Hacker News Reader

Comment by gboudrias | Hacker News Reader

> The automators will continue to get very, very rich, while the automated will have to find something else to do.

Part of me thinks this will lead to make much more apparent the paradox of capitalism: Those with the most money can... make the most money. Over time this means it all gets funneled into just a few families' pockets, until they own everything (which is silly and most likely would not be put up with after a certain extent).

Another part of me thinks that automation has been part of our society for a long time, and that taking it further only changes the kind of jobs everyone has, maybe even promoting equality by eliminating the more menial tasks (and therefore making every worker's skill and experience important, instead of them being just a replaceable cog).

Either way it seems absolutely inevitable in the long term, and I don't see a point in pretending it's not going to happen.

View on HN

A huge part of the problem here is the low, low interest rate environment we're in. Low interest rates make expensive machines cheaper than humans. High interest rates make humans cheaper than machines. (For a range of values of cheaper)

Being able to afford a $300k house on $50k/year is great until you don't make the $50k/year anymore because you've been automated out of a job.

In the past people with a lot of money could invest their money and see a fairly stable, secure return of 5-15% a year depending on the exact risk profile of their investments; savings accounts -> stocks.

Now if you have a lot of money you can buy businesses and replace people with capital equipment. Better returns than investing in the stock market (or bank accounts!) and it's basically off-limits to people who don't have 7+ figure amounts of cash. Which statistically is basically everyone.

Higher interest rates would let the rich enjoy more work-free income but it also keeps normal people in jobs. And it provides some kind of incentive to save; interest in a bank account is basically scale-free.

I think maybe that's another one of the problems we're seeing, opportunities to grow your wealth aren't scale invariant. A little money has very, very few opportunities to earn interest. A lot of money has a many opportunities to be grown.

Not really - that's kind of a false dichotomy.

Low interest rates are present to stimulate spending. It allows for investments in capital expenditures (e.g. machines) that would otherwise have up front fixed costs that made it unattainable.

This can allow for businesses to invest into new markets or expand where they otherwise would have not spent anything.

This is nonsense. Various funds available to ordinary people, including pension funds, do invest in the hedge funds and private equity funds. If anyone were actively buying businesses and replacing people with capital equipment, it would be these groups.

> opportunities to grow your wealth aren't scale invariant.

This can be solved by pooling your capital with a group of people until you reach the scale at which the opportunity you were looking for becomes available. One way this has been done for over a century is via mutual funds.

> Being able to afford a $300k house on $50k/year is great until you don't make the $50k/year anymore because you've been automated out of a job.

It doesn't even work like that. Rent is basically fixed by the average income level. The price of the house is then annual rent / interest rates.

I agree completely. But "owning" a $300k house makes people feel a lot richer than "owning" a $120k house, even if they're the exact same house! You're right that payments are the same, because housing prices are driven by monthly payments not the dollar amount that you're signing for.

Which is part of the problem, too. When mortgage interest rates fluctuate between 6-10% say, a person doesn't get terribly underwater that fast. But when they fluctuate between 3% and 7% a person can go from equity->underwater in a hurry, because the price fluctuations are so much larger.

It's basically a y=1/x kind of problem and the closer you get to x=0 the closer you get to y=infinity in a very nonlinear way. Which is great if you bought a house a x=more and sold a house at x=less because you made huge gains. But if you bought at x=less and are trying to sell at x=more you can be very, very underwater.

Over time this means it all gets funneled into just a few families' pockets, until they own everything...

Yes, this is a popular theory, but it doesn't appear to be happening. To quote Larry Summers:

> the Forbes 400 list also provides only limited support for Piketty’s ideas that fortunes are patiently accumulated through reinvestment. When Forbes compared its list of the wealthiest Americans in 1982 and 2012, it found that less than one tenth of the 1982 list was still on the list in 2012, despite the fact that a significant majority of members of the 1982 list would have qualified for the 2012 list if they had accumulated wealth at a real rate of even 4 percent a year.

http://www.democracyjournal.org/33/the-inequality-puzzle.php...

For comparison, I'm personally achieving well over 4% returns by trading my own account. Those with the most money can't possibly do the same thing (my strategies won't scale).

Index funds do about 4% real return over time and scale very well.

You are right that wealth is more dynamic than most people think but the reason is rather that the most effective way to get really rich is to leverage everything into one company, which is a strategy that hurts you on the way down as much as it can help you on the way up.

> it found that less than one tenth of the 1982 list was still on the list in 2012

It doesn't seem like a good argument. It's not because they're not at the very very top anymore that their fortune hasn't grown a lot.

>a significant majority of members of the 1982 list would have qualified for the 2012 list if they had accumulated wealth at a real rate of even 4 percent a year.

This part of the quote, which you snipped, is in fact a good argument that they didn't accumulate much. As I said before, that's far less than I earn from trading.

Well, it could simply just mean that others have surpassed them in terms of overall wealth, sure. Not that they lost the wealth.

Presumably very few people in that list are rich in purely cash or near cash terms. They're rich because they own businesses at the apex of their valuation? As technology changes businesses rise and fall which probably accounts for the turnover in the list.

The churn in the top .0001% means basically nothing as to how serious wealth accumulation is. It's a horrible indicator.

Looking at the world cup results of 1982 (Italy, W. Germany, Poland, France) and 2014 (Germany, Argentina, Netherlands, Brazil), you could assume that there is a healthy amount of churn happening, so there's not really an accumulation of medals happening. Every country has a chance! In reality, there are only 8 winners and maybe 12 who have ever had a chance to win (out of 24 who have ever even made it to the top 4). There are ~196 countries in the world. The wealth distribution of FIFA medals is not good if you haven't already been a winner.

http://en.wikipedia.org/wiki/FIFA_World_Cup#Teams_reaching_t...

I'll give you the same reply I gave to yodsanklai, who made the same (incorrect) response:

> a significant majority of members of the 1982 list would have qualified for the 2012 list if they had accumulated wealth at a real rate of even 4 percent a year.

Further, your world cup analogy shows how inequality of results is not strong evidence of gboudraias' claim: "Those with the most money can... make the most money. "

Owning a 2006 or 2010 world cup does not give you a greater ability to play soccer - the World cup inequality between Germany/Brazil and India/USA is purely meritocratic.

Well, that "few families' pockets" end up owning everything is not incompatible with WHICH those families are changing over time.

Besides, 1982 was a very different time compared to 2012. China wasn't as much of a world player in manufacturing etc yet (and Japan was only beginning to get big). Plus, the whole computing explosion happened since then, throwing off many older established players, and then the Internet one. Lots of old fortunes died in those 30 years. In the last 10-15 years though?

> Another part of me thinks that automation has been part of our society for a long time, and that taking it further only changes the kind of jobs everyone has, maybe even promoting equality by eliminating the more menial tasks (and therefore making every worker's skill and experience important, instead of them being just a replaceable cog).

This is the theory, but the realities of this promoting equality are really to be doubted. An increase in the supply of labour has two fairly predictable effects - it both lowers wages overall (more supply, less demand) and provides a pool of cheap workers should things pick up. When there are more unemployed there is also typically an increase in the intensity of work in the jobs that remain - people tend to work much harder for the same or less pay under the looming threat of unemployment. One prevailing myth is that the unemployed workers are simply retrained for new jobs created by automation, however even with government subsidized retraining this doesn't seem to work [1]. In reality automation seems to be a source of widespread immiseration beneficial to only a few.

[1]: http://www.manhattan-institute.org/html/ir_21.htm

> In reality automation seems to be a source of widespread immiseration beneficial to only a few.

Are you serious?

The printing press alone has benefitted humanity as a whole immensly over the last centuries.

And there were piles and piles more of automation necessary for us to be able have this discussion right now - on a global scale and at virtually no cost to either of us.

I don't think printing is a good example, there were not that many scribes around who were suddenly put out of work.

How about the Combine Harvester?

This talk about technology taking jobs from people is utter nonsense. This has happened before - the Industrialization. People are freed from unnecessary chores like harvesting or standing at the cashier, for example. Soon enough new types of jobs were in high demand - mathematicians, engineers, professors. This enabled further development in other professions - more doctors, teachers, historians, etc.

Same thing will happen now. Just look at the education landscape. Data science is taught EVERYWHERE.

Industrilization made humanity able to focus on exact sciences. The advancement in exact sciences will enable us to focus on social studies, philosophy, history, ecology, medicine, law, system design.

Not everyone is happy with an intellectual job.

It's not only about intellectual jobs; it's also about more powerful abstractions. With the same resources spent on the same amount of workers for example, a company will produce more than in the past. It's not necessarily true that in such conditions, the company would choose to keep the productivity constant employing less workers.

One big issue that people tend to miss when comparing current situation to the previous years is the fact that machines are making more machines. See, when combine harvester or loom was invented people were needed to build these machines, so it was fairly easy to move them from doing one manual job to another. Right now however, that one machine can be build by another machine.

When I recently helped build a machine to automate some test processes at a local company it took us about 120 hours in total to do it. The machine will replace three people. All of the components that were used were either computers, software or electronics. This is not something that any of these employees will be ever able to do, as they are simply a very low skilled, uneducated workers, from impoverished neighborhood. The next machine that we may be building for this company will take us even less time, as we have everything already designed, and tested. We just need to put few components together and verify that it works. It will maybe take us 16 hours to do it.

When it takes you less than a week of work to replace three people, there is no way that the jobs for these employees can be created fast enough.

You seem to be discussing net human effects and cherry picking a particularly positive example. I'm discussing the immediate effects of automation on the lives of the workers formerly employed in the automated jobs.

> An increase in the supply of labour has two fairly predictable effects - it both lowers wages overall (more supply, less demand) and provides a pool of cheap workers should things pick up

It's more complicated than that though:

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

"and therefore making every worker's skill and experience important, instead of them being just a replaceable cog"

Problem with this is that most workers don't have the skills and experience, and self-teaching can only go so far in a lot of specialized areas.

Further, you run into the issue of a worker who is, say, 50 and has just been 'automated out'. Learning an entirely new skill set and trying to get into a field where your previous experience may not even be applicable at that age would be brutal, if its even possible in many instances.

To make matters worse, some developed countries (UK and USA to name a couple) have taken to the mindset of cutting education and skills training and prioritized those areas even lower so that it'll only become that much harder for workers to build up skills in the first place...

Copying a comment of mine from a few days ago, seems relevant:

I encourage everyone to watch Humans Need Not Apply, a mini-documentary by the excellent CGPGrey:

https://www.youtube.com/watch?v=7Pq-S557XQU

I suspect most of what is said in there is well known to the HN crowd, but it's all put together in a very pleasant video and is a nice thing to link to people who don't understand just how far along we are in the process of automating a lot of people's jobs.

>Part of me thinks this will lead to make much more apparent the paradox of capitalism: Those with the most money can... make the most money.

This is not workable. Capitalism value is built on having people transfer their assets (money) in exchange for products and services. For this to work it needs paid laborers, not just in order to create the stuff but also to form the large market that buys it.

One could conceive of a future society were automators can make everything and those "automated" don't have anything to do. But that society wont be capitalistic anymore, as it wont contain any market (the majority of the people, those automated wont be part of any market).

Isn't what you've described capitalism's analog to physics' perpetual motion machine? It just seems that there should be frictions and inefficiencies and equilibriums such that it would require severely anti-capitalist intervention to make it happen.

Now, that's not to say we aren't seeing severely anti-capitalist intervention on a daily basis. A major consulting firm in my state got passed a bill to require consulting software developers to be licensed as engineers, then promptly got themselves exempted from it, claiming their "economic interest" compelled them to secure well-trained engineers, even if they weren't licensed.

But if a single actor has all of the money, then wouldn't the money becomes useless? People need things--shelter, food, clothing--not money. Money is only useful if it is a convenient medium for trading labor for things. And if people can't get the things they need, then wouldn't we all just walk away from the old system.

And if a single actor has all the property, then what's the point if there's nobody to charge rent?

Again, barring severely anti-capitalist intervention, like a government ban on travel.

I don't think it is specifically about capitalism.

When you look at how empires fall you can argue that it can often be attributed to increasing wealth inequality to unsustainable level.

And revolutions / civil unrests are often romantically portrayed as a battle for democracy / free rights while they seem to me more about food and jobs.

I know we always talk about rich families accumulating wealth, but what about corporations? Apple has $150 billion in cash sitting idly, accumulating more every time someone buys an iThing. That's money that's disappearing in the short term.

From my understanding of this issue, that money is "owned" by the Apple stockholders, and is incorporated into the company's stock price. If a corporation has a lot of money, its shareholders have that money. Same with anything else the company owns--buildings, equipment, etc.

From my understanding, all the capital (money) and other assets (equipment etc) all belong to Apple, where Apple isn't just "a group of people", but is an entity in itself.

To make it clearer, if apple has $150 Billion cash lying around, none of the stockholders can just withdraw some of it, or even their percentage share of it. What they can do is sell their "share", where the share itself can be considered a separate object that is less related to the company's assets, and more concerned with the market's perception of the company. To illustrate the last point, suppose all of Apple's stockholders suddenly decide to sell their shares. The market would be flooded by sell orders and the price of the shares would begin to plummet, even though there has been absolutely no change in the company's assets. The price drops simply because apparently no one wants to own the share. Similarly when a lot of people want to own it, the price rises. (yes, this is an over simplification)

When you buy a share, you aren't buying a percentage of the company's assets, you are buying:

- a stream of future dividends. - A sell-able "share" which you may hope to sell at profit in the future.

They could also all get together (a shareholder meeting) and vote in a board of directors who would then agree to disburse that capital to shareholders as dividends. It's not quite as direct as going to an ATM and taking out cash, but there are mechanisms for the shareholders to reclaim some of the 'locked-in' value.

It seems that you're missing the fact that the reason outsourcing to India is less profitable is that wages in India have increased enormously.

Reply on news.ycombinator.com