My take on this is that people are used to paying very low prices on consumer goods because of what I’d call “The China Effect”. Goods made in offshore or otherwise low cost societies flood the market. Modern consumerism makes us all “need” much more than we actually do and the prices in big box stores and other discount outlets encourage this behavior. Peoples inability to critically think about their actual needs versus their wants, in addition to the diminishing purchasing power of the middle class means that they have a skewed view of the value of something as opposed to merely the price of the item.
I work as a machinist. I’ve owned a shop and worked in this capacity for years. One type of customer is another business owner who can view the cost of a replacement part for a machine he uses in his business as an important facet of his profit making ability and judges the cost of said part accurately. Another type is someone off the street who needs a part made for an item in their hobby or home. They are used to Wal-Mart and internet pricing on mass produced low cost consumer items and have no idea of how much time and effort goes into making one of anything. Guess who gripes and bitches the most about how much the part is going to cost? I’ve actually told some of the latter folks, after they freaked out about my cost estimate to hand make a crankshaft or piston for their 1904 Hoyt-Clydewell (or somesuch item) that they should check Wal-Mart or Amazon. “I hear they’re having a sale on that very piston”. “Really?” No…JK
There is going to be a very long tail when it comes to automation. There will be a few industries that collapse quickly and comprehensively (like truck driving), several that take a bit longer (several decades) and are never really completely automated (surgery perhaps), and some for which there’s no evidence that we’ll ever significantly automate (writing, programming).
Basically, we have no idea yet how to build a general purpose intelligence. What we can build are special-purpose intelligences. Driving can be decomposed into small, solvable problems, and so if you throw several thousand man-years at it, you can solve it to a “99%” level. But that approach doesn’t work for anything that requires what you might call improvisation.
Also, some things won’t be automated because it won’t be worth it. Even in the Tesla factory, significant parts are assembled by hand. Not because they couldn’t get a robot to do it, but because the work is just fiddly and would take too much effort to really automate well. Snapping a bunch of odd-shaped trim pieces into place is just faster with a dextrous human.
At any rate, I’m certainly not frightened for myself (I could manage even if my job were replaced tomorrow), but I do worry a bit about 10 or 20% of the workforce being displaced in a short time period. There will be more career options, but change is probably happening too rapidly for the kind of transition that’s needed.
I actually agree with this approach for other reasons and so this is just one more argument. But it’s not mutually exclusive with training and education. Silicon Valley needs way more engineers to solve just auto-driving, let alone the numerous other problems out there.
This job vacancy just came across my email, for a Senior Systems Admin. I am not familiar with this particular career, so perhaps this isn’t unusual, but when I saw it my eyes fell out of my skull and fell into my gaping jaw, Beetlejuice-style.
[QUOTE=Job Application]
Specifically, you will be responsible for:
Managing and assigning workload to at least one IT Support person
Windows Active Directory 2008 R2 administration - Servers and Workstations
OS X administration of edit and composite suites
Optimising network infrastructure for best performance in digital image workflows
Managing multiple networks LAN, VPN, WAN
VMware Administration
Backup and Archive on LTO Libraries
Developing and applying operational practices and procedures for Production system technical support.
Monitoring, troubleshooting and reporting on systems reliability and redundancy
Maintaining installation, configuration and systems documentation with best practice procedures. Defining policies and reviewing procedures for servers and workstations
Project related systems solutions to support changing business requirements
Maintaining licence register and renewal timeline (in conjunction with IT Support)
Maintaining asset register(in conjunction with IT Support)
Arranging any new asset purchases as approved by Management team
Tracking software installs, hardware usage and workstation assignment
To be considered for the role you will need:
prior experience in Systems Administration or as Systems Engineer within digital post production
experience with administration of Windows Domain Services
to be comfortable juggling and shifting priorities and exercise excellent organisational skills
demonstrated experience with networking, firewall security and VMWare infrastructure maintenance
be happy to work flexible hours in order to perform upgrades or system changes
have great project management skills, with a strong ability to define future technical needs and manage through sign-off and implementation
be easy-going, flexible and enjoy having a laugh
to be able to work autonomously, efficiently and effectively and recommend sound technical solutions
excellent communication skills to communicate with management, employees, clients and suppliers (both written & verbal)
an outgoing personality, be vocal, inquisitive and a natural troubleshooter
extensive OS knowledge and experience in OSX, Windows Win7, Windows 2008 R2 – 2012 Serverr
3 years minimum experience in SAN, SAS, iSCSI and FC Connectivity; LAN, VPN and WAN; Active Directory NAS Administration
an ability to quicklylearn new technologies
a positive ‘can-do’ attitude and an ability to work well under pressure – with the skills to identiffy issues before deadlines are not met
Additional skills we would love to see include:
Python, BAT or Shell Scripting ability
Exposure to 2D/3D applications such as Maya, 3DS, Nuke, Photoshop and After Effects
Renderfarm design, build and management (Deadline).
Camera formats: RED, Alexa transcodes and proxy generation
Post Production workflow: Camera, comping, colour and finishing for TV
Experience managing a team or another individual
An inquisitiveness about the post-production industry and new technology trends
[/QUOTE]
That’s a pretty typical systems admin job for a small company, and most small business systems engineers will be able to check off most of those boxes. I work with four of those guys every day that fit that description - except they haven’t worked in what is obviously some sort of print house. It sounds like more qualifications than it really is.
The issue with that posting, a small print shop won’t be able to pay competitively against Fortune 500, and someone with that skill set could work for a big company for $10,000 more. So you do need to find someone who is willing to get paid less for a small business environment - and then need to treat them well enough to stay (IT unemployment in Minnesota is 1%).
In my small/medium sized company experience, that’s why you end up with marginally qualified people; there’s not always the strict pay scale privacy you see with bigger companies, and execs get pissy about it.
Say a company of 200 people wants to hire a new sys admin and pay him $85k. It all sounds well and good, until the CFO gets his underwear in a twist because his senior managers who have 15 years of experience in the company are only paid $75k, and that’s unfair in his eyes. Ultimately, it becomes a revolt, and the CEO/CIO end up hiring someone for 65k to keep everything ‘fair’, and then they have some jerkwad who is willing to do that job for that much money, and then everyone bitches that their IT support sucks, or that the IT guys never stick around, or that he doesn’t answer his phone after hours or whatever.
“What the market will bear” is pretty much a truism when it comes to this kind of thing. Landscapers make minimum wage or below, because there are a metric shitload of them, they’re all interchangeable, there’s no premium for experience, and literally anyone who can heft a weedeater or push a lawnmower can do the job. Similarly, a guy with 15 years experience with big iron UNIX system administration is going to cost much, much more, because there aren’t very many of them.
One thing that’s always surprised me is that most companies won’t pay a premium for a simple job, but one in which the difference between a good employee and a mediocre one is huge. Data entry is one example. I generally did 3 times more than everyone else. That eventually landed me in IT, so I did get my just rewards eventually, but what’s wrong with paying double for someone who is going to do 3 times more and carry fewer employees?
I had the same issue in the pizza business. A good pizza maker can produce up to 3 times faster than an average one. Instead of me employing three guys at minimum wage, why wouldn’t corporate let me employ one at double minimum? They’d still be saving money, and if a person is good enough but not the kind of person who is good at learning complex skills, then a person could reasonably support themselves for life off their top notch pizza making or data entry skills. I’m sure the same dynamic works in a lot of relatively simply jobs: some people just figure out how to do things more efficiently and become a ton more productive, but usually don’t get more pay for it. With data entry especially I really don’t understand it. Employers will demand X amount of keystrokes, but the pay rate is seemingly totally disconnected. So I’ll have one ad that wants 8000 keystrokes for $10/hr, and another that wants 10,000 keystrokes for $9/hr.
Well how fast the clock is ticking for particular jobs is a matter for plenty of debate. But the thing is, as computer software “learns” to do one job, the code will be transportable, at least in part, to other jobs. A piece of software that can manipulate a robot hand to do one thing can manipulate it to do another thing. So I expect to see a lot of acceleration once things really get going.
For almost all jobs, you don’t NEED general purpose intelligence, most especially the office and service sector jobs that compose the bulk of our economy now that manufacturing has poofed and blown away. Here’s the thing: many people are prone to dismiss concerns about automation of the workforce as groundless, because “people will find something new to do for a living, just as they always have.” But even as people are looking for new jobs, the people who own the ever-improving software and robots will be looking for new markets for their products. I wonder who’ll win that race?
At present. See above.
Well that is the key issue, isn’t it? There’s nothing more logical than that the software and robots will improve and gobble up ever larger chunks of the labor market, the question is, how fast? If it’s gradual, moving very slowly at first and taking maybe a century, we may be able to make an easy transition to a laborless post-scarcity society. I wouldn’t count on that. I see it as much more likely that we’ll be looking at a robot job holocaust within 20 or 30 years. 50 percent unemployment? If we’re lucky!
And the problem is, I don’t see our social values changing. Particularly the conservatism and libertarianism that dominates the One Percenters’ world views. With increasingly large portions of the world’s wealth going straight to the One Percent, that’s a very bad thing indeed. They, and this includes the ones who inherited all their wealth, think of us middle class Americans as weak-willed, short-sighted, feckless losers who lack the brains and moxie that they have – in short, they think of us as peasants. I believe they will face the prospect of the bulk of Americans going homeless and/or starving with a certain equanimity, in the sense that they won’t give a shit. Their major concern abouit us will be that the peasants don’t come after them with pitchforks and torches, or in the US, our many, many, many guns.
Sure, there will be jobs for programmers and some others long after the robot job holocaust puts most American’s out of jobs. And yeah, you and I will likely be retired or dead by the time the shit hits the fan. But I dunno, I would like to hand off to my son and his prospective children something other than a world with a climate run amuck and no jobs to be had, and a social climate in which starving to death is the Thing To Do. YMMV, of course.
It’ll never get that far. In a world where robots take most of our jobs, the economy will be so productive that giving everyone a lifestyle commensurate with what the upper middle class enjoy today will cost a relative pittance. Working will be for those driven to accomplish things, or who just want to get really rich, while everyone else enjoys a life of leisure.
That’s the kind of socialism that is probably inevitable and which is hard to argue against. But as long as we live in a world where the vast majority who want work can get it, and where we NEED human workers to function, the carrot of wages and the stick of poverty are still necessary for the system to work.
Sure, we will probably get to a post-scarcity society where robots do almost all the work and all the uninteresting work. It’s the transition that could be problematical. And “it’ll never get that far” is mindlessly optimistic, if you are at all familiar with recent human history. Hell, current human history.
“Work” in such a scenario would likely consist of voluntary pursuit of the arts or scientific research.
Of course “post scarcity” evangelists always act as if human labor is the only resource (i.e. “replaced with robots”). Robots require energy and manufacturing anything requires physical resources (metal, chemicals, etc). Not to mention other finite resources like drinkable water, arable land, even just finding decent locations for people to live (i.e. everyone can’t live with views overlooking Central Park or the beach).
Not to mention that in the future, people wouldn’t want to live a 2015 “upper middle class lifestyle” any more than we would want to live an upper middle class lifestyle from the 1700s. You’d look like a jerk trying to entertain your friends with a 3D television when everyone else’s TV has 6D.
Anyhow, my point is that “post scarcity” is a bullshit concept. At least in the context of the modern economy within any reasonable timeline.
Skills gap is the same thing as lack of available jobs. Society didn’t run out of work to do. It’s just that the work that needs doing isn’t the same as it was 20 years ago.
It is optimistic but I don’t think mindlessly so. If we’ve got an economy generating say, $100 trillion in output in 2100(in today’s dollars), but with 80% unemployment, all it would take to give that 80% a living standard equivalent to $100,000/yr would be a 20% average effective tax rate. And $100 trillion is being pessimistic. We’ll probably hit that GDP level by 2070 assuming current growth projections.
Of course, if growth is slower than that, or robots take over most of the jobs long before then, we will have a very difficult transition.
Of course not. In any case, goods won’t cost much, that’s not even a problem for middle class households today. It will probably be even less of a problem in the future. Non-working families will probably have all the same goods rich households do, probably just not as awesome. Where the middle class struggles is in paying for services provided by other people. Which hopefully will also be taken care of by robots. Get rid of the human doctors and educators, replace them with robots, and suddenly medicine and education aren’t so much of a financial strain anymore.
Of course people will WANT more than that, but if they do, they’ll acquire the super advanced skills needed to live that way. But no one will need to live in poverty.
Imagine a business owner who needs a piece of machinery, but “can’t afford” to pay the going rate for that equipment. And whines about it. And complains about the manufacturer expects him to have a stash of spare cash for that machine. Total yutz, right?
Your employer is just as much of a yutz. We’re not talking about them bumping salaries out of the goodness of their hearts. We’re talking about them raising salaries until they are competitive enough to hire the people they need. If they can’t afford to do that they better figure out how to improve productivity until they can, or else go out of business. And good riddance to them.
I have no trouble hiring scarce high level technical talent out of universities. That’s because my company pays a ton in starting salaries. I worked for another company where I couldn’t get anyone, because my boss cut the salaries for new hires HR proposed. Yes he was an idiot, and yes I left as soon as I could.
Employers who can’t find talent these days have only themselves to blame.
Krugman notes that if there were a real shortage of STEM people the salaries for STEM jobs should be increasing - as salaries for IT jobs did during the bubble. So, pretty much in line with your OP.
If you have someone you want to hire and need to get a visa for, in the old days at least you crafted an ad where the requirements matched this person perfectly. If by some odd chance the guy had HVAC experience on it went. Ditto weird language requirements, and weird software tool requirements.
There is another factor which employers have given up on these days, which is training. They want a school or other employer to teach a person about their specialized tools, and then hire that person. And they complain that the schools teach the basics of a field and not whatever is hot today.
I’m in an extremely specialized area. The number of people who could sit down and do the job from day one is very small, and they’d be very expensive. Instead we hire people with the basic background from universities and don’t expect them to be fully productive for six months. And we have very good retention. If employers were willing to invest they’d get good people for a lot less money.
I need very elite people. Now, most resumes I see don’t have the basic background, but it is so easy to submit these days getting rid of them takes a second or less each. I have not seen the bimodal distribution you are talking about. Maybe I do a good job filtering. Maybe 30% of the people I phone screen don’t got it. Almost everyone we invite in would be able to do it - and we’ve hired enough of them that I’m pretty confident I’m right about the ones who turned us down.
We do pay good money - but that determines if they take the offer, not if we make it.
What specific part “is not true”? That temp agencies take a few dollars per hour of every hour you work? That if you are making $17, the company pays them $20… per employee…per hour worked. Specifically list where this is not so. And no, I’m not going to post camera phone shot the poster behind the desk of the recruiter at a temp agency stating that $3 per hour was required “or they don’t get hired”.
Its an unnecessary expense that lowers take-home pay artificially during a time when wages are artificially deflated far enough.
…and crackpots who think that just because they can run a payroll they are Ogs Gift.
I don’t, at least not past the first phase. We aren’t going to get a singularity where we invent strong AI that then automates everything in a short period of time. We’ll pick the low-hanging fruit and then it’ll be a long, hard slog to go further. There might even be some false starts.
I think it’ll be a bit like the space race, actually. If one extrapolated our progress from, say, 1955 to 1969, one might have expected that we’d have a Mars colony by the 80’s. Didn’t turn out that way because the number of problems to solve just grew exponentially. Even now, the technology has just barely progressed to the point of making it possible with an incredible amount of effort.
We’ll just have to disagree on the rate of progress. I can just barely see 50% of current jobs becoming obsolete in 30 years. But that’s also an entire generation of workers who can choose a career that won’t be obsoleted.
I also think there will be large classes of jobs that will stick around due to social and/or political obstacles. I would say the way classes are taught today is obsolete, and should be replaced by online systems. The tech is simple and here today. But the change isn’t going to happen universally anytime soon.
Who you callin’ they? I slip in just under the 1% wire, thankyouverymuch. You best not be getting uppity, peon.
That’s enough, isn’t it? The only reason I support police departments, for instance, is because I don’t want criminals wandering around and making trouble for me. Beyond that, it’s a pure net loss. Likewise, basic income keeps the unemployed masses from turning to criminal activity, and is cheaper than keeping them imprisoned. Whether you think it’s the moral thing to do or just a form of bribery to keep the peasants at bay, it’s still a good idea (and popular even among the rich in this part of the country).