Will robots steal our jobs? The humble loom suggests not | Hacker News Reader

Will robots steal our jobs? The humble loom suggests not

washingtonpost.com

27 pointsGlennCSmith45 comments

"In the 19th century, the maturation and standardization of weaving technology finally brought long-overdue raises to textile workers."

Wow. Way to completely write out of the story the PEOPLE who fought tooth-and-nail to get those raises, WaPo....

http://en.wikipedia.org/wiki/Lowell_Mill_Girls

Of course, acknowledging the real struggle might seriously impact the whole "things will get better by themselves, just keep sleeping" argument of the piece...

Keep in mind that The Washington Post was recently acquired by Jeff Bezos.

This article makes much more sense when viewed in the light of Amazon's current PR problems due to their treatment of poorly paid warehouse workers. They are desperately trying to automate as fast as possible, and this piece tries its best to spin it as a win-win for everyone.

You of course have evidence that Bezos is interfering in the editorial operations of the Post, right?

It's happened enough with other figures & papers that some people take it as good practice to assume, proof or no, and simply disregard the paper on suspect topics.

Certainly nothing you can take anyone to court over, but perhaps prudent.

The idea that Bezos has already turned the WaPo into Amazon's personal PR factory seems highly unlikely. Esp. when you see articles like this:

http://www.washingtonpost.com/blogs/the-switch/wp/2014/01/16...

Bezos wouldn't need to "turn the WaPo into Amazon's personal PR factory" in order to get this published though. You don't have to have a massive conspiracy to get a story published. Everyone in PR knows that.

All he'd have to do is mention the research to a few writers. If you're a writer and the owner of the company mentions a story that sounds interesting to you, you might be just a little etra inclined to cover it.

The first season of House of Cards is a nice illustration of how the combination of powerful people and ambitous writers can lead to a "journalistic" PR campaign without any mass conspiracy. Granted, it's fictional. But it's a plausible mechanism.

Consider Fred, who is paid $1 and produces $10 in value. Every employer would love to have Fred come and work for them. In fact, another employer would likely offer Fred $2 to come and generate $10. Fred, not being an idiot, takes the offer.

Another employer offers Fred $3, and so on, until Fred is making ($10 - opportunitycost).

In other words, Fred being way underpaid according to the value he produces is an unstable situation.

Unless of course there are 10,000 Freds and only 1000 jobs like Fred's.

There's no reason to suspect it's a static system.

People have a history of rushing to start new businesses where there's lots of value being created at little cost.

For a couple examples: the 1849 gold rush, and the current internet startup rush.

The 1849 gold rush was a terrible idea and cost most people more than they got back. Most startups also fail. I feel like that's not a good way to run an economy.

1 9 8 4 rot

Fixed, thanks

Thought experiments cannot be used to prove real points.

On the other hand, the mechanization of agriculture had devastating social consequences when introduced to the UK in the 1700s, leading to mass unemployment, crimewaves and the introduction of the death penalty for stealing food. The industrial revolution eventualy created replacements but things got very ugly for a few generations.

There's no reason to assume that replacement forms of employment will automatically appear when needed.

> as the market for cotton cloth matured in the mid-20th century, further automation no longer generated such increases in demand and new technology slowly eliminated jobs in the textile industry.

This seems to be a counterpoint to his thesis: Technology does indeed create unemployment if all demand has already been satisfied.

His point should thus be stated with a caveat: Technology does not destroy jobs as long as demand increases along with the increased supply. In order to know whether robots will steal our jobs, we have to first determine whether human demands are already being met or not, and it seems plausible to me that they nearly are today (though the inverse also seems plausible).

"Demand" can be a bit of a fuzzy thing. For instance, all those people who were now had "enough" clothes and were saving money from the lower prices were instead buying other things, creating increased demand for some other sector of the economy. Which, presumably, would eventually lead to more jobs elsewhere.

I think a better moral for the story is to expect technology to have significant local effects on employment in particular industries (either up or down), but the evidence for the economy as a whole seems to be much more neutral.

The loom was an example of a technology that actually had a more egalitarian effect on wealth distribution: expensive skilled labor was replaced with greater numbers of cheap unskilled labor.

Given the right prices, this sort of technology can make sense even if it doesn't actually "save" labor at all -- replacing one efficient skilled person with two inefficient unskilled ones for the same output makes business sense if the skilled labor is more than double the price.

Today's market has a lot of disparity in labor prices like this, especially when you look at it globablly. So it's entirely possible for new tech that doesn't actually save labor in terms of output per worker to find a niche.

At the same time, it's also possible for tech to do the inverse, and replace three expensive unskilled workers with one skilled one who commands double the wage.

So in terms of income inequality, at least, it's not entirely clear where technology will take us.

To write an entire article using historical textile production as an example and not mention the impacts of contemporary globalization on the same industry is confusing at best and hopelessly optimistic at worst.

The point of the article isn't what is happening presently in terms of textiles and globalization, but how technological innovation affected the industry in 18th and 19th century Massachusetts.

That said, there are some omissions. The American Colonies / Early United States were labor-constrained. The economy was growing rapidly, and there simply weren't enough warm bodies to meet demand (discussed in Smith, Wealth of Nations, Book 1, Chapter VII). Wages were high, supplies of finished goods were limited, and the benefits of industrialization were high. This contrasted with England in which factories drew masses of labor from the countryside, city populations exploded (Liverpool, Manchester, and Birmingham grew from small towns of 4,000 - 6,000 in 1685 to major cities of 393,000 - 552,000 by 1881, see Arnold Toynbee's Lectures on the Industrial Revolution), and working conditions in general were abysmal.

Which suggests other reasons why picking the Lowell factories as the analysis point is somewhat misguided. Toynbee's Lectures paint a markedly different picture of how industrialization proceeded in the UK.

The challenge now is that virtually all basic needs of survival ... and then some ... are satisfied. Does this mean that there's no additional demand or demand growth possible? No. But ... the situation's markedly different from the late 18th century.

Too: the dawn of the Industrial Revolution was a period in which there were vast untapped natural resources -- virgin land (or, if you prefer, newly cleared land by way of the aboriginal genocide and pandemics) in the Americas, forests, untilled prairies, and of course, untold gigatons of coal, oil, and gas in the ground. William Stanley Jevons noted that increased efficiencies tended to increase consumption of resources by increasing their marginal utility (the so-called Jevons Paradox)

We're looking at a constrained resource environment going out, which may well invert the logic of Jevons Paradox and see fewer resources consumed with increasing efficiency. Which would mean that automation would idle, not employ, hands.

One of the arguments of this article is that as technology stabilizes we will be able to demand higher pay.

Technology will likely not stabilize.

Internet technology has stabilized already: we're still using HTML, HTTP, javascript, etc. The Space Jam website from 1996 still loads in a browser. Most servers and clients still use IPv4. Etc.

Using the article's analogies, they talk specifically about web designers.

Would a web designer from 1996 have a job in today's market? In the same vein, it would be like saying the autoloom would still work today, obviously it would, but no one would use it to produce goods to sell commercially.

http://www2.warnerbros.com/spacejam/movie/jam.htm

Well, it must, some day. But I'm not betting on any specific millennium.

I just love those articles that state that "we have no problem" because "everything just settles out at the end", while forgetting that one of the mechanisms for everything to settle up is a huge share of the population dying. It's survival bias acting at slaughter time.

I don't know about England in the 1700's, but during the industrialization of the US, 1800-1900, the life expectancy and just about every other measure of health got dramatically better throughout that century.

Another way to look at it as go visit Fort Williams in New York, they have a display of 18th century clothing. Look how tiny it was. It's a dramatic illustration of how hard life was in pre-industrial America.

Next visit some Civil War museums and look at the uniforms on display. They're about halfway between colonial sizes and modern sizes.

Overall, technology hasn't really "stabilized" since the loom, either.

It has, however, stabilized locally for periods of time in one area of industry or another.

And it is likely to do similar things again.

Richard Posner's view:

http://www.becker-posner-blog.com/2014/01/secular-stagnation...

TL;DR:

> It is difficult to imagine productive activities that cannot be automated—mining, construction, many medical services, house cleaning: the list goes on and on.

> A decline in the demand for labor, caused by automation, will result in lower wages, without necessarily producing an increase in employment (work effort will not grow if there is no work), let alone an increase in national income.

However, as he also states, "with no truck drivers to pay, the cost of truck transportation will fall, and price will follow". In my view, this will further reduce the main necessities of living (food, hygiene products, energy) which will in turn allow people to live with much less money. Therefore, the poverty levels should fall, as well. It is not hard to imagine the government paying enough social welfare so that everyone can afford basic stuff. Actually, government could also just buy enough robots and produce all the necessities itself.

To get some sense of the numbers:

labor currently takes around 65% of gdp[1]. Owners of the automation will take some of that gain ,so at best we'll see 50% reduction in prices.

According to one calculation, a $10K per person over 21 is a possible basic income proposal[2]. So for family , after adding the 50% deflation, each family will receive $40K in today's dollars.

$40K is close to the median family income in 2012($53K). And we still haven't factored non-automation based innovation.

[1]http://en.wikipedia.org/wiki/Wage_share

[2]http://nextbigfuture.com/2013/11/a-basic-income-of-about-100...

To extrapolate on what you said: "which will in turn allow people to live with much less money", I think this is absolutely true but the endgame is that most things will be nearly free but people will have no money to buy them.

For example computers are effectively free now, they are widely available on Craigslist for under $100 and capable smartphones are bundled into most cellular plans. Flat screen TVs, for all their perceived excess, halve in price each year. However houses and cars still cost $100,000 and $10,000, even used. What's beginning to haunt me is that soon there will be no money in so many industries that people think of as lucrative now: computing, transportation, services, etc, yet the things that constitute the core infrastructure of our society only appreciate in value. You'll be able to get what you want, but not what you need.

My prediction for America in 10 or 20 years, assuming there is no intervention soon, is that corporations will replace the state as the primary wealth manager and that nearly the entirety of the population, probably 400 million people by then, will have no wealth to speak of. People will be struggling to find high paying jobs because raises will be a thing of the past once automation progressively lowers costs each year. There will be a profound sense of being "locked in", running a perpetual rat race where work that used to be worth $100 per hour slips to $50, then $25, and so on until even the most trying work is not profitable, and stabilizes on some rate set by fear, perhaps $10 an hour. With everyone making this across industries, we'll have neo-communism. Your boss won't be able to afford a home, and neither will his or her boss. But go high enough up the ladder and houses will be gifted out as perks.

Then there will be some kind of revolution, it's tempting to think we don't know what form it will take, but if it's to be successful it will be a peaceful one. Basically we'll all look around and wonder why we're all working so hard in the most productive economy that's ever existed, and we'll throw out the bosses by deciding not to participate in the activities that they depend on. Suddenly everyone will have a substantial level of public wealth, you'll instantly own the home you're in, and it will be unclear whether you'll still pay a mortgage. The way it works is that the people and policies that help society are carried over post revolution, and the ones that hinder it are cast aside.

I kind of hope I'm wrong about all of this, but I know in my heart I'm not. We're living in the last heyday of the industrial age, "enjoy" it while it lasts. We can even pretend that we don't know what the post-revolution age will look like, but that's not true either. Everyone's standard of living will go up to what it would have been without oppression. Only this time it will feel unthinkable to coerce people into doing work you would disdain to do yourself, so robots will fill that role instead of other ethnicities, genders or foreigners. It may very well be the last revolution before we're a spacefaring society.

In fairness, and because this got so long, I just want to say that this is only one possible future, and that we may very well adopt robots on a personal level like the PC (hydroponic gardens, off-grid homes and cars built of recycled materials, a sharing economy like Airbnb for everything) before then and the next generation may follow a completely different path. I just think this is the most likely forecast, based on the evidence I see around me now.

My friend, you are an incredible optimist and I hope this is our future. But history shows that power gained is never relinquished without a vicious struggle.

How do you think a non-violent revolt can succeed in a society where mercenaries keep the peace? Furthermore, what if these mercenaries are autonomous systems which require very few skilled individuals to operate? Surely they would simply work for the highest bidders, and viciously enforce whatever "society" these bidders require.

The truth is, going forward, an ever larger percentage of people (specifically their labor) will be deprecated by superior technology (capital). In fact, as technology progresses, most people will become an operating cost in society, as the cost of keeping the majority of people alive will outweigh their total lifetime labor contribution. This is the reality of capital in the technological era. Economics doesn't break down, it just means capital gets more and more of the production pie, and labor gets less.

It will be a brave new world, but in no way will it be better than now. The Dark Ages happened, and they can happen again.

"But history shows that power gained is never relinquished without a vicious struggle."

People say this, but it's not true. Obviously there are countless cases where violent struggle was used, and quite possibly in many of those cases violent struggle was required (maybe even all), but there have clearly been cases of individuals and groups giving up power. Sometimes in pursuit of ideals, sometimes out of fear that violence might otherwise follow.

Ya "how do you think a non-violent revolt can succeed in a society where mercenaries keep the peace?" is a good question and I was never sure either until I watched a show on FSTV called "How to Start a Revolution" about Gene Sharp and his book "From Dictatorship to Democracy":

http://www.cnn.com/2012/06/23/world/gene-sharp-revolutionary...

It's hard for me to say what the gist of it is, but if I had to try, it would be that nonviolence is a powerful weapon against totalitarianism because it's covert. If all the people communicate simultaneously online, and there's no leader, and they reach a consensus that they're going to convince the military and police not to stop them (beginning with persuading servicemen in their own families), then nobody stops them on the day they revolt. So the real struggle happens in the weeks and months before the revolution in the hearts and minds of the institutions that prop up those in power. If the revolutionaries find themselves throwing rocks at police on revolt day, then they've probably already lost. Unfortunately we only hear one side of this on mainstream news, specifically to keep us demoralized and in the dark about how nonviolent revolution works.

Of course in the western world, we're all democracies already, so we may never have a revolution like this. I guess I'm most concerned about voiceless democracy, where as you said capital trumps humanity, and they subvert us all into working toward their goals instead of ours. I really worry about people being perceived as a "cost" on society, because it frames any argument for the social welfare in a bad light in people's minds. And I totally agree with you that it's getting progressively worse with machines lowering the value of labor. I hope some clever people meditate on what may end up being the problem of our time and think of a way out of this mess!

As long as population stays manageable, which sounds feasible in developed economies, and the owners of capital accept being taxed to fund that, or go the Bill Gates way.

> the owners of capital accept being taxed to fund that

Given that the alternative seems to be riots and potentially a revolution, I think they will see that it is in their benefit to keep the population living reasonably well.

Investment in riot control gear and militarization of police in first world countries doesn't suggest that this realization is going to come any time soon.

The incentives are wrong. Each taxpayer would like to reduce his own burden specifically, but everyone shares the cost of riots and revolution (which is mostly harder on the poor, anyway). Not having riots and revolution is a common good.

But the richest have the most to loose. If you're starving and freezing most nights (the poor), you don't really loose a lot by revolting (i.e. destroying the wealth of the rich).

why would anyone want to job in the first place...? If you're in a place where you need to mindlessly execute duties, it's not a good place.

If you have some kind of basic income then what you said makes sense. If not, people still need food and shelter.

This is conveniently forgetting that demand (which eventually ends in consumer demand) isn't necessarily infinite. We have this thing called a planet we live on, that needs to remain intact (or Elon Musk needs to get a move on getting us to other planets :)

Could the inventors of the loom, at the dawn of the Industrial Revolution, have imagined the types of jobs most of have today, in the morning of the Information Revolution? I doubt it.

There will be industries and jobs enabled by today's technologies, that we can't picture today. My bet is on biotechnology. We know only a very small amount about the systems of life.

Yeah there may be new jobs, but why would you have a human doing them, when a robot could?

We can't pretend that a new tool for humans to use is equivalent to a fully autonomous robot which can think for itself beyond any human capacity. Those two things are fundamentally un-alike.

I thought humble loom is a name of some weird startup.

> Will robots steal our jobs?

It's not stealing if I don't want it.

Reply on news.ycombinator.com