I was listening to Andrew Yang on the JRE podcast a few days ago and I found out that his entire political platform is built on automation taking away manual jobs, specifically truck driving jobs.
Truck driving jobs aren't going anywhere, not in 10 years, not in a 100 . Neither is that of plumbers, electricians or many such skilled labors. Most programmers will lose their jobs before any of that happens.
Indeed. I think jobs like plumber, electrician, etc are actually the safest there are in term of automation.
But self-driving trucks, especially long-haul ones that drive mostly on motorways between warehouses are likely near the top of the list of replacement by self-driving vehicles.
Agreed. Plumbers, electricians, and other complicated/messy jobs will probably see some minor efficiency gains, but won’t be replaced by automation in any foreseeable future I can imagine.
I'm not sure why society getting poorer means fewer plumbers and electricians will be needed...
Provided you're not actually talking about poor people just dying off in large numbers, we'll still have roughly the same number of households, which means the same number of sinks to unclog and outlets to wire. I suppose it might be true that fewer people will be able to afford the services of skilled professionals, but that's not at all the same as saying that fewer will be needed...
> TuSimple is hardly the only company working to making fully automated shipping a reality. Several companies, including Aurora, Daimler, and Embark Trucks, are competing for a slice of the future of self-driving freight trucks. Alphabet-owned Waymo confirmed on Tuesday that it will be expanding its own self-driving trucking routes throughout the American Southwest and Texas, following previous tests in Arizona, California, Michigan, and Georgia
Gotcha, sounds like a miscommunication. I meant real as in, people are working on it and it’s not vaporware. In the same way that I don’t see autonomous plumbing companies when I search.
I actually see plumbers getting highly specialized as the infra gets more sohpisticated.
Automation in that regard is probably a later iteration of the automated household where now we have auxillary vacuums, embedded iot light controllers, and so forth.
One can also search for lizard people and alien takeover which may seem just as "real" to some, but fully automated trucks that will replace drivers is not happening.
Lmfao, long haul drivers are very much on the chopping block this decade. Every major logistics corporation is lining up to snap up self driving trucks between warehouses.
Short-haul/local drivers however yes, they will continue to exist for now due to the additional complexity preseneted by everyone and their mom's loading dock being all kinds of differently designed, often in ridiculous ways.
I’ve been hearing about self driving cars since I was a kid a few decades ago.
I have yet to see any evidence of a self driving vehicle that can handle anything other than the simple roads of sunny AZ or CA.
Show me one that can handle a snow squall or heavy rain or missing signs and lane markers and construction zones, and I might think it’s possible this decade.
Otherwise, I assume there’s a lot of work to be done.
> meaning that they’re capable of performing all driving functions autonomously under certain conditions, and they drive on public roads mostly during weekdays (both during the day and at night) with occasional testing on weekends “only during favorable weather conditions.”
Since they don’t advertise being able to drive in less than ideal weather, I assume it means they can’t yet.
I don’t want to be pedantic but you said:
> I have yet to see any evidence of a self driving vehicle that can handle anything other than the simple roads of sunny AZ or CA.
I’m not disagreeing about weather constraints in the ODD. But I don’t think anyone who’s driven in Pittsburgh thinks it has “simple roads” like Chandler, AZ. I should have been more explicit I was only responding to that aspect.
Yeah, that was news to me and I’m happy there are advances being made. I was responding to claims of automation hitting truck drivers relatively soon (in the next decade), and causing drivers to lose employability, which I think is pie in the sky thinking.
If anything, I think truck drivers’ jobs will be resilient and in demand with all the move to delivery in the next decade or even two.
It’s going to take a lot of time and data to prove the safety of self driving trucks and sort out the calculation of liability and who is insuring what.
There's money in it. Whether its transponders on the roads to help guide them, RFID tags in the cement for accurate lane location, beacons at terminals that share route information, or all of that, it'll happen because it will pay.
We built the railroads because it paid. It's not about where we are, but where we'll be going.
Those are all corner cases that won't convince spreadsheet guy to stop automating. And once automation does take hold, there will be plenty of incentive to solve them.
> Show me one that can handle a snow squall or heavy rain or missing signs and lane markers and construction zones, and I might think it’s possible this decade.
And the next step would be correct recognition and behaviour
based on the current location. Different countries use different
colors and styles for road markings. Some use yellow lines in
the middle, others orange in the case of redirection around
construction sites. Then there are the cases where the lines are
simply incorrect or out of date(which famously lead to the fatal
Tesla autopilot crash).
It is impressive how far the technology has come, but it's an
area with countless edge cases and unpredictable situations
coupled with high stakes if something goes wrong. I can see
self-driving trucks being adopted slowly on specific routes, but
a full switch not labeled as "pilot project" will still take a
while I think.
I imagine for well-plied, fixed routes (interstates between two warehouses) there could be training runs (with a driver)
to check edge cases along routes before going to prod. There could also be periodic retraining with drivers, as well as offline model retraining with the data that the trucks will be collecting continuously. This simplifies the problem significantly (even though it doesn't solve it completely -- there's always the 1% off case).
I don't think one would just plop a general system on any route without human-supervised customization -- at least we're not there yet.
I get it. I think self-driving cars are way over-hyped, and I'm not expecting them to be real level-5 automation for 10-20 years. But I think trucking is different. There is a lot of work to be done, but we can cheat a bit.
The first cheat is to only automate the long-haul portion. As anybody who has driven cross country knows, long stretches of American highway barely require you to be conscious, let alone skilled. The truck pulls over before needs to get on city streets and a human driver gets in.
The second cheat is what Waymo is doing. If the car doesn't know what to do, it slows or stops and asks for help from a central driver that plots a path through the hard part for it. This works for trucking because network coverage on highways is very good.
The third cheat is to work around the limitations. If the weather is going to be bad, the truck waits or takes an alternate route. If it gets into trouble like a freak snowstorm, it either proceeds very slowly or just stops until either conditions improve or a human driver can get to it.
This approach isn't perfect automation, but it will significantly drop labor cost right away, and it gives you a lot of road-miles to iteratively chip away at the problems.
It's simply too risky and will be for the foreseeable future. The liability cost for a semi-accident is huge, not just in lives but in cargo and in the truck itself.
Highways have decent network coverage but nothing close to the levels of reliability necessary to satisfy the actuaries. From a risk perspective, failure mitigation strategies like slowing or stopping are largely hamstrung by the same reliability concerns that prevent the vehicle from simply driving on its own, the vehicle may simply not recognize that the situation is dangerous until it's too late, that's without factoring in the reality that stopping or slowing can actually be extremely dangerous and can't be relied upon as a catch-all fallback. Additionally, I think you're greatly underestimating the practical problems involved in operating a remote emergency-on-call driver system, there's also some questionable economics around the idea of paying to keep someone on call when you could just pay them to drive the truck.
On top of all that there are various other hard problems that need solving like fueling and security. You can be sure that automated trucks hauling valuable cargo will find themselves regularly waylaid by sophisticated criminals who will be greatly encouraged by the lack of a human element to resist their efforts. Truck drivers don't really get paid that much money, the risks just aren't worth it.
These concerns are unlikely to surface in ways that are important enough to overcome the advantages of automated long-haul trucking.
Network coverage? On most big interstates it's perfectly fine, even through mountain ranges, and it's known in advance which highways don't have it. But even then, you can just use satellite systems such as Starlink as a backup.
Questionable economics for keeping people on-call? No, it's quite economic, because one on-call person can provide coverage for a wide area.
Criminals? This is an unserious idea that's always brought up in these discussions. Why not break into containers at ports? Why not break into dry vans stored at the back of warehouses overnight? If the idea is just that criminals will break into unattended cargo vans, there are manifold opportunities to do this already, and the small amounts of property crime isn't a serious issue.
Anyway, it's covered by insurance.
Conversely, the advantages are numerous:
* Much faster cross-country freight. No sleeping, no work-rules, no need for team drivers.
* A safety profile that may even be better (no falling asleep).
* Lower costs.
* Better availability. At times, it is very difficult to schedule a truck to arrive, depending on how tight the market is. It's easier to flex equipment availability up and down than truckers.
Many large shippers will see this as an important area to focus on and money is being poured into this today.
Most real cross-country freight (say, from west coast ports to Chicago or other major hub cities) goes by intermodal rail already. There are of course some odd lots that go by truck but most over-the-road trucking is already medium/short range today.
The problem is that long haul trucking isn't just "most big interstates", the only advantage of long-haul trucking is the logistical flexibility, if you're constrained to "most big interstates" you might as well use a train.
The remote on-call strategy doesn't scale, yes it'll save some money relative to drivers, but there's a delicate balancing act involved in mitigating risk of multiple simultaneous failures, yes this might be rare, but it'll also be costly.
See my response to wpietri for more about security, but with respect to warehouses and the like, they are frequently staffed by armed guards or dogs, it's also orders of magnitude easier to secure a stationary facility compared to a mobile autonomous vehicle traveling across the country. Insurance will only cover it if they can tolerate the risk.
Increased availability is definitely a win, but the other advantages you listed are just alternative ways of saying "lower cost". Yes, the costs are lower, but truck driver salaries aren't that high to begin with and the cost of building out and maintaining automated trucking infrastructure will be passed on to any trucking companies that decide to adopt the technology.
I would like to see your math on the risk. It's not like human drivers are perfect; for decades, long-haul truckers have been known for drug use just to keep awake. [1] And trucks break down for all sorts of reasons, so "robot has issues and pulls over" could plausiblly be just another one of those.
I think the economics of remote backup drivers is working for Waymo, so I don't see why it wouldn't work for truck drivers, which are higher value. And for the you-need-somebody-on-site problems, there's already a network of tow and repair services that will surely be willing to send someone out.
Fueling seems similarly surmountable. Semis have very long ranges, and fuel up at specialist facilities that are just off the highway. I don't think it will take much persuasion for truck stops to offer full service fueling on a $500 fill-up. And it's not obvious that security will be worse. An automated truck is going to have a lot of high-quality cameras and be in continuous contact with a staffed operations center. The vehicle may be effectively impossible to steal, and transferring a semi's worth of cargo is not a fast process.
I think it's also important to note that to start it doesn't actually have to be better. It just has to be a) within shouting distance, and b) exciting enough that some exec thinks they'll get a promotion out of it.
Waymo is an extremely specialized closed system that can't be generalized to the needs of cross-country long haul trucking, if it could we wouldn't be having this conversation.
The fueling concerns are pretty trivial from an engineering perspective, the problem is one of infrastructure. Yes, industry could prop up some fueling stations, but nothing approaching the coverage necessary to make it more practical than e.g freight by rail.
Security concerns are very real, there is no such thing as "impossible to steal" and the threat of cameras are trivially defeated by way of masks under cover of darkness. Transferring a semi's worth of cargo is only slow if you're not prepared to do it in advance. I also want to reemphasize the shift in psychology associated with removing the human element, criminals are absolutely going to be emboldened by the fact they are dealing with a harmless predictable computer and not a human person that might resist their efforts, cause them injury, or otherwise complicate the risk involved. It's also well known in criminal psychology that people are much more willing to justify criminal behavior they perceive as victimless.
Setting aside my practical objections, yes, it's true that it doesn't have to be better to start, but I think there's no chance it gains momentum since there is a lot of up-front cost and complexity for relatively small financial gain vs just e.g. freight by rail.
You make a lot of assertions here, but I don't see any data backing them.
Fueling stations already exist. The only "infrastructure" needed is an agreement with a couple truck stop chains than when a robot truck pulls in, somebody goes out fill it up. Full-service gas stations are not a radical innovation.
I think your security concerns are a fairy tale. I especially suspect you have never loaded or unloaded a semi. A mask might keep a person's face from being seen, but it can't keep the theft from being detected. Trucks generally drive on well-policed, limited-access roads. For this to work, time of trouble to time of police response has to be larger than the time to break into, unload, and load a semi's worth of goods. By the side of the road. At night. Plus you need more time to get a semi, which is not an inconspicuous, far enough away from the scene of the crime, that the cops won't find it.
Will people try it? Sure. But I don't see any reason to think it becomes a major issue. Certainly not a dealbreaker, because if I'm wrong and there's a crime wave like this, a bunch of well-funded, powerful groups will be immediately demanding better police response and new laws.
outside of the major east coast (and maybe some west coast?) cities, US interstates tend to be pretty straightforward and well-marked. I wouldn't be surprised if we saw trucks driving themselves most of the length of I-70 in the near future. they would have to stop in some conditions where a human driver could keep going, but in ideal conditions, they could also keep driving when a human would be required to sleep.
They could band together as pirates as their alternatives for employment become null. That or becoming digital servants doing shopping and other erands for the well to do for peanuts.
The U.S. will not spend the amount on infrastructure required to make this happen in the next decade.
We can solve self driving, but it requires better maintained infrastructure, you can’t have self driving trucks which break down due to pot holes and black ice and no internet access clogging up the logistical supply chain.
I can see long distance high way truck driving jobs on set routes going away. That scenario is relatively simple to automate. That is moving goods from one big hub to an other. On other hand the last delivery leg likely will stay human.
On other hand I do agree that skilled labour jobs in maintenance and renovation jobs aren't going very soon.
Then again in new buildings automating these jobs with good design or even pre-build components is not impossible eventually.
Oh you sweet summer child. Truck drivers are among the first that will be completely automated away.
The other thing to keep in mind, is that we automate tasks not jobs. The effect of this is that workers become much more productive, it is possible to do the same work with fewer people. One truck driver oversees a fleet of trucks remotely, ready to take over if needed. The good news is that these should be higher paying jobs than the original, because they are more productive and higher skilled.
Whether it gets rolled out will depend on the level of liability self driving vehicle operators will face (or be able to avoid).
Self driving cars as a concept could in theory be killed entirely by a few high profile crashes and a few successful prosecutions for criminal negligence.
Im skeptical that the US legal/political system would kill it, though. The original automobiles were rendered "safe" by criminalizing jaywalking after all.
I agree. Passenger spaceflight would be offered widely and more affordably in short order if you remove all liability from vendors for deaths and injury of passengers. That doesn't mean it's ready yet.
Unfortunately, it’s good news only for the driver who gets to manage the fleet, but bad news for the rest of the fleet’s drivers who are now unemployed.
Agreed. I think the first step here is to automate the long-haul portion, which is pretty straightforward. You can then have manual drivers take over at the switchover to local streets, which require real skill for an 18-wheeler.
The problem is that trucks are one of the things that needs to be right essentially every time. If a truck is right 99.9% of the time it is doing considerably worse from a financial liability standpoint than a truck driver.
One truck accident of any severity will incur liability expenses in the millions and often tens of millions of dollars. And with automated trucks there are no owner operators to externalize that liability to.
Outsourcing truck driving via tele-driving on the other hand, seems incredibly likely if regulatory questions can be answered.
It doesn't have to be right; it just has to fail safely and then wait for a driver to take over, which is a much easier problem. Waymo has been pioneering that approach for a while and it seems like it's working for them.
I have doubt that more productive and higher skilled == higher paying. True in some/many cases but our overall society is more productive than 50 years ago, wages haven't really reflected that.
No, there won't be a single person controlling a fleet of trucks on a highway. It's not tiny drones we're talking about. Huge trucks - on highways, where people are in their cars (driving or not). Won't happen. Not in a 100 years.
Now replacing drivers for trucks driven on public roads is much harder to achieve, but even here systems that reduce the number of drivers required are well advanced:
Consider pilots in an aeroplane. Most of flying is automated but there are more pilots than ever. There will never not be pilots.
Same will happen to trucks and trains. You may see some sort of fully automated cars in some small cities in the next couple of decades. But trucks? Nope! Not happening.
The problem is that workers get Manna in place of the human management - automated surveillance systems that track every moment of their working life and slice and dice this data to figure out who can be fired or harassed and where to tighten the screws on them.
Manna by Marshall Brain is a great short book about this that's available free online.
What are you talking about? Every large company is jumping aboard the automatic worker surveillance, tracking and scoring. Where are you gonna go work? To a competitor who uses the exact same surveillance system?
Name them please. I work for a largish company, this subject was openly addressed in a townhall and stated it's not gonna work by CEO and the board.
As for where you gonna go, like everything else, it just creates opportunities for companies that might not paid as much but offer better environment, that's where I'm gonna go.
I have often complained about my employer on here, for their lack of foresight, their lack of empathy from the top, and (right now) their lack of really giving a shit about us as humans during a pandemic.
But. At the end of the day, I stay here for less money than I could make somewhere else for two reasons:
A) I don't like moving; and
B) They leave me the fuck alone. They don't really know what my office does. They don't really know how we do it. But they do know that we provide massive value for the expense. And they let us do it with as little oversight as they can stomach.
I can't name my employer (too small - less than a thousand egineers) but I can tell you they are tracking using a plugin for Atlassian. We are an Atlassian shop through and through including our source code control. They deployed this plugin last year. What is really sinister is that it's open to all employees so they can see in real time how they rank against their peers ącross the various metrics.
I'm pretty sure Atlassian peddles this product to everyone who uses their suite.
Exactly. We already know that such metrics do not identify the best performers, they identify those who are willing and able to game the system the most.
It's already here in place at my company and many others. PRs, commits, comments and even slack and email activity are being tracked and compared between employees.
Way past that. "Agile" (i.e. micromanagement) methodologies and tooling are already firmly seated which is basically this same special hell.
I guess we have a generation of technologists that didn't realize when to stop foaming at the mouth at efficiency over humanity or perhaps they were lead here with bags of money and deceit.
I love efficiency and optimisation too (with certain processes), but not at the cost of human happiness. Our current optimizations are ever growingly sacrificing basic happiness for efficiency which is not a tradeoff we should want or agree to.
I think you underestimate the line of future workers willing to work under almost any condition.
People said this about consulting companies when they retracted offers in 2001. People said this about wall street firms when they reigned in party culture in the early 2000s. Yet, the show goes on because fresh graduates and immigrants at all levels need jobs and are willing to bear with it all.
Well, I'm not the GP so I can't tell for sure. But "lose competitiveness" without any further said normally refers to a company as a seller, on its products markets.
When referring to the company as a buyer, like in the labor market, people usually qualify it.
It's not about workers being "willing" - it is about handicapping worker "ability." You surveil me and it will reduce my performance, creativity, etc. It isn't necessarily conscious on the workers part - it's like a virus that infects all that worker does.
> I think those firms will likely lose their competitiveness overtime. Unless they are defacto monopoly.
I suspect that's wishful thinking that assumes a greater alignment than actually exists between what we want for society and what the market will lead us to.
I think a more useful term for that might be "management automation". And it's not something I worry about is because a primary function of American-style management is to adapt too-rigid top-down systems to the complex, ever-changing nature of the work. Companies trying to automate management are even more vulnerable to competition and disruption than the ones automating labor.
Surely some portion of that can be attributed to agency ("doing") but I'd contend a non-trivial amount is due to being recipients of American hegemony ("being")
But is that something actually right?
IE in the sense we would want more of it in our lives.
Or is the profitability less about agility as per the GP comment and more to do with employee manipulation(for lower wages) and political sway (for lower taxes).
Hell, given where FAANG pay taxes, are they even American?
I'm pretty sure FAANG pays quite a lot in payroll tax, property tax, and they also ensure that FAANG employees pay quite substantial income tax, property tax, and sales tax.
Don't get me wrong; I think American-style management is on average terrible. I just don't think it's easily automated, because the point is to take the nonsense from on high and make it somehow work, for some sufficiently lax definition of "work".
This is a really bizarre generalization. There are good managers and bad managers, it just depends on the organization and the individual. Bad managers aren't always easy to spot, but it's very obvious who the good ones are because everyone in the organization implicitly looks to them for guidance and support.
I'm a relatively new engineering manager in an older company in Canada (turning itself into fin/insurance tech), and a big part of my role has been about breaking down silos.
The silos where created at a time when being really efficient at 'your piece' was what the system rewarded, whereas now we want to be efficient at shipping code, which means efficient value streams (from inception to production)...which has basically meant many rounds of silo deconstruction.
But I'd be VERY curious about why silos get built in the first place & what incentives are in place to make this happen...and why this is bad. To be honest, maybe I'm just idealistic, but I see silo building as not desirable.
Not OP, but imo the classic thought is that your managerial influence is directly correlated with the count and quality of employees under you. In this sense, silos get built not quite for specialization, but because your individual managerial incentive is to try to grow you and yours at the expense of (internal) competitors for a fixed budget; interoperation can be a direct threat to your influence and growth.
A famous study of behavior like this might be Kodak, and how internally the film "silo" crushed the digital "silo".
For the background and how things go wrong, you might enjoy the book "Confronting Managerialism" by Spender and Locke. The Lean folks also have a lot of good critiques. A book that sticks in my mind is "Simple Excellence: Organizing and Aligning the Management Team in a Lean Transformation", but there are a bunch. And for software specifically, Mary Poppendieck's work is great, as well as Reinertsen's "Principles of Product Development Flow".
Beat me to it. My mind went immediately to Manna. Can't think of a book (or anything else for that matter) that sparked more of my imagination in terms of what can be possible (good and bad) in the near future with our existing tech.
But the biggest short-term takeaway was how middle-management is the most vulnerable to automation. Leadership may be hard to replace with a robot, but a lot of management is pretty straight-forward.
Is there a word for waiting-for-a-twist-that-never-comes? I keep thinking that the Australia Project was really a front for Manna "leveling up" and secretly turning the Terraform dwellers into Vites to further keep them easier to control and out of sight.
I started reading this, got sucked in the first few chapters and lost the link a few years ago and could not find it since. Thank you so much for sharing it!! Now I can finish it.
Well duh, do you really think robots would take the shit jobs first? Would you? Or if you don't think so, it only makes sense they'd come for managers first, when you think about it. Not that there's anything wrong with that.
> "Contrary to the popular notion that robots will replace human labor, we find robot-adopting firms employed more people over time. Any displacement of labor came from firms that did not adopt robots. These non-adopting firms actually lost their competitiveness — and they had to lay off workers."
I don't find this very compelling. How many workers were lost by companies that didn't automate vs companies that did? It's entirely possible for Amazon (or another company with heavy automation) to add 50,000 jobs, while their competitors cut 75,000 jobs, meaning a net loss for workers.
These sorts of articles and the subsequent threads always end up nitpicking which job will get automated and which jobs are safe without considering that the more we automate away jobs, the less society will be able to pay for the goods and services that aren't automated. Maybe tradespeople are safe this decade, maybe not... but one thing is for sure, there will be a smaller subset of people able to pay for renovations by 2030 because their jobs will have been automated.
Assuming your job/role is unmodifiable is a more serious problem that is only slightly aggravated by it being automated.
Some jobs are more easily automated than others, sure, but the problem is that some people don't realise the market might render your current job useless much before automation comes, so you should be ready to adapt always, and in any case.
We like to talk about pivoting business models on this website, but when it applies to what a person considers "sacred" everything falls apart. (not just here, as the automation discourse is happening globally)
Globally it may have hit harder. Most of the simple tasks that were offshored in the early 00's are the first to come back onshore and be automated away
I am fully aware that my job as a CRUD API developer has an expiration date and I have no idea what to do about it. At this point, I don't think the onus is on the workers... automation is too widespread and fast to plan.
Ultimately, automation is good because it means more goods and/or services to more people. More goods and services means that society as a whole can support more people.
A lot of American life is tied down to your job however. That seems to be the problem: health care is tied to your job, retirement benefits are often tied to your job (401k or pension), etc. etc. One can argue that its the overemphasis of your job that's the problem... not the grossly improved productivity due to inventions.
Indeed. Provide robust safety nets and universal healthcare and we can proceed with Automating All The Things with wild abandon. Until then, based on the automation you’re building and deploying into the marketplace, you’re relegating subsets of workers to possible long term or permanent unemployment, despair, and possibly death (people die without health insurance, which many can’t get without an employer or holding on until they’re eligible for Medicare at 65, in the US).
At some point, we have to move past the concept of jobs as the measure.
> Provide robust safety nets and universal healthcare
and the question becomes "who" does the providing? In the US, the going mentality is 'user pays' (which leads to insurance, which leads to insurance company dealing with healthcare provider, which leads to employers providing good insurance as a side benefit to compensation).
The US populous can't easily change their mentality to "society pays". Therefore, this is where the issue should be tackled.
If housing, healthcare, education, clothing, and food were free, and I still had income, I could spend it on fitness trainers, life coaches, interior designers, personal shoppers, performance artists, geanealogists, ...
There are a lot of potential mass-market service industries that don't exist because all the benefits of productivity growth (and slightly more than all, now) go to the top 0.1%.
There are still plenty of potential jobs - jobs that we will never have because of inequality.
And we should keep the eusocial aspects of jobs in mind. They provide life structure, self respect, and respectability in the eyes of others as well as a means of subsistence.
Edit: I would love to live in a world where tens of thousands of people have "street party organizer" as their job.
Retirement is not at all tied to your job, IRAs exist for this reason. Obamacare was supposed to address the health insurance tie to employment, however all that seems to have done so far is increase costs
In 2012 Obama promised premiums would go down, my main point was that health insurance is theoretically no longer tied to employment. (note: I am aware that in practice it is not so simple as Obamacare plans are a small fortune)
ACA would have caused insurance prices to rise no matter what since healthcare spending would have risen since more people are getting healthcare since the law required insurers to cover everyone.
People that complain about ACA increasing prices are complaining about having to pay more since more people are getting healthcare. ACA also requires younger people to pay disproportionately more than older people.
Health insurance premiums did go down from infinity to somewhere between $400 per month and $1,200 per month for people who previously couldn’t buy it.
It's very rare to have direct proof of the effect of an economic/regulatory intervention, but luckily, in this case, we do. We've had a decade a survey data that indicates exactly what the effect the ACA had on employer provided healthcare was.
Some relevant highlights:
* More than one-third of organizations now have increased out-of-pocket limits, in-network deductibles and/or participants’ share of premium costs in response to ACA. More than one in five organizations have increased copayments or coinsurance for primary care, increased participants’ share of prescription drug costs and/or increased the employee proportion of dependent coverage cost.
* The excise tax on high-cost group health plans (a.k.a. Cadillac tax) is considered the top ACA cost driver beyond 2015. Since 2011, a steadily increasing percentage of organizations has taken action to avoid triggering the excise tax — a trend likely to continue. More than one in ten organizations already have adopted changes to prevent them from triggering the tax, 21% are working on changes and 28% plan to act sometime prior to 2018. Only one-quarter said changes were not necessary either because they have no high-cost plans (23%) or because they plan to pay the tax (2%).
That being said, the year following the passage of the ACA saw a large anomalous spike in employer healthcare costs that wasn't present in Medicaid/Medicare. It's almost undeniable that this was due to the ACA.
That’s what happens when poorer people who tend to work for employers that don’t offer benefits such as health insurance start getting healthcare.
Someone has to pay for it, and it shows up as increased premiums/deductibles/employer share of premium, but in exchange, we get rid of pre existing condition exclusions, we get in network out of pocket maximums, and older people don’t get hit so hard because younger people have to subsidize their healthcare.
Growth in healthcare costs also slowed in the years after ACA was passed as the various provisions were phased in:
65+ has Medicare. Health problems start in your 40s, and greatly accelerate in 50s. Most Americans have clogged arteries, diabetes, or hypertension and need treatment many years before getting to Medicare age.
This isn't an argument over the merits of the ACA. Many of are fine with Universal Healthcare but not with deluding ourselves into thinking that it's a free lunch. It's a statement of fact intended to address the false premise that we don't know the causal effect of the ACA on various healthcare costs.
We know exactly what it was as we have direct evidence as to what actions employers took in response to it e.g. increased premiums, increased out of pocket minimums, reductions in coverage, etc. The result is that, pound for pound, the same healthcare is more expensive.
> Growth in healthcare costs also slowed in the years after ACA was passed as the various provisions were phased in:
This is unsupported by the chart. In fact, it shows the opposite: the largest decreases in healthcare spending growth occurred before the passage of the ACA. That aside, the parent comment is seeking the causal effect of the ACA on healthcare costs, not a low correlating trend.
> Correlation does not imply causation - is there any proof that obamacare is the cause of the increases in costs, or would that have happened anyways?
And even if it did, cost isn't the only measure of success. It's main goal was to increase access to medical care, which I understand it was successful at.
That's very.. one dimensional point of view. You can't just call something a success when you made other people worse off in the process and simply ignored it.
> That's very.. one dimensional point of view. You can't just call something a success when you made other people worse off in the process and simply ignored it.
Not really. It made people better off in the way it was designed to do. It was always going to cost more to give more people access to medical care (e.g. those with preexisting conditions). Rather than "making people worse off", it set the trade-off to a different point than it was before.
In theory, it seems plausible that the additional cost of more medical care was more than offset by the economic value of letting people switch jobs more easily, thus spending more time doing things they are better suited to.
As well as the economic value of saving the lives/careers of people who go on to contribute to society in excess of the cost of their treatment.
I'm not saying I've seen a study proving that helping sick people is a net positive for society, in dollars and cents, but it should at least be considered...?
The maximum IRA limit wouldn’t be a measly $6k if that was true. 401k limit is tens of thousands to disadvantage smaller businesses that can’t afford to offer 401k.
> without considering that the more we automate away jobs, the less society will be able to pay for the goods and services that aren't automated.
That may not matter to those making these automation decisions. Ultimately, ownership is about power and control, not about actually satisfying the needs of the masses. As long as the owners at the top maintain their power and have their personal wants met, they may find the situation acceptable where the market ignores an expanding pool of penniless masses and focuses on a shrinking pool of other elites who still have money (see North Korea for a similar stable result from a different economic system).
For this to work, you need a huge, well paid and loyal army. America has a huge army, but their people have more guns, and pretty much everyone else has less army. By some theories, when a day comes, that fathers are unable too feed their children, the revolution starts, and those usually don't end very well for the elites.
> I look at the videogame market and see the future of ownership: you don't get to have any.
Everything you have will be only a service... and sometimes people will get locked out of their accounts. Sometimes it will mean losing your videogames, sometimes losing your e-mails (and contacts) and documents saved in cloud... and sometimes losing your home and money. That will be fun.
Societies get richer by automating things. How much money people have to pay for things is completely dependent on how much they produce, and as a consequence, how automated is their work. The more automation, the more of everything going around.
But automation has a strong short term effect on wealth distribution. How health your economy will be by 2030 depends completely on how your society deals with this effect. And "stop automating" is not a viable course.
It's not true that productivity is correlated with purchasing power. This hasn't been true since the 70s. This is true for an economy in general, but not for individuals.
How much were pocket supercomputers back then? Or cordless phones? Or chicken pox vaccinations? Or cars designed to protect childrens' lives and health? What did almond milk cost in Duluth? How much were encyclopedias?
Some things are absolutely more expensive, but we definitely have improvements since then, and automation (like modern logistics) contribute to almost all of them.
To put it another way, I would rather have the median income today while doing day-to-day activities now than any past decade. And to the extent that there are downsides in those decisions, I think it's odd to say automation is to blame versus other forces like anemic growth in housing supply.
Not necessarily. Every dollar saved on automation is a dollar still in the pocket of the automation owner. The money supply doesn't go down. The type of demand shifts, but it doesn't vanish.
Oh, sure, you've got the risk of an unemployable underclass full of dystopian ennui. But before the guillotines come out, the .01%ers will have lots of cash for buying art, personalized AAA video games, solid gold plumbing fixtures, and the like.
There's good reason to think that we need to reconsider the value of human life in an automated era, when there is plenty of stuff to keep people alive if only we could distribute it just a little bit. We've actually been doing that for well over a century, and there's good reason to think that we've avoided the torches and pitchforks so far only because of that.
The aggregate demand of "society" needn't go down without it. It just might behoove us to consider that the median might be more important than the sum.
The money supply doesn't go down, but the distribution of that wealth matters a lot. The 0.01%ers don't spend the same way everyone else does. They hoard. The more they get, the more they will hoard and for the rest of society, that means there will be less to go around.
If the money is never spent, it is worth 0 and doesn't effect the rest of economy.
If Mr Bezos begins to buy every yellow plastic duck at $100, you surely will see their price spiking, until then nothing happens. Same goes with whole economy.
Another way to see it: if Jeff is only allowed to spend $100K/$1M/$10M a day, how his total worth is effecting the whole economy ?
He's not hoarding anything. People who claim the rich "hoard" wealth is under a misconception that investing wealth takes away wealth from other people.
It doesn't. Investing wealth creates more wealth. Of course, the instigator of the investment gets a disproportionate reward for such investment, but taxation, and the general increase in technology level due to such investments will benefit society at large, even if the poor is an unrelated party to the investment.
Literally anything you do with wealth can be described as "investing".
Therefore it's implausible that investing innately creates more wealth - there is an opportunity cost.
(In this case: the opportunity cost is first giving everyone a decent standard of living, so that everyone can have wealth and invest wealth and create more wealth, as you say)
> Literally anything you do with wealth can be described as "investing".
except for spending it on consumption.
> ... implausible that investing innately creates more wealth - there is an opportunity cost.
the _only_ thing investing wealth does (or hopes to achieve) is to increase the amount of wealth that exist, and most importantly, capture that increase as much as possible.
"giving everyone a decent standard of living" is indeed an increase in wealth, but this increase is uncaptureable by the investor. Therefore, i will say that this sort of "investment" is more closely described as charity than investment.
> But before the guillotines come out, the .01%ers will have lots of cash for buying art, personalized AAA video games, solid gold plumbing fixtures, and the like.
That doesn't pass the smell test. For instance: it doesn't help the 99.9%ers if one .01%er pays another 10x for a Picasso and a 0.1%er makes a commission on the sale, and I don't imagine the elite demand for new-art production would increase much from current levels. Production of personal goods, luxury or otherwise, for elites can't fill the gaps caused by automation, since they already buy them today. Even over-the-top conspicuous consumption won't cut it either (e.g. employing tens of thousands in North Korean Mass Games-style performances just to employ tens of thousands), is getting into pretty dystopian territory.
Automation has been going on for 300 years. Jobs have disappeared, but other jobs appeared. People can be replaced with automation, or with other cheaper people. What difference does it make if you lose your job because your output can be done now more efficiently by a script, or by someone in Mumbai, or by a kid fresh out of school? I personally came to terms with the idea that at some point in my career I will be subject to a round of layoffs. I am not looking forward for that moment, but if/when that happens, I'm sure I'll feel so happy when I'll deposit my first paycheck from my new job. It's like being young again. What's not to like ;) ?
Agreed. A ready example is the jobs created by those who would produce, maintain and improve upon such automated solutions. Ubiquitous automation would imply robotics, permitting demand for classes of non-knowledge worker labour, who are typically viewed as the victim of this kind of shift in society.
The only problem tho, is if the person becoming obsolete cannot retrain to something else.
I personally don't believe this to be true - humans are insanely adaptable. Learning new skills is part of it.
If an AI is able to automate and produce everything humans need, there would still be jobs that need a human. For example, prostitution. Or premium service jobs like butlers, etc.
I wonder how communities like the Amish fit into the larger scheme of things. It seems like they've decided to ignore anything later than the 1700's technology. Are there other groups with different technology cut-off dates? If not, any theories as to why not? Do you need a religious component to keep you cohesive over the generations? I think I might be OK with regressing to a 1980's level of technology. I wonder how large of a group of people would you need to be mostly self-sufficient at various technology levels. It seems like you might need a lot of people, if you wanted to have even very old-school integrated circuits. Probably quite a bit fewer if you only had discrete transistors. It seems like sustaining at a previous level should be easier than inventing it in the first place, since the knowledge is already available. And you could siphon off a bit from the more advanced society (e.g. you can potentially use scrap materials left over from the advanced society, instead of maintaining your own mining technology, etc.). And you could still have some (lopsided) trade between the societies (anyone want a bespoke Walkman produced by a skilled 1980's artisan?). Any thoughts on the estimates below?
- 1700's, 100's of people in your community (crude metalworking)
- 1800's, 1000's of people in your community (machines needing more metalworking and machining skills)
- 1900-1940, 100,000 people (radio)
- 1940-1980, 1,000,000 people (TV, automobiles)
- 1980-2000, 10,000,000 people (personal computer, mass air travel)
I sued to go shopping at an Amish market and it seemed that they were just very deliberate in their use of technology. They used trucks to get their stuff to the market, they had cell phones, they had freezers. I think their society is just very focused and rejects distraction.
But it’s not all golden. There were a lot of stories of spousal and child abuse but they stay hidden.
The various Amish sects that I'm familiar with in PA use technology much more recent than the 1700s (including smart phones in some cases).
Most of it is via generators and gas-powered appliances, as they won't have electricity running into their homes.
I'm not an expert by any means but it's just not correct that they stopped using technology developed after the 1700s.
I agree with your larger point that people may be better served by simpler technologies in many instances.
Alan Watts has some interesting (and prescient) writings about how misuse/misunderstanding of new technology leads down a hostile path to societal destruction. Sounds like you may find it worth reading.
> I wonder how communities like the Amish fit into the larger scheme of things. It seems like they've decided to ignore anything later than the 1700's technology.
No, the Amish specifically have decided to carefully review new technology on a moral cost/benefit level before deciding whether and how to allow it in their communities. They don't ignore technology, and don't have any "cut-off date".
An example I heard on the radio was about the men who work in Chicago. The buggy takes them. As they are leaving their family, it should not be a trivial to do that. They had cell phones for emergencies, but when at home they were kept in the out house as they had no place when everyone was home.
A believe a lot of it also revolves around being dependent on outsiders. For example, I think in some cases it's OK to generate their own electricity with a diesel generator but they will not use electricity from the utility grid.
> there will be a smaller subset of people able to pay for renovations by 2030 because their jobs will have been automated
That's assuming new jobs are not going to replace the old ones. Human society has a tendency to use up all the available capacity by scaling up its goals. Do you think we'll stop inflating our expectations?
Even if our expectations don't drive up demand for new jobs, the government can. The transition to full automation, not to mention the global warming crisis, are hard work and require lots of jobs.
The premise of the article supports this idea - the decline in jobs happens mostly in companies that avoid automation because they lose out on quality, price or are less innovative.
In the extreme case that companies automate almost all jobs, then we're left with billions of people with no ability to earn an income. But they have lots of free time and unfulfilled needs. So they will use that sweet automation to solve most of their needs by working for themselves and being less reliant on external help.
There's always work to do when there are people, especially when they are forced to be self reliant. Of course I don't wish self reliance to look like North Korea, hopefully we can do this in a humane way.
If the FED doesn’t succeed in fixing the productivity to compensation ratio ( https://wtfhappenedin1971.com/ ), there will be more revolts by people who cant afford to send their own kids to the dentist.
The article's point is that fewer managers are required as organizations become more automated, not that managers are being directly automated. However, I'm not sure I agree with the study author's statement "Because managers, by definition, supervise other human beings, so we really cannot replace their functions until the singularity occurs with artificial general intelligence."
This came up on HN a while ago: https://apenwarr.ca/log/20190926 . The thesis was: "To paraphrase the book, the job of an executive is: to define and enforce culture and values for their whole organization, and to ratify good decisions. That's all. Not to decide. Not to break ties. Not to set strategy. Not to be the expert on every, or any topic. Just to sit in the room while the right people make good decisions in alignment with their values. And if they do, to endorse it. And if they don't, to send them back to try again. There's even an algorithm for this." All of that sounds automatable to me using current or near-future technology.
Managers, especially middle managers, also do many other things; but they universally end up spending a whole lot of time in meetings. I'm not a manager, but my own role in many meetings feels automatable. In a handful of meetings I have to be there because I have detailed knowledge about the subject to share. In the other meetings I think it would honestly be fine for an AI that had learned to emulate my writing and speech to represent me in the conversation and write a summary for me afterward.
I'm not arguing that managers _will_ be automated anytime soon, just that AI probably _could_ automate some of the things that take up a big chunk of managers' time.
In my country, companies are required by law to provide their employees a way to clock-in and clock-out at work. Companies are also required to keep a copy of these records that can be audited any time. This was seen as a great improvement for a country famous for its long workdays and systemic unpaid overtime.
The response of many company owners has been to forge their employees records. No way to clock-in and clock-out is provided, instead employers craft all this information and keep it to themselves. Everything is exactly as it used to be.
I'm one of those people who think AI could replace many management jobs, since my manager struggles at things a computer does easily, such as calculating annual leave. Many days all my interactions with him could be improved by a well-crafted bot. The main issue is that many companies are actively fighting AI and management automation, since it would expose the schemes that keep their business running.
When it comes to organizing Sprints, I want to get automation maxed out there. Managers usually want to do things "their way" but I think the mechanics of Sprints are well enough understood to just go by the process and enjoy it.
On the other hand I think it's good to have a manager to talk to, also someone that deals with conflicts at work and someone who is able to reverse toxic dynamics. No way this can be automated in foreseeable future and neither do I see that "self-organized teams" find any way to solve this. Also dealing with product requirements is hard to automate.
> Rather, it is “middle-skilled” jobs that are threatened, Wu found.
I think this would be a better headline. If "middle-skilled" refers to manually moving information across an organization using people, systems, and spreadsheets, then yeah, those jobs are very much the bottom of the iceberg as far as automation goes.
So I guess from that perspective, an employer with many managers who manually manage scheduling/time tracking/etc. could consider deploying an ERM/ERP product and consolidating the org structure an efficiency.
Manager as leader though is harder to consolidate, both during growth ("we're adding resources to your team, you should be happy") and through adding offshore teams without a project manager, cutting down on the bandwidth of the onshore manager.
I'm sorry but any article that refers to "the singularity" so casually and cursorily loses my respect. At the very least this needs to be defended and/or explained, because as far as I'm concerned it was a big marketing ploy for Mr. Kurzwel to brand himself with.
Yes, the singularity is science fiction, but I don't think that's a fair characterization of Kurzweil. If you've read any of his books on the topic he lays out a pretty meticulous if flawed case for the singularity, he may be wrong but I don't see any reason to suggest he is anything other than earnest.
I worked as a field construction supervisor (as a junior civil engineer) and remember that the senior supervisor used to told us that we were "his eyes of the job site". Because at the end he was the one who aproved the important things. Of course I also used to take decision in the field and aprove little things. But if placing "video cameras and AI" could replace 70% of my job at that time and pass the rest of the tasks directly to the senior supervisor to aprove it in a faster way, then the junior field supervisor might not be necessary or might have to be retrainned in other areas.
There's a science fiction book about someone whose boss is
"Microsoft Middle Manager 2.0 with passive-aggressive set to low".
But, yes. If everything you do goes in and out over a wire, replacing you is easier.
Span of control, the number of people directly supervised, has been increasing for decades. Thus the "flattened organization", with fewer middle managers.
I believe the paper discussed is here [0]. I've skimmed it and am somewhat underwhelmed by both the methods and theory:
* All of their regressions are in levels.
* Look at Table 1, Column 1.
* Robotics investment is positively correlated with overall employment.
* What this tells me is that, "Large companies, by dint of their being large, make large investments. Some of those investments are in robotics. Likewise for small companies."
* To me, it has nothing to do with the claim "More robots => more employment."
* Running regressions in first differences would be better (but not enough).
* Then you could begin to work towards an interpretation like, "companies that _increased_ their investment in robotics in year X tended to increase/decrease employment/management/turnover by X." You can account for lags and the like via your specification.
* Fixed effects _are not_ equivelant to first differences in this setup!
* Separate out the long-, medium- and short-term consequences:
* Investment in new technology may require hiring new "talent" that is knowledgeable of the tech.
* In cases where there is a tech shock and firms are all buying the same machine, demand for a certain skillset will grow quickly. So you'd likely see:
* An uptick in turnover among affected workers
* A move to a "flatter" organizational structure (various reasons). So maybe fewer people who are managers per se.
* Possibly short-term productvity decreases arising from implementation issues. (e.g. have to hire people to check the work of machine in staging manually while still running previous process.)
* Over time, though:
* Implementation issues/bugs are smoothed out
* Rare skills are commidified. The need to hire "superstars" ends.
* It is important to ask, "what might have caused investment by a given firm over a given time frame?"
* Common industry technological shocks might spur investment.
* (And the simplest way to account for technological shocks in this sort of regression you'd want `[industry] x [year]` fixed effects.)
* "We went on a manager hiring spree in year 1. Those managers, in turn, bought a lot of robots in year 2. In year 2 we hired fewer managers because we'd just hired a bunch in year 1."
* Reverse causation issues are pervasive.
* Finally, I think it is important to notice their notion of "robot adoption":
* They are talking about physical hardware _only_.
* Their measure of investment is coming only from import data. The set of firms that are going to go through the rigammarole of importing physical hardware is probably raher special: Probably:
* Larger
* Have a high expected ROI from investment
At the very least, this paper needs first-difference regressions and a good instrument that can help to disentangle what might be going on. For the instrument, it might make sense to look into changes in tariff duties over time [1].
We're automating away until our creations become self replicators.
Humans are self replicators, so are the other living things, culture (+memes), technology, companies and the economy. If we could have self replicating factories, then all we'd need to do is to import such a factory to a poor country and let it run. Of course automation is just half the problem, the other is raw materials. Presumably we could recycle almost everything and not need an unlimited quantity.
I bet we'll reach technological self replication singularity before AGI, and the impact is going to be huge. Robotics, 3D printers, IoT, sensors, computer assisted manufacturing and logistics are going to be interesting in the next decades. I just hope they will be open source and not restricted by patents.
Perhaps in companies intelligent enough to understand which part of their workforce drives actual revenue: in my job I can say for a fact that there’s more redundancy in management than in engineering, and I don’t think that’s because of intelligence on the company’s part, but rather an engrained culture and assumption of how necessary managers are, often linked to already per-company arbitrary objects like departments or projects.
So a more interesting question I think is: how many companies out there are intelligent enough to go against the axioms of corporate culture and investigate value from the recently so popular term “first principles”? I’m unfortunately cynical enough to believe the answer is “not many”.
197 comments
[ 3.5 ms ] story [ 208 ms ] threadTruck driving jobs aren't going anywhere, not in 10 years, not in a 100 . Neither is that of plumbers, electricians or many such skilled labors. Most programmers will lose their jobs before any of that happens.
But self-driving trucks, especially long-haul ones that drive mostly on motorways between warehouses are likely near the top of the list of replacement by self-driving vehicles.
Provided you're not actually talking about poor people just dying off in large numbers, we'll still have roughly the same number of households, which means the same number of sinks to unclog and outlets to wire. I suppose it might be true that fewer people will be able to afford the services of skilled professionals, but that's not at all the same as saying that fewer will be needed...
> TuSimple is hardly the only company working to making fully automated shipping a reality. Several companies, including Aurora, Daimler, and Embark Trucks, are competing for a slice of the future of self-driving freight trucks. Alphabet-owned Waymo confirmed on Tuesday that it will be expanding its own self-driving trucking routes throughout the American Southwest and Texas, following previous tests in Arizona, California, Michigan, and Georgia
Automation in that regard is probably a later iteration of the automated household where now we have auxillary vacuums, embedded iot light controllers, and so forth.
Short-haul/local drivers however yes, they will continue to exist for now due to the additional complexity preseneted by everyone and their mom's loading dock being all kinds of differently designed, often in ridiculous ways.
Of course they're lining for it, but if no one delivers L5 autonomy trucks, then they'll still be stuck with drivers.
I have yet to see any evidence of a self driving vehicle that can handle anything other than the simple roads of sunny AZ or CA.
Show me one that can handle a snow squall or heavy rain or missing signs and lane markers and construction zones, and I might think it’s possible this decade.
Otherwise, I assume there’s a lot of work to be done.
https://venturebeat.com/2019/04/26/5-companies-are-testing-5...
(Maybe they’ll finally figure out the belt system /s)
Since they don’t advertise being able to drive in less than ideal weather, I assume it means they can’t yet.
I’m not disagreeing about weather constraints in the ODD. But I don’t think anyone who’s driven in Pittsburgh thinks it has “simple roads” like Chandler, AZ. I should have been more explicit I was only responding to that aspect.
If anything, I think truck drivers’ jobs will be resilient and in demand with all the move to delivery in the next decade or even two.
It’s going to take a lot of time and data to prove the safety of self driving trucks and sort out the calculation of liability and who is insuring what.
Look, they're just machines, they're not miracle workers ;-)
We built the railroads because it paid. It's not about where we are, but where we'll be going.
And the next step would be correct recognition and behaviour based on the current location. Different countries use different colors and styles for road markings. Some use yellow lines in the middle, others orange in the case of redirection around construction sites. Then there are the cases where the lines are simply incorrect or out of date(which famously lead to the fatal Tesla autopilot crash).
It is impressive how far the technology has come, but it's an area with countless edge cases and unpredictable situations coupled with high stakes if something goes wrong. I can see self-driving trucks being adopted slowly on specific routes, but a full switch not labeled as "pilot project" will still take a while I think.
I don't think one would just plop a general system on any route without human-supervised customization -- at least we're not there yet.
The first cheat is to only automate the long-haul portion. As anybody who has driven cross country knows, long stretches of American highway barely require you to be conscious, let alone skilled. The truck pulls over before needs to get on city streets and a human driver gets in.
The second cheat is what Waymo is doing. If the car doesn't know what to do, it slows or stops and asks for help from a central driver that plots a path through the hard part for it. This works for trucking because network coverage on highways is very good.
The third cheat is to work around the limitations. If the weather is going to be bad, the truck waits or takes an alternate route. If it gets into trouble like a freak snowstorm, it either proceeds very slowly or just stops until either conditions improve or a human driver can get to it.
This approach isn't perfect automation, but it will significantly drop labor cost right away, and it gives you a lot of road-miles to iteratively chip away at the problems.
Highways have decent network coverage but nothing close to the levels of reliability necessary to satisfy the actuaries. From a risk perspective, failure mitigation strategies like slowing or stopping are largely hamstrung by the same reliability concerns that prevent the vehicle from simply driving on its own, the vehicle may simply not recognize that the situation is dangerous until it's too late, that's without factoring in the reality that stopping or slowing can actually be extremely dangerous and can't be relied upon as a catch-all fallback. Additionally, I think you're greatly underestimating the practical problems involved in operating a remote emergency-on-call driver system, there's also some questionable economics around the idea of paying to keep someone on call when you could just pay them to drive the truck.
On top of all that there are various other hard problems that need solving like fueling and security. You can be sure that automated trucks hauling valuable cargo will find themselves regularly waylaid by sophisticated criminals who will be greatly encouraged by the lack of a human element to resist their efforts. Truck drivers don't really get paid that much money, the risks just aren't worth it.
Network coverage? On most big interstates it's perfectly fine, even through mountain ranges, and it's known in advance which highways don't have it. But even then, you can just use satellite systems such as Starlink as a backup.
Questionable economics for keeping people on-call? No, it's quite economic, because one on-call person can provide coverage for a wide area.
Criminals? This is an unserious idea that's always brought up in these discussions. Why not break into containers at ports? Why not break into dry vans stored at the back of warehouses overnight? If the idea is just that criminals will break into unattended cargo vans, there are manifold opportunities to do this already, and the small amounts of property crime isn't a serious issue.
Anyway, it's covered by insurance.
Conversely, the advantages are numerous: * Much faster cross-country freight. No sleeping, no work-rules, no need for team drivers. * A safety profile that may even be better (no falling asleep). * Lower costs. * Better availability. At times, it is very difficult to schedule a truck to arrive, depending on how tight the market is. It's easier to flex equipment availability up and down than truckers.
Many large shippers will see this as an important area to focus on and money is being poured into this today.
The remote on-call strategy doesn't scale, yes it'll save some money relative to drivers, but there's a delicate balancing act involved in mitigating risk of multiple simultaneous failures, yes this might be rare, but it'll also be costly.
See my response to wpietri for more about security, but with respect to warehouses and the like, they are frequently staffed by armed guards or dogs, it's also orders of magnitude easier to secure a stationary facility compared to a mobile autonomous vehicle traveling across the country. Insurance will only cover it if they can tolerate the risk.
Increased availability is definitely a win, but the other advantages you listed are just alternative ways of saying "lower cost". Yes, the costs are lower, but truck driver salaries aren't that high to begin with and the cost of building out and maintaining automated trucking infrastructure will be passed on to any trucking companies that decide to adopt the technology.
I think the economics of remote backup drivers is working for Waymo, so I don't see why it wouldn't work for truck drivers, which are higher value. And for the you-need-somebody-on-site problems, there's already a network of tow and repair services that will surely be willing to send someone out.
Fueling seems similarly surmountable. Semis have very long ranges, and fuel up at specialist facilities that are just off the highway. I don't think it will take much persuasion for truck stops to offer full service fueling on a $500 fill-up. And it's not obvious that security will be worse. An automated truck is going to have a lot of high-quality cameras and be in continuous contact with a staffed operations center. The vehicle may be effectively impossible to steal, and transferring a semi's worth of cargo is not a fast process.
I think it's also important to note that to start it doesn't actually have to be better. It just has to be a) within shouting distance, and b) exciting enough that some exec thinks they'll get a promotion out of it.
[1] e.g. this from 1973 https://genius.com/Jim-croce-speedball-tucker-lyrics and this from the Trucking Alliance: https://truckingalliance.org/new-study-supports-projection-o...
The fueling concerns are pretty trivial from an engineering perspective, the problem is one of infrastructure. Yes, industry could prop up some fueling stations, but nothing approaching the coverage necessary to make it more practical than e.g freight by rail.
Security concerns are very real, there is no such thing as "impossible to steal" and the threat of cameras are trivially defeated by way of masks under cover of darkness. Transferring a semi's worth of cargo is only slow if you're not prepared to do it in advance. I also want to reemphasize the shift in psychology associated with removing the human element, criminals are absolutely going to be emboldened by the fact they are dealing with a harmless predictable computer and not a human person that might resist their efforts, cause them injury, or otherwise complicate the risk involved. It's also well known in criminal psychology that people are much more willing to justify criminal behavior they perceive as victimless.
Setting aside my practical objections, yes, it's true that it doesn't have to be better to start, but I think there's no chance it gains momentum since there is a lot of up-front cost and complexity for relatively small financial gain vs just e.g. freight by rail.
Fueling stations already exist. The only "infrastructure" needed is an agreement with a couple truck stop chains than when a robot truck pulls in, somebody goes out fill it up. Full-service gas stations are not a radical innovation.
I think your security concerns are a fairy tale. I especially suspect you have never loaded or unloaded a semi. A mask might keep a person's face from being seen, but it can't keep the theft from being detected. Trucks generally drive on well-policed, limited-access roads. For this to work, time of trouble to time of police response has to be larger than the time to break into, unload, and load a semi's worth of goods. By the side of the road. At night. Plus you need more time to get a semi, which is not an inconspicuous, far enough away from the scene of the crime, that the cops won't find it.
Will people try it? Sure. But I don't see any reason to think it becomes a major issue. Certainly not a dealbreaker, because if I'm wrong and there's a crime wave like this, a bunch of well-funded, powerful groups will be immediately demanding better police response and new laws.
https://en.wikipedia.org/wiki/Luddite
https://www.bbc.com/news/uk-england-northamptonshire-5497278...
Why anyone thinks companies would do anything more then keep an eye on their insurance rates, I don't know.
We can solve self driving, but it requires better maintained infrastructure, you can’t have self driving trucks which break down due to pot holes and black ice and no internet access clogging up the logistical supply chain.
On other hand I do agree that skilled labour jobs in maintenance and renovation jobs aren't going very soon.
Then again in new buildings automating these jobs with good design or even pre-build components is not impossible eventually.
The other thing to keep in mind, is that we automate tasks not jobs. The effect of this is that workers become much more productive, it is possible to do the same work with fewer people. One truck driver oversees a fleet of trucks remotely, ready to take over if needed. The good news is that these should be higher paying jobs than the original, because they are more productive and higher skilled.
Self driving cars as a concept could in theory be killed entirely by a few high profile crashes and a few successful prosecutions for criminal negligence.
Im skeptical that the US legal/political system would kill it, though. The original automobiles were rendered "safe" by criminalizing jaywalking after all.
Take the snarkiness back to reddit please.
The problem is that trucks are one of the things that needs to be right essentially every time. If a truck is right 99.9% of the time it is doing considerably worse from a financial liability standpoint than a truck driver.
One truck accident of any severity will incur liability expenses in the millions and often tens of millions of dollars. And with automated trucks there are no owner operators to externalize that liability to.
Outsourcing truck driving via tele-driving on the other hand, seems incredibly likely if regulatory questions can be answered.
No, there won't be a single person controlling a fleet of trucks on a highway. It's not tiny drones we're talking about. Huge trucks - on highways, where people are in their cars (driving or not). Won't happen. Not in a 100 years.
Go back five years and these trucks would have be driven by a human driver.
Today any new mine will be designed to use autonomous trucks and many of the older mines are being retro fitted to replace those human driven trucks:
https://www.amsj.com.au/autonomous-mining-truck-safety/
Now replacing drivers for trucks driven on public roads is much harder to achieve, but even here systems that reduce the number of drivers required are well advanced:
https://www.autocar.co.uk/car-news/industry/daimler-trials-c...
Same will happen to trucks and trains. You may see some sort of fully automated cars in some small cities in the next couple of decades. But trucks? Nope! Not happening.
Manna by Marshall Brain is a great short book about this that's available free online.
https://marshallbrain.com/manna
But. At the end of the day, I stay here for less money than I could make somewhere else for two reasons:
A) I don't like moving; and
B) They leave me the fuck alone. They don't really know what my office does. They don't really know how we do it. But they do know that we provide massive value for the expense. And they let us do it with as little oversight as they can stomach.
That's worth something.
I'm pretty sure Atlassian peddles this product to everyone who uses their suite.
I guess we have a generation of technologists that didn't realize when to stop foaming at the mouth at efficiency over humanity or perhaps they were lead here with bags of money and deceit.
I love efficiency and optimisation too (with certain processes), but not at the cost of human happiness. Our current optimizations are ever growingly sacrificing basic happiness for efficiency which is not a tradeoff we should want or agree to.
People said this about consulting companies when they retracted offers in 2001. People said this about wall street firms when they reigned in party culture in the early 2000s. Yet, the show goes on because fresh graduates and immigrants at all levels need jobs and are willing to bear with it all.
When referring to the company as a buyer, like in the labor market, people usually qualify it.
I suspect that's wishful thinking that assumes a greater alignment than actually exists between what we want for society and what the market will lead us to.
Or is the profitability less about agility as per the GP comment and more to do with employee manipulation(for lower wages) and political sway (for lower taxes).
Hell, given where FAANG pay taxes, are they even American?
This may change in the case of Amazon with its fulfilment centres, but see above comments re: Manna.
I'm a relatively new engineering manager in an older company in Canada (turning itself into fin/insurance tech), and a big part of my role has been about breaking down silos.
The silos where created at a time when being really efficient at 'your piece' was what the system rewarded, whereas now we want to be efficient at shipping code, which means efficient value streams (from inception to production)...which has basically meant many rounds of silo deconstruction.
But I'd be VERY curious about why silos get built in the first place & what incentives are in place to make this happen...and why this is bad. To be honest, maybe I'm just idealistic, but I see silo building as not desirable.
A famous study of behavior like this might be Kodak, and how internally the film "silo" crushed the digital "silo".
I think I might just be idealistic and having thought about it this way. Thanks for the context.
But the biggest short-term takeaway was how middle-management is the most vulnerable to automation. Leadership may be hard to replace with a robot, but a lot of management is pretty straight-forward.
[1] https://www.strike.coop/bullshit-jobs/ [2] https://en.wikipedia.org/wiki/Bullshit_Jobs [3] https://www.youtube.com/watch?v=kikzjTfos0s
https://www.youtube.com/watch?v=yK9Cs_UcTEE&ab_channel=Artre...
The Leisure Management Suit Advantages: Remote Ops. Centralized Logistics. Lower Stress. High Fashion.
https://www.youtube.com/watch?v=Wj-gisszHM0&ab_channel=Patri...
I don't find this very compelling. How many workers were lost by companies that didn't automate vs companies that did? It's entirely possible for Amazon (or another company with heavy automation) to add 50,000 jobs, while their competitors cut 75,000 jobs, meaning a net loss for workers.
Some jobs are more easily automated than others, sure, but the problem is that some people don't realise the market might render your current job useless much before automation comes, so you should be ready to adapt always, and in any case.
We like to talk about pivoting business models on this website, but when it applies to what a person considers "sacred" everything falls apart. (not just here, as the automation discourse is happening globally)
edit: typo sacred
A lot of American life is tied down to your job however. That seems to be the problem: health care is tied to your job, retirement benefits are often tied to your job (401k or pension), etc. etc. One can argue that its the overemphasis of your job that's the problem... not the grossly improved productivity due to inventions.
At some point, we have to move past the concept of jobs as the measure.
and the question becomes "who" does the providing? In the US, the going mentality is 'user pays' (which leads to insurance, which leads to insurance company dealing with healthcare provider, which leads to employers providing good insurance as a side benefit to compensation).
The US populous can't easily change their mentality to "society pays". Therefore, this is where the issue should be tackled.
There are a lot of potential mass-market service industries that don't exist because all the benefits of productivity growth (and slightly more than all, now) go to the top 0.1%.
There are still plenty of potential jobs - jobs that we will never have because of inequality.
And we should keep the eusocial aspects of jobs in mind. They provide life structure, self respect, and respectability in the eyes of others as well as a means of subsistence.
Edit: I would love to live in a world where tens of thousands of people have "street party organizer" as their job.
People that complain about ACA increasing prices are complaining about having to pay more since more people are getting healthcare. ACA also requires younger people to pay disproportionately more than older people.
Health insurance premiums did go down from infinity to somewhere between $400 per month and $1,200 per month for people who previously couldn’t buy it.
See here for one such example: https://pnhp.org/news/impact-of-aca-on-employers-and-their-e...
Some relevant highlights: * More than one-third of organizations now have increased out-of-pocket limits, in-network deductibles and/or participants’ share of premium costs in response to ACA. More than one in five organizations have increased copayments or coinsurance for primary care, increased participants’ share of prescription drug costs and/or increased the employee proportion of dependent coverage cost. * The excise tax on high-cost group health plans (a.k.a. Cadillac tax) is considered the top ACA cost driver beyond 2015. Since 2011, a steadily increasing percentage of organizations has taken action to avoid triggering the excise tax — a trend likely to continue. More than one in ten organizations already have adopted changes to prevent them from triggering the tax, 21% are working on changes and 28% plan to act sometime prior to 2018. Only one-quarter said changes were not necessary either because they have no high-cost plans (23%) or because they plan to pay the tax (2%).
That being said, the year following the passage of the ACA saw a large anomalous spike in employer healthcare costs that wasn't present in Medicaid/Medicare. It's almost undeniable that this was due to the ACA.
Someone has to pay for it, and it shows up as increased premiums/deductibles/employer share of premium, but in exchange, we get rid of pre existing condition exclusions, we get in network out of pocket maximums, and older people don’t get hit so hard because younger people have to subsidize their healthcare.
Growth in healthcare costs also slowed in the years after ACA was passed as the various provisions were phased in:
https://www.statista.com/statistics/720767/medical-cost-tren...
We know exactly what it was as we have direct evidence as to what actions employers took in response to it e.g. increased premiums, increased out of pocket minimums, reductions in coverage, etc. The result is that, pound for pound, the same healthcare is more expensive.
> Growth in healthcare costs also slowed in the years after ACA was passed as the various provisions were phased in:
This is unsupported by the chart. In fact, it shows the opposite: the largest decreases in healthcare spending growth occurred before the passage of the ACA. That aside, the parent comment is seeking the causal effect of the ACA on healthcare costs, not a low correlating trend.
And even if it did, cost isn't the only measure of success. It's main goal was to increase access to medical care, which I understand it was successful at.
Not really. It made people better off in the way it was designed to do. It was always going to cost more to give more people access to medical care (e.g. those with preexisting conditions). Rather than "making people worse off", it set the trade-off to a different point than it was before.
As well as the economic value of saving the lives/careers of people who go on to contribute to society in excess of the cost of their treatment.
I'm not saying I've seen a study proving that helping sick people is a net positive for society, in dollars and cents, but it should at least be considered...?
That may not matter to those making these automation decisions. Ultimately, ownership is about power and control, not about actually satisfying the needs of the masses. As long as the owners at the top maintain their power and have their personal wants met, they may find the situation acceptable where the market ignores an expanding pool of penniless masses and focuses on a shrinking pool of other elites who still have money (see North Korea for a similar stable result from a different economic system).
What happens when most of what we use isn't made by humans or ever really owned by any consumer?
Everything you have will be only a service... and sometimes people will get locked out of their accounts. Sometimes it will mean losing your videogames, sometimes losing your e-mails (and contacts) and documents saved in cloud... and sometimes losing your home and money. That will be fun.
Societies get richer by automating things. How much money people have to pay for things is completely dependent on how much they produce, and as a consequence, how automated is their work. The more automation, the more of everything going around.
But automation has a strong short term effect on wealth distribution. How health your economy will be by 2030 depends completely on how your society deals with this effect. And "stop automating" is not a viable course.
How much were pocket supercomputers back then? Or cordless phones? Or chicken pox vaccinations? Or cars designed to protect childrens' lives and health? What did almond milk cost in Duluth? How much were encyclopedias?
Some things are absolutely more expensive, but we definitely have improvements since then, and automation (like modern logistics) contribute to almost all of them.
To put it another way, I would rather have the median income today while doing day-to-day activities now than any past decade. And to the extent that there are downsides in those decisions, I think it's odd to say automation is to blame versus other forces like anemic growth in housing supply.
Oh, sure, you've got the risk of an unemployable underclass full of dystopian ennui. But before the guillotines come out, the .01%ers will have lots of cash for buying art, personalized AAA video games, solid gold plumbing fixtures, and the like.
There's good reason to think that we need to reconsider the value of human life in an automated era, when there is plenty of stuff to keep people alive if only we could distribute it just a little bit. We've actually been doing that for well over a century, and there's good reason to think that we've avoided the torches and pitchforks so far only because of that.
The aggregate demand of "society" needn't go down without it. It just might behoove us to consider that the median might be more important than the sum.
It doesn't. Investing wealth creates more wealth. Of course, the instigator of the investment gets a disproportionate reward for such investment, but taxation, and the general increase in technology level due to such investments will benefit society at large, even if the poor is an unrelated party to the investment.
Therefore it's implausible that investing innately creates more wealth - there is an opportunity cost.
(In this case: the opportunity cost is first giving everyone a decent standard of living, so that everyone can have wealth and invest wealth and create more wealth, as you say)
except for spending it on consumption.
> ... implausible that investing innately creates more wealth - there is an opportunity cost.
the _only_ thing investing wealth does (or hopes to achieve) is to increase the amount of wealth that exist, and most importantly, capture that increase as much as possible.
"giving everyone a decent standard of living" is indeed an increase in wealth, but this increase is uncaptureable by the investor. Therefore, i will say that this sort of "investment" is more closely described as charity than investment.
That doesn't pass the smell test. For instance: it doesn't help the 99.9%ers if one .01%er pays another 10x for a Picasso and a 0.1%er makes a commission on the sale, and I don't imagine the elite demand for new-art production would increase much from current levels. Production of personal goods, luxury or otherwise, for elites can't fill the gaps caused by automation, since they already buy them today. Even over-the-top conspicuous consumption won't cut it either (e.g. employing tens of thousands in North Korean Mass Games-style performances just to employ tens of thousands), is getting into pretty dystopian territory.
The richest ones will also buy their own automated armies, so there will be no need to worry about the guillotines.
I personally don't believe this to be true - humans are insanely adaptable. Learning new skills is part of it.
If an AI is able to automate and produce everything humans need, there would still be jobs that need a human. For example, prostitution. Or premium service jobs like butlers, etc.
But it’s not all golden. There were a lot of stories of spousal and child abuse but they stay hidden.
Most of it is via generators and gas-powered appliances, as they won't have electricity running into their homes.
I'm not an expert by any means but it's just not correct that they stopped using technology developed after the 1700s.
I agree with your larger point that people may be better served by simpler technologies in many instances.
Alan Watts has some interesting (and prescient) writings about how misuse/misunderstanding of new technology leads down a hostile path to societal destruction. Sounds like you may find it worth reading.
No, the Amish specifically have decided to carefully review new technology on a moral cost/benefit level before deciding whether and how to allow it in their communities. They don't ignore technology, and don't have any "cut-off date".
But also so nobody can tell you're using it, who you're talking to or what you're saying.
That's assuming new jobs are not going to replace the old ones. Human society has a tendency to use up all the available capacity by scaling up its goals. Do you think we'll stop inflating our expectations?
Even if our expectations don't drive up demand for new jobs, the government can. The transition to full automation, not to mention the global warming crisis, are hard work and require lots of jobs.
The premise of the article supports this idea - the decline in jobs happens mostly in companies that avoid automation because they lose out on quality, price or are less innovative.
In the extreme case that companies automate almost all jobs, then we're left with billions of people with no ability to earn an income. But they have lots of free time and unfulfilled needs. So they will use that sweet automation to solve most of their needs by working for themselves and being less reliant on external help.
There's always work to do when there are people, especially when they are forced to be self reliant. Of course I don't wish self reliance to look like North Korea, hopefully we can do this in a humane way.
https://www.theatlantic.com/magazine/archive/2018/06/the-bir...
If the FED doesn’t succeed in fixing the productivity to compensation ratio ( https://wtfhappenedin1971.com/ ), there will be more revolts by people who cant afford to send their own kids to the dentist.
This came up on HN a while ago: https://apenwarr.ca/log/20190926 . The thesis was: "To paraphrase the book, the job of an executive is: to define and enforce culture and values for their whole organization, and to ratify good decisions. That's all. Not to decide. Not to break ties. Not to set strategy. Not to be the expert on every, or any topic. Just to sit in the room while the right people make good decisions in alignment with their values. And if they do, to endorse it. And if they don't, to send them back to try again. There's even an algorithm for this." All of that sounds automatable to me using current or near-future technology.
Managers, especially middle managers, also do many other things; but they universally end up spending a whole lot of time in meetings. I'm not a manager, but my own role in many meetings feels automatable. In a handful of meetings I have to be there because I have detailed knowledge about the subject to share. In the other meetings I think it would honestly be fine for an AI that had learned to emulate my writing and speech to represent me in the conversation and write a summary for me afterward.
I'm not arguing that managers _will_ be automated anytime soon, just that AI probably _could_ automate some of the things that take up a big chunk of managers' time.
The response of many company owners has been to forge their employees records. No way to clock-in and clock-out is provided, instead employers craft all this information and keep it to themselves. Everything is exactly as it used to be.
I'm one of those people who think AI could replace many management jobs, since my manager struggles at things a computer does easily, such as calculating annual leave. Many days all my interactions with him could be improved by a well-crafted bot. The main issue is that many companies are actively fighting AI and management automation, since it would expose the schemes that keep their business running.
On the other hand I think it's good to have a manager to talk to, also someone that deals with conflicts at work and someone who is able to reverse toxic dynamics. No way this can be automated in foreseeable future and neither do I see that "self-organized teams" find any way to solve this. Also dealing with product requirements is hard to automate.
I think this would be a better headline. If "middle-skilled" refers to manually moving information across an organization using people, systems, and spreadsheets, then yeah, those jobs are very much the bottom of the iceberg as far as automation goes.
So I guess from that perspective, an employer with many managers who manually manage scheduling/time tracking/etc. could consider deploying an ERM/ERP product and consolidating the org structure an efficiency.
Manager as leader though is harder to consolidate, both during growth ("we're adding resources to your team, you should be happy") and through adding offshore teams without a project manager, cutting down on the bandwidth of the onshore manager.
But, yes. If everything you do goes in and out over a wire, replacing you is easier.
Span of control, the number of people directly supervised, has been increasing for decades. Thus the "flattened organization", with fewer middle managers.
* All of their regressions are in levels.
* Separate out the long-, medium- and short-term consequences: * Investment in new technology may require hiring new "talent" that is knowledgeable of the tech.* In cases where there is a tech shock and firms are all buying the same machine, demand for a certain skillset will grow quickly. So you'd likely see:
* It is important to ask, "what might have caused investment by a given firm over a given time frame?"* Common industry technological shocks might spur investment.
* (And the simplest way to account for technological shocks in this sort of regression you'd want `[industry] x [year]` fixed effects.)
* "We went on a manager hiring spree in year 1. Those managers, in turn, bought a lot of robots in year 2. In year 2 we hired fewer managers because we'd just hired a bunch in year 1."
* Reverse causation issues are pervasive.
* Finally, I think it is important to notice their notion of "robot adoption":
* They are talking about physical hardware _only_.
* Their measure of investment is coming only from import data. The set of firms that are going to go through the rigammarole of importing physical hardware is probably raher special: Probably:
At the very least, this paper needs first-difference regressions and a good instrument that can help to disentangle what might be going on. For the instrument, it might make sense to look into changes in tariff duties over time [1].---
[0] https://content.tcmediasaffaires.com/LAF/lacom2019/robots.pd...
[1] https://cbsa-asfc.gc.ca/trade-commerce/tariff-tarif/2020/htm...
Humans are self replicators, so are the other living things, culture (+memes), technology, companies and the economy. If we could have self replicating factories, then all we'd need to do is to import such a factory to a poor country and let it run. Of course automation is just half the problem, the other is raw materials. Presumably we could recycle almost everything and not need an unlimited quantity.
I bet we'll reach technological self replication singularity before AGI, and the impact is going to be huge. Robotics, 3D printers, IoT, sensors, computer assisted manufacturing and logistics are going to be interesting in the next decades. I just hope they will be open source and not restricted by patents.
So a more interesting question I think is: how many companies out there are intelligent enough to go against the axioms of corporate culture and investigate value from the recently so popular term “first principles”? I’m unfortunately cynical enough to believe the answer is “not many”.