Comment by AndrewKemendo | Hacker News Reader

Comment by AndrewKemendo | Hacker News Reader

The article doesn't really examine productivity, it is examining wages.

To wit: industrial automation has grown 6% YOY since 2003[1]. Manufacturing is up worldwide and robotic integration is exploding, not to mention software automation.

The reality is that these increases are not labor productivity, but capital productivity. You also won't see these productivity numbers easily on a balance sheet because they currently take longer to find return on value when doing replacement work than scaling human labor. This is largely because it takes time to transition labor force into a new role within a company or out of the company altogether.

At the end of the day though, from a labor productivity perspective, there is no raison detre to pay workers more for doing less unless they are providing more value per time period.

While history has shown show that technology does not eliminate jobs, merely creates new ones, the pace at which this is occurring is increasing [2] and is arguably faster than workers can adapt.

This is also not to even mention where finance is putting investments, being far from what I would consider proportional with respect to consumer goods and industrial technology base.

In the end, there is every reason to believe that as capital takes more share of the work, there will be less work and thus less pay for labor. That is unless the market for labor, in places which are not yet ready for automation, grows in such a way that it can absorb transitioning workers and new entrants. I don't really see that happening without some major changes in the ethos of how we allocate "value" to labor and define "work."

[1]http://www.thefinancialist.com/automation-a-trend-thats-stic...

[2] http://www.technologyreview.com/featuredstory/515926/how-tec...

That's like arguing that workers who plough a field with their bare hands are less productive than workers who use a plough because someone else owns the plough.

How much 'capital productivity' would be left if all workers downed tools and stopped producing?

As for [2] - if you bother to read Katz, you'll see that even he admits there's a serious issue with wage polarization, and that in fact his own argument is incorrect - there has been no bonanza of well-paid jobs created by technology, and currently there is no prospect for same.

What has actually happened - and he says as much [1] - is that wealth capture through productivity gains has polarized the market into a few well-paying jobs and a much larger number of poverty-line menial jobs, which a majority of formerly middle class workers have been forced into.

Where are these jobs workers have 'failed to adapt' to? What are they, exactly? If they exist, where can they get the capital to retrain for them?

A lot of unemployed and underpaid people would love to know.

[1] https://www.aeaweb.org/econwhitepapers/white_papers/David_Au...

- You're citing a document from 2010, when IBM Watson hadn't beaten humans on Jeopardy, Google driverless car wasn't demonstrated, Boston Dynamics big-dog, atlas hadn't happened, DARPA robotics competition hadn't happened, quadrocopters weren't doing crazy things, Amazon kiva systems, Baxter, etc, etc. Things are picking up pace. (EDIT: hadn't happened or not widely known).

- Since then, Paul Krugman himself has covered tech-unemployment in at least one of his blog posts, in which he has expressed surprise at how things are changing but has concluded the post in humor and essentially dismissing the idea (not very wise!).

- Politicians lag behind economists in staying up to date, who themselves lag behind technologists. "This time" it would really help if economists try to wake up.

- CGPGrey's youtube video 'Humans need not apply' is a must watch for anyone serious about staying aware of upcoming technology.

- The accusation of lump-of-labor fallacy is itself a strawman fallacy. Tech-unemployment doesn't require labor to be fixed. It can happen even if labor lump is increasing but not as fast as humans are becoming obsolete, i.e., "this time".

- Side-comment: Adaptation almost always happens because of new-entrants (youth learning new trades) so it's a generational thing. Those thrown out of the job market are more likely to retire early or something like that. And "this time", even youth are not going to adapt.

- Side-comment 2: Recessions cause quite a bit of damage to social structures which intentionally or unintentionally doesn't get media coverage, or at least doesn't register in public's minds. Just the recent European recession caused suicide in the millions (I came across this article recently but can't find it ATM).

- I've put "this time" multiple times. I have to admit it's a claim without scientific evidence, but circumstantial evidence combined with conversation with technologists who agree what is possible in principle, really is a strong indication. Brain emulation has nowhere to go but in the direction of "a solved problem". The estimated time frame is 10 to 15 years.

I found 'Humans need not apply' utterly unconvincing. For instance, the example cited regarding computer-generated music that was supposed to prove that computers could write great music. But the computer-generated music that was featured in the video sounded obviously like computer-generated music, and not something that I would ever listen to.

The irony there is that music is going downhill anyways, without needing to be automatized, because now you buy songs instead of albums/discs or just listen to them using youtube or some free streaming service.

Plus there may be a finite number of songs -distinguishable by the human hear- than can be created.

Plus, single examples are low-hanging fruit, you need to attack the main reasoning or just assert plausibility.

It's more great music produced now than ever in history, and it is cheap and accessible for everyone.

We are many orders of magnitude before we are even starting to approach the limitations of what humans can distinguish.

Really, "more great than ever in history"? Tell me one single group that is better than Queen, one single singer better than Syd Barrett, one single song better than Stairway to Heaven.

I believe there is good music out there, but saying "more great than ever in history" is just plain false.

And most people are stuck in the same music they listened to during their formative years and will always consider everything done after that garbage.

Consequently, for you there will never be anything better than those groups, and if I suggested anything made today you would find it incomprehensible that anyone could think that music better than your icons.

Not sure about you but my experience is that my life is full of more amazing music than it ever has been before.

You don't have to buy each point in the video (I agree about the computer music) to agree with the larger trend: human brain labor is being phased out. Some already has (aspects of car manufacturing), other large swaths are coming soon (car driving), and the laws of economics will ensure that nothing will stop it. A job is something wealthy people begrudgingly pay you to do until they can figure out how to replace you with a machine.

When I watched "Humans need not apply" the first time, it was a near-religious experience for me. It framed some of the observations and frustrations I see and feel every day and helped clarify the core trend: capitalism has worked really, really well during the labor-based phase of human civilization. Will it work well when human labor is no longer required? What happens then?

The obvious extension of 'Humans need not apply' is that the financial rewards of the march of technology have been concentrated in the hands of the few. The 100+ workers it used to take to work a farm aren't now chilling on a beach somewhere. The founders and stockholders of John Deere are (or their grandkids are). The workers are struggling to find work driving cabs or what-have-you, only to be crossed-out by Google in a few years. There's nowhere to hide. I applaud the entrepreneurs at John Deere and Google and they deserve every penny they made, but the concentration of liquid wealth (excluding homes and cars) into the hands of just a few is fostering profound opportunity inequality. While some folks are becoming wealthy on a scale that's hard to even wrap your brain around, what chance does an Appalachian or inner-city kid have to make it in today's world? Almost none, especially when computers can do things like drive cars.

As the video states, the trend is everywhere. Take Canva, for instance. For some portion of small companies and startups, guess what? They don't need a full time graphic designer anymore. Web frameworks and cloud services have lead to the rise of the "full stack" engineer when there used to be a dedicated database person, app tier person, frontend person, etc. Everywhere

A major, major inflection point of human history is coming in the next couple of decades, and there is no clear solution yet. European-style socialism doesn't work very well and obviously Russian-style oligarchy sucks. So what will we do?

> European-style socialism doesn't work very well

Citation? Because even though Americans are working themselves to death with no vacation, healthcare, or retirement guarantee, the French are more productive and enjoy (what I would argue) is a much higher quality of life:

http://krugman.blogs.nytimes.com/?s=french+productivity+2011

We should not judge quality of life by how many SUVs are in the driveway, and how many big screen TVs are in the 3K sq ft home.

We should not judge quality of life by how many SUVs are in the driveway, and how many big screen TVs are in the 3K sq ft home.

But what about my own goals in life? Would you look down at me for wanting huge TVs and a large home? Those are things that I actually would like to have someday. When I earn enough money to get that stuff, would you advocate taking it away from me?

No, he won't. He will advocate making you pay taxes on the relevant income, but nobody cares if you want to sink post-tax money into pleasures we all consider vain and stupid. We have lives to be getting on with too, you know.

>Would you look down at me for wanting huge TVs and a large home?

Yes. Especially if there are huge sacrifices in your overall quality of life for such meagre rewards.

The idea behind culture and civilization is that not all ways of life and aspirations are equally valid and good.

It might be debatable who is right and who is wrong, but people can and should "look down" on other people for wanting certain stuff.

>You're citing a document from 2010... Things are picking up pace

That's only 4 years ago. Thing's haven't changed much, it's a human perception problem.

Its a misconception that we believe that the time they are living in, right now, is the most important time of all time. Why? Because it's the present and we cannot see the future, and we do not care about the past.

The thing to remember is that we have always thought this! In the 1970's they thought that we were living in the most important time ever, and that things were picking up pace. In 1980s we thought the same, in the 1990's we said the exact same thing. pick any year and humans would say the same, "we are living in special times".

>In the 1970's they thought that we were living in the most important time ever, and that things were picking up pace. In 1980s we thought the same, in the 1990's we said the exact same thing. pick any year and humans would say the same, "we are living in special times".

Actually no. Most of the millenia of human history have been repeatitive and slow moving. Most people in history lived in exactly the same material conditions, and general culture, as their parents and grandparents.

This is something quantifiable, in productivity, in number of inventions, in being able to do extremely important things for the first time in human history (electricity and artificial light, urbanization, nuclear power, flying, space exploration, tele-communications, computing, medicine and biology, etc). All those things are totally unlike the 16th or the 9th century, and they have been picking up pace since the industrial revolution.

And the 70's to today (50 years) is not that much in the course of human history. Not even the whole 20th century is. So it's not like those were district periods that "got it wrong", all those decades belong to the same historical period, and were indeed "special times".

you are comparing the past with the present. Which is exactly my point. If you were in the past, you would think that you were in the most exciting times. You wouldn't know about the future. You would never say "oh, in 30 years time will be the most important time". It's a human failing, we always think that right now is the most important.

We can compare now with the past. But whenever you are in a now, you will always think it's the most important time.

So. We have to remember that everyone always thinks that their now is the most important, and we have to remember this when we think about what the future will be like.

Basically what I am saying is that we need perspective, but that humans are naturally biased to whatever moment they are living in.

I don't think it's really automation that's hurting US workers so much as outsourcing. Globally there are significantly more factory workers doing low to moderately skilled tasks now than at any point in history. We just don't pay them much because there is currently a massive oversupply globally and transportation is cheap.

However, wages in India and China are rising rapidly so that may change fairly quickly.

Yes, but those wages are rising only to the level below which robots can do the jobs cheaper.

The days when working in a factory was enough for a good middle class life are dead and will never come back. Manufacturing, long term, is likely to look like agriculture does today: employing only a relative handful of people.

This leaves the question of what everyone else does. Computer programming? I'd really like the answer to be "not much because we're all so rich we don't have to work", but it seems our societies are nowhere near rich enough to pull that one off.

> This leaves the question of what everyone else does.

I have full confidence in humanity's unquenchable appetite for consumer goods. People with extra money tend to spend that money on something (hand-woven organic angora sweaters). People will all their needs met tend to make up new needs (car seats that are 100.000001% safer).

How is the vast majority going to buy those goods if they don't have jobs to pay for them?

Because you're making it out to be some sort of binarized chicken/egg problem, where I'd say it's more of a grey-mush of chicken and egg mixed into one. They feed off of eachother, whether up or down the spiral of wealth.

This line of reasoning still works even if it's only a few people buying things (think excessively rich people buying ever increasing numbers of luxury goods).

In reality the marginal propensity to consume is greater the less money you make, so there's an argument to be made that increases in wealth disparity might in and of themselves cause a decrease in consumption. It gets complicated though, as when people don't consume with dollars they invest them, and invested dollars are ultimately spent on something too.

Invested money is spent on something but only if someone wants to borrow the money. Now the financial system has got very good at recycling the savings of the rich into more borrowing of the poor (subprime loans), but it is hard to see this continuing at that rate. The rich largely want risk free savings (in aggregate), so the banks turned high risk loans into risk free savings, with the government backstopping them. This has not been working too well. The vast majority of wealth is looking for safe assets (except the Elon Musks) which effectively means people need to have jobs to pay the rentiers.

That, and invested money is not cyclical in the economy. Yes, the invested spend it on something, but only on the pretense that future returns of theirs will go back to the original investor with interest. The only exchange of resources there is the time between having money now and giving more money back later.

Compared to more equitable economics where the baker gives the mechanist bread to fix the milkmans car who gives the baker milk, and each of them is using cash received from the other to provide goods for their wants and needs, is much more tantamount of a functioning economy than loans, especially when investment money seeks the safest of avenues for returns, thus the vast majority of potential entrepreneurs are out of scope of getting the money anyway.

We go explore space because that opens up a wide realm of new possibilities and problems to solve.

I sincerely hope we grow as a society in this regard. And not allow ourselves to be held back by the constant drain of propping up the unproductive, instead of letting them stumble and reach for their dreams.

>The days when working in a factory was enough for a good middle class life are dead and will never come back. Manufacturing, long term, is likely to look like agriculture does today: employing only a relative handful of people.

I'd modify that somewhat. The days when working in a factory as an unskilled labourer paid a middle class income are over, but someone still needs to install, maintain, and program the robots. Those jobs are likely to pay better than the unskilled labour the robots made redundant, but there will be fewer of them.

>This leaves the question of what everyone else does. Computer programming? I'd really like the answer to be "not much because we're all so rich we don't have to work", but it seems our societies are nowhere near rich enough to pull that one off.

In the First World, we are that rich (in terms of per-capita GDP/national income/mean income), but people have an irritating tendency to hold that not-working-for-a-living is a capital-V Vice and that working-for-a-lot-of-money is a capital-V Virtue. It can go to such extreme lengths that people often believe (sometimes against their own class interest) that (high productivity) white-collar jobs are less virtuous than (low productivity) blue-collar ones, and often believe low-salary jobs (like teaching, factory work, manual labor, service-sector positions, and trades) are less virtuous than high-salary jobs (like banking, corporate law, programming, and management).

I dunno.

I like work. I like going into work, I like solving the problems that get presented to me at work, I like the challenge of working with other people to get shit done.

A lot of people hate their jobs. I think it's these people that are hoping one day that the technological wonderland comes where nobody has to work, but the real answer is to manage human labor better. I think that even if we implement basic income, I think after a little while we'll all be back to work.

Reason is people have this deep need to feel useful to each other. They want to feel like they matter. Most people don't at their jobs, but like I said, I think it's because management hasn't evolved to that point.

I don't think we'll ever get a world where no one has to work, at least not until we find the world where nobody wants to not work.

> I think that even if we implement basic income, I think after a little while we'll all be back to work.

The point of a basic income is that people will go back to work on things they're truly passionate about, without fear of becoming homeless.

e.g. a mailroom worker might be really passionate about woodworking, but isn't skilled enough (or confident enough) to make it his profession. With a basic income, he might decide that he can make a go at woodworking without starving to death.

I get that's the point, I just don't see it being politically feasible until after the other thing does.

>I think it's these people that are hoping one day that the technological wonderland comes where nobody has to work

Frankly, I consider it a very strong statement to say that your job is the best possible thing that you would enjoy the most of all to do with your time.

>I don't think we'll ever get a world where no one has to work,

You think nothing can replace human labor economically? Then what do you think human labor is, and where do you think it comes from?

> Frankly, I consider it a very strong statement to say that your job is the best possible thing that you would enjoy the most of all to do with your time.

I spent much of my twenties doing whatever I could to avoid work. Once I realized that all those things boiled down to 'working', and that programming was the most fun thing I'd ever learned in the cause of not working, I figured I'd just get a job doing it.

Could I have better jobs? Sure. But I'd need to have the skills in order to do the job. How do I get these skills? By working on them.

We seem to be performing a lot of mental gymnastics to convince ourselves that certain things are and aren't work. Economics is the study of how to get people the things they want. The things they want don't necessarily exist yet, so we have to work to make them exist. Then we have to work to get them to each other.

Whether it's by making machines that do the physical work or by fixing the machines or by programming the machines, we're still doing work. We will always have to do work. Even if we completely automate the supply chain, we will still have to think of new things to want and then add them to the supply chain.

Even if we realize that all these physical things are a lie and we just want mental / digital things, we'll still have to build / create them. Work. It might feel like play, but there's always going to be things about the building and creating that we like and things we don't like.

To paraphrase Voltaire, if work didn't already exist, we would find it necessary to create it.

Our societies as a whole are rich enough that we all could work much less than we do. That would of course require quite a different distribution of the wealth that we produce.

That point about manufacturing looking like agriculture hit me hard - it's brilliant - thank you.

Worrying too, but insightful.

> The article doesn't really examine productivity, it is examining wages. To wit: industrial automation has grown 6% YOY since 2003[1]. Manufacturing is up worldwide and robotic integration is exploding, not to mention software automation.

Which doesn't seem to have led to huge productivity gains on the order of 6% annually:

> Yet the timing does not seem to support Mr Gordon’s argument. The big leap in American economic growth took place between 1939 and 2000, when average output per person grew at 2.7% a year. Both before and after that period the rate was a lot lower: 1.5% from 1891 to 1939 and 0.9% from 2000 to 2013. And the dramatic dip in productivity growth after 2000 seems to have coincided with an apparent acceleration in technological advances as the web and smartphones spread everywhere and machine intelligence and robotics made rapid progress.

While true, we can't really say that there's a causative relationship or even a particularly solid negative correlation between automation and productivity. In fact, it would be extremely counterintuitive (low prior) to find that there is.

6% industrial automation hasn't been resulting in 6% productivity gains because most work is not industrial.

Indeed. And so it is a little disingenuous to point to narrow gains in one area in a discussion of broad gains, is it not?

A 6% growth in the industrial automation market does not, by itself guarantee any set increase in productivity. All it guarantees is that companies are buying more robots.

Very well said, and the last sentence is key for the future of society in my opinion.

Given your statement regarding wages and productivity - typically the two are used interchangeably in economic conversations / calculations. I wonder if we will start to see those split more to account for the contributions of technology / capital to "productivity". Direct wages would be divided by a combination of societal value + economic value to calculate actual productivity. Perhaps somebody should create the indicator and take home a Nobel or something hehe.

>there is no raison detre to pay workers more

Pedantic comment: it's "raison d'être", and it's used wrongly here. It means "reason for existance".

Reply on news.ycombinator.com