the way of it - when the technology reaches a point where some set of tasks is suddenly cheaply done by AI, one company will implement it and the dominoes will fall, one way or the other, throughout whatever industry pretty suddenly.

    It's the sudden part that is the fly in Rand's ointment. If too many people are in the bread line too quickly, chaos will ensue. And another fly is the quality of the new job - a big part of the problem in the US is the old coal miner / steel worker / factory worker job that paid pretty good goes away and the work those folks find pays less and/or has crappier benefits -- it isn't just about that unemployment rate. There are a lot of flies for the ointment, actually. Nothing is simple. But when AI really gets rolling, if those dominoes fall quickly, it will be chaos.
    The Rules
    Copper- behave toward others to elicit treatment you would like (the manipulative rule)
    Gold- treat others how you would like them to treat you (the self regard rule)
    Platinum - treat others the way they would like to be treated (the PC rule)

  11. #11
    I don't think you can necessarily extrapolate the current speed of development into something society changing like AI all the way into the future. Compare where the Internet was two decades ago (a niche pass-time for technically minded kids) to the way it's become a completely ubiquitous part of society over past decade in a way that was predicted by absolutely no one. There is an non-trivial chance of a similar explosion in AI.

    I mean, if I had to guess and place I bet I'd probably go with the AI pessimists but it's all pretty up in air as far as I'm concerned. We have no idea how far we can take it until we actually try and go there.

    Also, automation doesn't have to replace *all* forms of work to have major society changing consequences. Simply doing away with most low-payed, unskilled work will do that. If every job must be something an AI *can't* do then we're really going to have to up our game on education.
    When the sky above us fell
    We descended into hell
    Into kingdom come

  12. #12
    I wouldn't underestimate the social/cultural chaos that will happen in our lifetimes, mostly because the rate of change is faster than ever.

    I'm still amazed at the timeline: my father worked with huge data machines but still needed a secretary (I still have some of the IBM punch cards!), I had first-generation home computing and the internet but still use a CPA, my kids had hand-held game devices and on-line computer games but still needed a pizza delivery car. Now practically everyone has smartphones with apps that can replace secretaries, CPAs, and (experimental) drone delivery pizza. It's not just that "working class" or manual labor can be replaced with automation, but that even "professionals" like teachers, lawyers, pilots, or surgeons can be replaced by AI. That's not an eventuality, that's real time now.

    Anyway, I suspect it will be years before long haul truckers will be replaced, at least in the US. But due to logistics and topography -- not any union fight. There are reasons we use ships and barges, planes and trains, and 18 wheeled trucks: no one mode of transportation solves all our delivery challenges. We've automated and computerized parts of every sector, and that will continue, so long as it's efficient and cost-effective. We already have 'Positive Train Control' to reduce human error, but still freak out when ANY train derails (causing human injury, property damage, and/or spewing chemicals). The same will happen with trucking, in fits and starts.

    But that doesn't mean we should ignore the real disruptions to work and labor that *will* come with AI. It's really not enough to say "people will just find another job", or compare the 21st century to previous eras. Quality of life and standards of living, those expectations are also higher than ever.

  13. #13
    Quote Originally Posted by EyeKhan View Post
    I see your conservativism reflex reacted to my characterization in "sucks to be you...". But if you read my post carefully, you'll understand it isn't about today, or the conditions workers are facing today. Find new work has always been a solution, but what happens when virtually all work is automated? Sounds impossibly ridiculous, I know, so I'll understand if you don't want to talk about that. But that's the topic.
    I understand what you're saying and I'm saying it won't ever be the case. When old work goes human ingenuity finds new work. New ideas. Ideas that would once have been either viewed as either impossible or unnecessary.

    At one stage the vast majority of population had to work on basics like agrarian society in order to survive. Then that became largely unnecessary and now a much smaller proportion do that.
    Then we had increasing proportions working on manufacturing to produce things we'd like rather than need. Then that became largely unnecessary and now a much smaller proportion do that.
    Now we have increasing proportions working on services to provide services we want rather than goods we want or basics we need. As that becomes unnecessary we move on to something else.

    For every one person working in agriculture now in America there are more than six working in leisure and hospitality. A service industry that is essentially largely a luxury that would have once been unimaginable to be so large compared to agriculture.

    EDIT: More than a third of a million people across the EU now work in professional sport and directly-related activities, that's not even counting ancillary hospitality areas like bars where people go to watch televised sport. Do you think we're going to replace footballers with robots?
    Last edited by RandBlade; 02-16-2018 at 02:09 PM.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  14. #14
    Quote Originally Posted by RandBlade View Post
    EDIT: More than a third of a million people across the EU now work in professional sport and directly-related activities, that's not even counting ancillary hospitality areas like bars where people go to watch televised sport. Do you think we're going to replace footballers with robots?
    Given the net gain in wits and charisma, this is appealing.
    There's a man goin' 'round, takin' names
    And he decides who to free and who to blame

  15. #15
    LOL!

    Kind of goes to demonstrate the opportunities that open up to even those lacking in wits, charisma and dare I say it education.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  16. #16
    Quote Originally Posted by RandBlade View Post
    EDIT: More than a third of a million people across the EU now work in professional sport and directly-related activities, that's not even counting ancillary hospitality areas like bars where people go to watch televised sport. Do you think we're going to replace footballers with robots?
    Humanity reduced to playing games as "work." I know that's not what you're saying, but it was funny -- a scifi short story premise. I do think the bars can easily be automated, btw, though people may not like it, and it might be cheaper to use people.

    I think virtually any work can be replaced by AI. But, it won't make sense to replace some work because it's cheaper for people to do it vs the complex robotics/ facility redesign required for AI to do it. I'm thinking of stuff like hotel/ home housekeeping, facility maintenance work - plumbing, electrical, etc. And other work won't be replaced because people like it better when humans do it - or it's literally defined by humans doing it. Now thinking of athletes, performing arts, and the like. I think engineering and design could be done by AI well enough, but I wonder about laboratory research - real out of the box idea work. Any work with curiosity at its heart should be safe for a long time because an AI with curiosity would be an actual entity - strong AI - which isn't, IMO, coming soon. But none of this other stuff requires strong AI.
    The Rules
    Copper- behave toward others to elicit treatment you would like (the manipulative rule)
    Gold- treat others how you would like them to treat you (the self regard rule)
    Platinum - treat others the way they would like to be treated (the PC rule)

  17. #17
    Many jobs done now by people are done that way as a service, not because it is necessary.

    I read a while back that there are more people employed in manual car washing today than there were twenty years ago - and fewer automatic car washes today than there were twenty years ago.

    As far as automation goes its a retrograde step but many consumers have decided that while the technology exists to get their cars washed by a machine they'd prefer to have a human do it.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  18. #18
    maintenance is going to remain human focused for a while, but I don't see house keeping as a safe bet. I see that going away just as quickly as most healthcare positions. We already have machines that vacuum, mop, wash, dry and even fold laundry. Making and fluffing a bed is more an issue of method rather than ability.

    Quote Originally Posted by RandBlade View Post
    As far as automation goes its a retrograde step but many consumers have decided that while the technology exists to get their cars washed by a machine they'd prefer to have a human do it.
    This is yet another example of method rather than ability. Most manual car washing places still include a lot of automation. Automatic car washes were once 1 size fits all, but as AI progresses its going to come down to price and speed, which is why most places pay minimum wage even though its back breaking work.
    Last edited by Ominous Gamer; 02-16-2018 at 03:07 PM.
    "In a field where an overlooked bug could cost millions, you want people who will speak their minds, even if they’re sometimes obnoxious about it."

  19. #19
    Quote Originally Posted by Ominous Gamer View Post
    maintenance is going to remain human focused for a while, but I don't see house keeping as a safe bet. I see that going away just as quickly as most healthcare positions. We already have machines that vacuum, mop, wash, dry and even fold laundry. Making and fluffing a bed is more an issue of method rather than ability.
    I have a robot vacuum and it does a decent job, though I have to go over the corners and along the walls manually. I also have to do quite a lot of prep work before I run it - picking up any kind of clutter on the floor, moving cords, moving chairs around and so on. A home is a constantly changing environment with items of all different characteristics and priorities getting in the way of cleaning. For an AI to actually clean a whole house without constant supervision, the way the weekly cleaning lady does (who also requires clutter to be put away, btw, back when I had one) it will have to be smart enough to deal with that complexity.

    This is yet another example of method rather than ability. Most manual car washing places still include a lot of automation. Automatic car washes were once 1 size fits all, but as AI progresses its going to come down to price and speed, which is why most places pay minimum wage even though its back breaking work.
    I use an automatic. The machine isn't terrible, but they all could be a lot better -- it's an engineering design issue, not an issue with difficulty to automate. (IMHO)

    Also - emergency medical situations are incredibly chaotic and complex. AI can handle routine, for sure. But emergencies? Not for a long time...
    The Rules
    Copper- behave toward others to elicit treatment you would like (the manipulative rule)
    Gold- treat others how you would like them to treat you (the self regard rule)
    Platinum - treat others the way they would like to be treated (the PC rule)

  20. #20
    I've gone through several automatic vacuums. Their evolution over time has amazed me. I'm using a $60 knock off from monoprice right now that would blow my original roomba out of the water.

    The main problem I see with people who don't wash their own car is that physical employees can more easily see where dirt remains. Thats the biggest hurdle for automated machines, and as they evolve I think we will see fewer garage designs and more nibble multi piece designs. We already have dirt sensors built into our vacuums and washing machines, the tech is there. They will become more precise and less wasteful.

    I'm in no means suggesting that all medical staff are going to disappear, but automation will do to nursing staffs what its done to big box retail. The employees will monitor the hardware more than the patient. Allowing more rooms to be assigned to each employee. Just like target/walmart have 1 employee per 6 self checks. Hospitals will eventually get there as well. I'll even go out there and say service will improve greatly as we rely less on over worked staff missing checkup and drug administration schedules and reduce how much opening there is for human error. I'm curious what Aimless has to say on automation in the healthcare industry.

    EDIT: Forgot to mention how impressive AI has gotten at medical diagnosis too. Treatment plan for brain cancer in 10 minutes? Making a diagnosis for ailments that stump actual doctors? Having so much information immediately available is going to really change how the game is played. People are going to look back at shows like House (i know thats not real doctoring) and laugh about how shit used to be handled.
    Last edited by Ominous Gamer; 02-16-2018 at 07:59 PM.
    "In a field where an overlooked bug could cost millions, you want people who will speak their minds, even if they’re sometimes obnoxious about it."

  21. #21
    I've already warned my friends that they'll all soon be replaced by robots

    In all seriousness, as with most academic and/or white collar jobs, I expect AI to enhance healthcare more than anything else. This is a sensitive issue for some physicians, many of whom are inclined to encourage the creation of a veritable mythology around their skills and are reluctant to entertain any notions of machines and software encroaching on their territory (just as they were for the longest time reluctant to embrace systematically gathered and analyzed evidence from medical research). I don't expect doctors in any major specialty to be made redundant by AI in the next few decades. Even when the technology will be here, and even when we have compelling evidence, there will be significant regulatory barriers and costs/obstacles arising from licensing issues. And there will still be a need for people who can conduct or evaluate research for the foreseeable future.

    In some cases, we already know that computerized algorithms based on actuarial data can outperform even expert physicians when it comes to diagnosis and prognostication. You still need medical expertise for eliciting important diagnostic information, identifying the patient's problem, collaboratively identifying the best decisions for a given situation and patient, establishing a good relationship with another human being etc. AI will help healthcare professionals work better, make better decisions, become better at using complex and messy data to solve real-world problems.

    Some well-developed and promising applications for machine learning in medicine can be found in imaging, pathology and dermatology (Watson and its brethren). There are obviously challenges that arise from eg. differences in methods between different hospitals, but, overall, I believe machine learning can significantly reduce the workload for doctors working in those fields in addition to giving people still living in underserved, remote, largely depopulated regions greater and more convenient access to healthcare. What concerns me a little is that training future radiologists and pathologists may become increasingly difficult. Meanwhile, advances in those fields will likely lead to greater demand for doctors in other specialties. Some of my friends are currently among very few experts sufficiently proficient in a couple of highly specific tasks involving radiologic diagnostics and targeting that will soon be replicated by machine learning software in less than a decade, freeing them up to focus on other skills and allowing many other teams all over the world to do similar work. Healthcare also depends heavily on the availability and productivity of eg. biomedical analysts, lab-technicians etc, which has been greatly increased already by automation (freeing them up to do more qualified work). The more we can do, the more we do.

    When it comes to automation of physical skills, we are of course further away, but even today's medical robots (eg. robot-assisted PCI, orthopedic surgery, abdominal surgery) are impressive, and in many cases they add great value (eg. greater precision sustained for longer time, slightly lower risks to staff from radiation exposure, much less tissue damage during surgery, much more precise and less damaging treatment of tumours, etc). Some basic steps of the procedures performed using robots are going to be automated within five years. The ARTAS hair transplant robot--banal an application as that may be--is already semi-autonomous, and I don't think it's unreasonable to expect it to be able to be almost fully autonomous in a few years' time. Whether physicians will then hand over the task of overseeing the robot to skilled technicians is hard to tell. I think some routine ultrasound exams (already robofied) will also become at least semi-autonomous in less than a decade. Some of the tasks previously done by nursing assistants, such as measuring blood-pressure, are already being done using automated systems, for better or for worse.

    Putting aside the sci-fi stuff for a moment and looking at realistic near-term applications, autonomous or semi-autonomous systems for applications like neurological or physical rehabilitation in the home are just a few years away. Gamified self-learning software and robots, set up and introduced by physiotherapists & occupational therapists, will enable us to rehabilitate many more people much more effectively in their own homes while reducing the burden on those professions, freeing them up to focus on those patients that need more support. For these applications, I think the greatest hurdle will be high costs eg. due to cost of R&D, licensing, lock-in effects. In larger developed countries with universal healthcare, this sort of equipment may still be worth it. At my hospital, probably not for another ten years for robots, maybe five years for VR and other software AI-assisted cognitive behavioural therapy in the form of games and the like will also be commonplace, I think, within ten years (already a few promising applications, but obviously need to select patients carefully as with any other approach).

    Currently, my favourite example of automation in healthcare--and a clear example of automation making people redundant--is in the area of hospital logistics. There are a lot of comical videos of autonomous robots in high-tech hospitals transporting equipment & supplies all over the place, but underneath the silliness of these robots telling each other to move out of the way, there's a vision of our future. Ten years from now, I expect almost all of the transportation and a majority of the cleaning at our hospital to be facilitated by autonomous robots, freeing up staff to do more human things, like talking to people. New hospitals such as the scandalous one recently opened in Sthlm will obviously have fewer such staff to begin with, which is kind of a shame.
    "One day, we shall die. All the other days, we shall live."

  22. #22
    Not directed at anyone in particular:

    A lot of people seem to believe that as long as robots aren't capable of fully replacing them, their jobs are safe. But this isn't actually true. If a robot can do 10% of your job, companies are going to find ways to reclaim that 10%. They can cut jobs, cut wages, or expand a job's responsibilities. The last one isn't always going to be available, and so you should expect one of the other options to occur pretty frequently as automation is increased.

    We also don't need everyone to have their jobs automated away before we start seeing serious problems. The easiest jobs to automate away are the ones where the workers are least able to be retrained for other positions. The remaining job openings are going to be increasingly high skilled, and as that bar rises more and more of the population is going to slip under it. And as automation eats more jobs, there's going to be more pressure to increase the pace of automation to stay competitive. What's going to happen once 10% of the working population isn't suited for any existing jobs? What about when it rises to 15%?

    I don't know how long it's going to be before this turns into an economic crisis, but I keep adjusting my estimates downwards. I'd really like to start working out solutions now, but very few people in power seem to be taking this seriously (because hey, it's probably not going to happen this election term), and I worry that we won't be able to move on it until the crisis has already started.

    For the OP's question, I'd ballpark 5-8 years before truckers starting being replaced by machines, and 10-15 years before human truckers start becoming a novelty. Once it starts, I expect it to catch on relatively quickly - the economic incentives will just be too high. I don't know what the social consequences will be, but they will be consequential.

  23. #23
    I can easily see truckers being transitioned from long haul drivers to an on call service for that "last mile" within 5 years. Similar to how ports are handled, and if its handled in the same mafia like way, transportation as a whole will be fully automated ASAP. Whats going to suffer are all the towns and industries that rely on truckers. God knows how many places exist solely because its X miles between 2 other places.
    "In a field where an overlooked bug could cost millions, you want people who will speak their minds, even if they’re sometimes obnoxious about it."

  24. #24
    You're all assuming that won't be a massive Luddite backlash that stops much of the technological progress, even if it's at the cost of crippling one's own economy.
    Hope is the denial of reality

  25. #25
    Many people view this question through the lens of theories about the Luddite Fallacy. The empirical research to date suggests that, overall, increased automation tends to lead to more jobs and greater employment in the long term, up to the point of saturation for the market for products produced using increasingly automated products. Automation makes products cheaper, leading to increased demand for related jobs. Those who lose their jobs switch to other jobs. RB's post is reminiscent of this type of analysis. However, this view is, I believe, overly simplified, in that it treats all jobs, people and locations as interchangeable. As Chaloobi pointed out, automation may force people to switch not to more qualified jobs but to less qualified ones, with lower pay, fewer benefits, and, consequently, lower living standards. This is consistent with findings from empirical research and applies to the kind of jobs that are most likely to be impacted by automation. Other recent research suggests that subjecting jobs to automation by robots leads to job losses that aren't fully compensated for by job growth elsewhere in the same or nearby markets:

    https://economics.mit.edu/files/12763

    As robots and other computer-assisted technologies take over tasks previously performed
    by labor, there is increasing concern about the future of jobs and wages. We analyze the
    effect of the increase in industrial robot usage between 1990 and 2007 on US local labor
    markets. Using a model in which robots compete against human labor in the production of
    different tasks, we show that robots may reduce employment and wages, and that the local
    labor market effects of robots can be estimated by regressing the change in employment
    and wages on the exposure to robots in each local labor market—defined from the national
    penetration of robots into each industry and the local distribution of employment across
    industries. Using this approach, we estimate large and robust negative effects of robots
    on employment and wages across commuting zones. We bolster this evidence by showing
    that the commuting zones most exposed to robots in the post-1990 era do not exhibit any
    differential trends before 1990. The impact of robots is distinct from the impact of imports
    from China and Mexico, the decline of routine jobs, offshoring, other types of IT capital, and
    the total capital stock (in fact, exposure to robots is only weakly correlated with these other
    variables). According to our estimates, one more robot per thousand workers reduces the
    employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5
    percent.
    Taking truck-driving as an example, truckers may become forced to trade a reliable, well-paid job for a much less reliable, less well-paid one. A decade ago, a long haul truck-driver out of a job might consider eg. working on an oil rig. But new oil rigs are highly automated, with small barebones crews. While the number of rigs in use has rebounded after the recent downturn, half of the jobs lost have not returned--a couple of hundred thousand jobs.

    The hospitality sector is notorious for low wages and poor job-security. The gig industry even more so. Switching jobs may also force people to move long distances, whether from rural areas to urban ones with their attendant problems (eg. higher and steadily increasing cost of living), or from one city to another, putting significant strain on families and relationships and making transitioning to a new job much more difficult. Most people don't have good networks in many cities, and most people get jobs through contacts. While retraining for more qualified jobs may seem like an appealing solution, this comes with additional costs that cannot be borne by everyone, nor is it a realistic way forward for eg. people nearing retirement age. Will society subsidize retraining laid-off workers? With what money?
    "One day, we shall die. All the other days, we shall live."

  26. #26
    Quote Originally Posted by Loki View Post
    You're all assuming that won't be a massive Luddite backlash that stops much of the technological progress, even if it's at the cost of crippling one's own economy.
    That concern is in fact the reason for the thread. I don't believe, honestly, that there would be such a massive backlash in most Western countries. But, if it seems likely to occur, I think most Western countries would begin giving UBI much more thought.
    "One day, we shall die. All the other days, we shall live."

  27. #27
    Quote Originally Posted by Loki View Post
    You're all assuming that won't be a massive Luddite backlash that stops much of the technological progress, even if it's at the cost of crippling one's own economy.
    Hence why I mentioned unions.

    Quite frankly I don't see how trucking can survive due to Luddism. Luddism works best in unionised monopoly environments and there is no union monopoly here. Unless the law gets in the way (unlikely) economics will ensure this happens.
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

  28. #28
    Quote Originally Posted by Loki View Post
    You're all assuming that won't be a massive Luddite backlash that stops much of the technological progress, even if it's at the cost of crippling one's own economy.
    I don't think a ludite backlash could freeze technological development like that. Slow it down at the cost of weakening their own economy, maybe, but I don't think anyone can realistically completely stop the automation from happening.

  29. #29
    Quote Originally Posted by Wraith View Post
    I don't think a ludite backlash could freeze technological development like that. Slow it down at the cost of weakening their own economy, maybe, but I don't think anyone can realistically completely stop the automation from happening.
    Too much money.
    "One day, we shall die. All the other days, we shall live."

  30. #30
    Quite right
    Quote Originally Posted by Ominous Gamer View Post
    ℬeing upset is understandable, but be upset at yourself for poor planning, not at the world by acting like a spoiled bitch during an interview.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •