60 comments

[ 3.6 ms ] story [ 124 ms ] thread
A useless article since we don't get to read the actual paper.
Yes. I wonder if they just looked into how many of jobs as they are right now will eventually be repacable. That would be wrong, as how we do a job has always changed, meaning that when we get to the point where job X would be replacable, it has changed so much to not be replacable. Also, if your job really gets replaced by a computer, there will also be tons of new types of jobs created for you to take.
I find it disturbing that we are stuck with a type of economy where this is a problem. The purely rational response: yay, now we don't have to work so much! But this never happens. Why? Why isn't there a market pressure to reduce the number of hours that people work (for the same pay) once more things get automated? Why is it that technology could bring us closer to heaven, but we choose over and over again to live in hell? How could the right incentives be created?
I think about this a lot. I think it has most to do with that people don't only care about absolute wealth, but also relative wealth - keeping up with the joneses, if you will. As long as social status is connected to wealth, this will be a problem.

Personally I advocate a gradual decrease in allowed work hours - 10 minutes shorter work day every year, or so. Although France has tried this and it doesn't really seem to work.

I think 10 min/day/year is too quick of a timeline. Perhaps 20 minutes a week per year would do. In 12 years we're down 4 hours. In 60, we're down to half what we are now, which is about where we should be, IMO, and is a long enough time to get excessive automation in place.
You can live cheaply, by opting out of the rat race. Though you do need to be pretty resourceful. e.g. http://www.atariarchives.org/deli/cottage_computer_programmi... It's just that most people don't want to, we are social animals and mostly define ourselves relative to the herd.

A very different application of automation to the relative-wealth issue: http://partiallyclips.com/2003/09/25/dome-house/

And sometimes, the herd tells the individual what to do.
I think you could compare this to sports or other game-like activities where success is by nature relative. There's no 'need' per se for athletes to practice 40 hours a week, but they do it in order to win. And the amount of time needed to win will almost always be right around the theoretical limit of how much a human can physically practice. There's an equilibrium in competitive practice that is very tough to regulate.
If for each two jobs automation removes, it creates one new job. The one new job pays less than two previous jobs. This changes the balance between capital and work. Workers share from production has been in steady decline in all OECD countries over two decades.

To achieve the goal of working less, you need to own capital. Technology changes the capital vs. work equation towards capital at the same time when there is increasing global supply of educated workforce. It's just supply and demand. Workers are simply competing towards bottom.

To get to position where you need to work less you would need to have:

1. Change taxation from work to capital.

2. Basic income.

The new job pays less but also raises the margin which in a healthy market lowers product prices. Purchasing power increases and everyone is happy.
Even if we had a healthy market, which we don't, the people who lost their jobs don't have "marginally less" income. They have zero income. They are not happy.
> If for each two jobs automation removes, it creates one new job.

Why do you think it's as many as 1:2?

That sounds like a lot of new jobs to me; How many travel agents did hipmunk and expedia et all shut down vs. how many employees that they have? One robot maker and a couple of robot service-people can make and maintain enough robots to remove a hundred jobs. Or maybe a thousand. I don't know, and I don't pretend to.

The result of growing inequity is that too much work remains undone, too few opportunities exploited.

Big money requires big investments for the big returns. Big funds cannot consider many worthwhile investments, due to transaction costs, the (relatively) lower expected ROI.

If the money keeps moving (the velocity of money) through progressive taxation (decreasing inequity), then more people will have access to more capital.

One reason I keep reading pg and HN (and watching crowd sourced funding) is because I'm fascinated by the effort to reduce transaction costs. I'd prefer progressive taxation. But this seems to be working too.

>Why is it that technology could bring us closer to heaven, but we choose over and over again to live in hell?

Working people don't choose. People in power and corporations choose. And the constant is that they want to expand the number of hours and reduce the pay.

The extra productivity that automation brings doesn't matter. It's not like they target X levels of productivity. They target "as much productivity as we can get, up to infinity".

Plus, what would they do with a population that had time in its hands to think and be involved instead of being a wage-slave bound by debt?

(That's not some dystopian tinfoil concept. It's a standard concern of those at the top of any given society from the age of the Romans ("panem et circenses"), to Machiavelli to modern era think tanks).

Thanks to competition people in power don't choose either. Higher margin will eventually lead to lower prices.
>> Higher margin will eventually lead to lower prices.

As long as countries don't have large economic areas based on corporativism, oligopolies and regulatory capture.

More work against that is needed from economic liberals in e.g. my native Sweden. (You won't see much criticism from media and politicians in such a situation.)

I agree, but I think it's mostly mutual - there are lots of opportunities for part time jobs but most people prefer full time employment.

In fact, in Sweden there's a law that allows parents of children under 8 to work 75% of fulltime (30 hrs/week), but very few make use of it. I suspect that part of the reason is a fear of annoying the employer though.

>I agree, but I think it's mostly mutual - there are lots of opportunities for part time jobs but most people prefer full time employment.

That's because they also prefer to have a full salary. It's not like employers, considering the increase productivity technology affords, offer full old-time salaries for part time work.

If anything it's the inverse: buying power has dropped since the seventies (only partially offset in some areas with cheaper and more capable electronics, but not for things like housing, food, etc).

I've seen it quoted that if Americans could accept a 1946 standard of living, we could have a 2.6-hour workday - as it is, some of that 5.4 hours goes to extra cars and televisions, and most of it goes to bolster the income of the suffering rich.
It is already happening. Automation = lower product prices = you don't need to earn as much money = start working part-time which is a perfectly viable option today.
Working part-time at minimum wage is viable?
Because the wealth is distributed upward, to a noble class that burns it all day for fun. As technology gets better, that class gets bigger, because we manufacture so many goods to burn that we need more nobles to burn them. Due to this process, in the year 3841 the entire population will have become nobles, and the sky will be black with smoke.
Orwell in 1984 addresses this subject extensively. People working less or not working are a problem for the government. In the book they invented a perpetual war to burn the surplus which would allow us to work less. In real life consumerism seems to to a pretty good job in keeping us busy instead.
War.. Most of the time I have been alive, the U.S. has almost had a military engagement someplace. What percentage of middle class' taxes are fed to the for-profit machine that in turn feeds the military?

War serves another purpose, too - wealth transfer, and that helps keep the work week at 40 hours.

I think Orwell's point was that excess labor can be funneled to war instead of work but so far that hasn't panned out. If anything, war is also becoming less labor intensive and more capital intensive due to technology.
Your comment would make sense if military spending were an order of magnitude higher. As it is only 5% of GDP, it is hardly enough to have much effect on total hours worked.

  > But this never happens. Why? 
If you pay an accountant $500/year to do your taxes but all that work could be automated for almost no cost, would you still pay your accountant $500/year?
No, because the rest of the world isn't doing that too. But if the WORLD doesn't need everyone to be doing crappy labor, why would we invent work for them to do, instead of just sharing in the new wealth created by our technology.
Suppose you have a job that pays $2000 for 40 hours a week, ie $50/hour. You might want to work only 30 hours, and get paid $1500/week. But your employer has fixed costs per employee: insurance, and the IT, HR, finance and management functions. This can be as much as 2X, ie another $2000/week. So they are currently paying $100/hour for your services. But if you go down to 30 hours and get paid $1500/week, they would still be paying $3500/week, so their cost would go UP to $116/hour.

You might be able to reduce all these administrative costs, but splitting work up among more people has its own cost. Remember Brooke's law, "adding manpower to a late software project makes it later" - the same effect is in play.

[Editted to fix arithmetic. This makes the example less dramatic, but still significant given the profit level of many companies]

> they would still be paying $4500/week

I don't understand this. If you are paying:

  $1500 for a 30 hour week
 +$2000 for administration costs
 -------------------------------
  $3500
Am I missing something on how changing the $2000/week pay to $1500/week makes the price go up by $500/week?
It's a typo I reckon, but the original point seems valid to me. It would be better to hire one person for 40 hrs than 2 for 20 hours each because of the overhead.
Yeah, sorry, it's a typo. Will fix.
it's per hour costing:

  2000 /wk labour
  2000 /wk overhead
  ----
  4000 /wk @40hrs/wk
  100 / hr
vs

  1500 /wk labour
  2000 /wk overhead
  ----
  3500 /wk @30hr/wk
  115 ish / hr
Demand for labour is decreasing while the supply of labour is fixed (or increasing) - everyone needs to work, because everyone needs to eat. Like others I think a basic income would go a long way towards mitigating this problem. Imagine a world where pretty much everything we need is automated - what do we do then? In the long term some form of basic income is inevitable.
> Why?

What good is it being rich without poor people for contrast?

Or, from another angle, I suspect (without any formal economic training) that hyper-rich with private jets is not possible in a society where most people are moderately rich, because there wouldn't be excess cash to monopolize.

So there's no question that automation creates economic opportunity.

If I can get the same product at a lower price, it frees up some of my income to buy something else. That creates demand, it is is demand that creates jobs.

The problem is that workers aren't fungible. People spend years or even decades acquiring skills, and then technology can automate away what they do, leaving them stranded economically speaking. They need new skills to keep earning because demand has changed.

Also on an individual level we all resist change. If you like your life and lively-hood now, and technology threatens that, it's not surprising that people will go Luddite in order to avoid the effort and risk that comes with changing jobs.

The answer is to put in place the social institutions necessary for people to be more adaptive and fluid in how they earn a living. Give people a safety net to fall on: social insurance, and a ladder to climb to climb out on: education.

Plenty of nations are using parts of this strategy and the data seems clear it works. But it's a difficult sell politically in the US. People here aren't willing to see the taxes they pay as investment toward bettering society as a whole. And to be fair, they're justified in feeling that way since our government's spending priorities are often insane.

We should definitely embrace the future as fast as possible, instead of trying to hold it back so we "don't lose (the old) job".

The problem is a lot of people need to be retrained, since things are changing so fast, and their jobs aren't lasting their whole lifetime, like they used to. And I think neither the government, nor the private education sector is properly equipped to handle that yet, nor are they making much progress in that. But they need to, and soon, because this is only going to get "worse" (for the people who aren't trained for the new types of jobs).

This is just machines and automation "disrupting" human jobs, which means us humans need to keep moving "up market" (doing stuff machines can't easily do yet, which tends to involve right-brain type of jobs). But it's only a matter of time before the machines/automation catches up to us there, too.

Hopefully, by that point, we'll be having a society where food is easily and very cheaply created by "replicators" for everyone, and it will be a society where money has less importance, and we don't need to "work for a living" anymore, and we can just be free to follow our passions.

I think all of this will happen in five decades to a century, but it will be a very tough ride, as we try to quickly adapt and keep pace with technology, as we try to make a living.

What makes you think the future will be better than the present? If history is the judge then it is more likely that we will end up in a situation where 10-20% of the people have everything and everyone else lives in super poverty relative to them. Most of history and even the present reflect that situation. Also it will require far fewer people to have absolute systems of control than it ever has before.

Or AI could take over singularity style.

I don't know why you assume the future will be better than the past. So far it honestly seems more like computers and robots have created a situation where most of the middle class is now useless to society. We seem to be reverting back to a system where there is only an upper class and a massive pool of peasants. For us to really replace the higher end of the workforce will require a singularity type of event.

Of all the possible futures it really takes a lot of optimism to not see some kind of dystopia.

Just look at the current privacy situation. Look at what the NSA is doing. Look at how google and facebook have been allowed to build a massive database of every action that we engage in with technology, and they are then allowed to sell it to 3rd parties without most people even knowing they do this and sharing none of the profits with us for selling our information. That is the reality of the future. Add largely automated robot armies and much more powerful systems of control plus finite resources.

I don't know why you assume the future will be better than the past.

50 years there were full blown dictatorships in Europe and South America, dozens of millions were dieing in China, and Soviet officials in submarines were blocking the launch of nuclear missils. Oh, and Vietnam war, Kent State, etc.

70 years ago we were in a world war, smashing half of Europeu to pieces and blowing up full cities in Japan, plus, you know, the holocaust.

80 years ago there was the Great Depression, and all of its effects. My grandmother, in an western European country, lived without electricty, running water, education and barely enough food.

And this is just a self- serving European perspective written in five minutes; there were plenty of horrors I probably don't even know about.

The NSA? it doesn't come even close.

If history is the judge then it is more likely that we will end up in a situation where 10-20% of the people have everything and everyone else lives in super poverty relative to them.

The key word here is "relative to them". On the other hand, if history is the judge, then the people living in relative poverty will still be vastly richer than most people alive today.

Dystopia? That depends on how strongly one feels envy. Personally, I can enjoy driving my flying car and stem cell therapy until age 100, even if some rich guy has nanotech rejuvenation allowing him to live to 1000 on his spaceship.

I'd argue that the biggest problem is not new technology and how it replaces workers - it's the increasing rate of change of technology, and increase in speed of new waves that replace / commoditize industries. In otherwords, I worry that we'll eventually get to a place where every month there's something dramatically better/more efficient coming out that will replace the jobs done by regular people. People take time to learn, and re-education for the jobs available will not be instant. That means that if the group of available jobs changes too quickly, we may find ourselves with a growing and unfillable skill gap, since nobody can learn the new skills fast enough to actually use those skills in production.

What are your thoughts? I'd like to pick apart this issue.

This article has a very fearmongering tone, and completely misses the fact that economically, people will still do these jobs for a long time, because replacing them with computers is not an easy or cheap task. Technology has been replacing jobs ever since it existed, and the workers somehow kept up, I see no reason why they wouldn't in future.
Except they don't. They haven't been keeping up for the last 30 years of growing automation and stagnating wages. In that time the world has seen massive population growth, which exacerbates the basic problem just as the information revolution accelerates it.

You can tell yourself whatever you want, but that won't close the massive gap between your perspective and actual reality.

I guess this "software is eating the world" explains github's $1 billion valuation (which doesn't make sense to me otherwise, developer tools aren't as large a market as say, the TV-replacing youtube, bought for around $2 billion).
I wanted to say that YouTube was acquired early in its history and would probably be worth much more today, but after looking into it, the data doesn't necessarily back that up. In July 2006 a few months before YT's acquisition, YT users were viewing 3 billion videos a month[1], and in 2010 it had "only" quadrupled to about 14 billion views a month[2]. This suggests in 2006 YouTube had already mostly caught on. So you may be onto something.

[1] http://usatoday30.usatoday.com/tech/news/2006-07-16-youtube-... [2] http://www.comscore.com/Insights/Press_Releases/2010/6/comSc...

The way I put it to my coworkers is that we are rapidly approaching the point at which we must choose between a society which redistributes the benefits of automation and one where the majority subsist on garbage dumps. But as long as many people don't vote, and a large number of those who do vote are distracted by idiotic "moral" issues we'll drift towards enclaves and garbage dumps.
The power of the people (via voting) you're suggesting relies on a) candidates that are willing to make the changes needed and b) actually being able to enact those changes once in office

Sadly we very rarely see candidates in set A and even less so of set B.

True, but if the voting public paid attention at least there'd be a small chance of fixing things.
UR says it best:

> All subterfuges and evasions to the contrary, the basic economic problem faced by 20th-century governments (somewhat less in the 19C; far more in the 21C) is unemployment. The cause of unemployment is simple: in an industrial economy, most human beings are economically useless. They are not productive assets at all. They are liabilities. For a brief transitional period, they could still be used as industrial robots. This period is close to its end.

[..]

> as the Singularity nears, the future of work becomes clear - there is an IQ threshold below which any human, no matter how cheap to feed, is a liability. Classic unskilled manual labor remains productive in some domains - gardening, housecleaning, and so on. Perhaps this will be true for another decade or two. It will not be true indefinitely.

It's a big problem all right, and we'll start to face it for realz by the end of this decade, IMO.

There's another way to look at this. As automation increases, the costs of running businesses decreases, making it easier for new entrants to compete (assuming the computing hardware and software is easily accessible). So as the workforce becomes increasingly computerised, you could see an increase in company owners (a job that is unlikely to be ever computerised).

There are two main drawbacks to this alternative view. Firstly, there will still be restrictions in place of physical resources (not everyone would be able to setup a mining company, for example), which will govern who can compete. Secondly, there would be limits on how many companies in a certain sector could be sustained. At some point I hope we resolve these issues by moving beyond personal greed.

Let's look at a counter example : amazon. Is it using technology to help online sellers ? or is it using technology to kill retail businesses ? It may well be that the end goal of amazon is to own and automate all retail, and even if it remains as a 3rd party platform, it could have so much power that sellers would make almost nothing.

In an automated economy, where businesses are easily scaled if you own capital, the amazon example could apply everywhere.

hershel, I believe you may have misunderstood me. Amazon is an example of a company that has centralised a sizeable portion of online retail, my argument was that technology can evolve to the point of encouraging decentralisation. Imagine if setting up your own "Amazon" was barely any harder than setting up a Wordpress site. A small warehouse space to store products, automated picking and packing, an easy to setup and maintain web presence, elements of AI to help field basic customer queries. It's all possible.
Say you set up you're "small" amazon. How can you compete with "big" amazon? They have faster delivery times - due to local warehouses. They have wider selection, better prices, better brand - due to their size. I don't see any advantage to "small" amazon.
Firstly, we're not talking about 1 small "Amazon", we're talking about a large number. Secondly, for such businesses to succeed, you do need a shift in culture. If consumers care about convenience above everything else, then larger businesses will always tend to dominate. However, if consumers hold other values in higher regard, then smaller businesses have a chance. The values that these customers hold will determine which approaches are successful. For example, companies that offer specialist knowledge, customisable products, a higher level of personal service, etc... could compete.

Speaking personally, Amazon is my go to source when I don't know much about what I'm buying, and just want something half decent. The reviews and the speed of checkout are Amazon's strengths for me. They're not necessarily the best on price, and if I research further I tend to find better products elsewhere too.

Whether we have this culture shift away from large retailers is up to us. We get what we deserve.

Small pendant:

automation != computerization (digitization)

Automatization is reworking processes to remove judgement. Simplification.

The systems I work on are all computerized. They are definitely not automated. Huge amount of human intervention, decisions, and judgement every day.

Half the jobs?! Whew, that's some chunk of work. I need to get busy...
Fascinating article! I wrote a short story a couple weeks ago that explored one possible impact of such an event from the perspective of a school teacher whose job is replaced by a computerized solution. It's not exactly a happy story, but it might be fun for some readers here.

Link -- http://192.241.231.77/stories/mission/

The silver lining to robots taking our jobs: remaining occupations will much better align with, extend and fulfill us in our humanity.