This feels quite political. I agree that robots are the future, but I think that some structural changes will be needed. Trying to sell it like no jobs are lost is quite naive at best, and purposefully misleading at worst.
If I remember correctly (don't quote me on that) Amazon used to employ quite a lot of temporary workers and had a high turnover (hire/quit) rate. Then they would no need to purposefully fire anyone to reduce their headcount. Just not temp hiring those people would be enough. Or not hiring as many people as befoe as they grow. Or ... My point: there are many ways of making it look good on paper; but when a robot does a human job, that job is not needed anymore.
What IMO we need to do is to keep on discussing structural unemployment from all points of view.
Well, it might be the case that no Amazon jobs are lost as a result of this automation - but this won't mean that the increased productivity doesn't cause job losses elsewhere (i.e. in their competition both bricks-and-mortar and online).
The problem is that we are losing the simple jobs. You have a lot of people with a basic education, who could do good simple work. The push for higher level of education kind of work is removing them from the market.
In Germany, the net effect is that you thousands of apprentice positions free and at the same time thousands of people looking for such positions. The requirements of today's positions are now to high for many young people.
So we have a gap between educated people and worker people which is growing extremely fast. This results in a society which is polarizing itself more and more. This is not good on the long term.
As an employer, I find it extremely hard to find qualified people, and my entrepreneur friends are having even harder times, with people showing up for their first day of work either drunk, or not at all.
I find it extremely hard to find qualified people. The problem is that a big part of the population is not qualified enough to work and will not be in the future.
The 3.7% unemployment rate in Germany is considering something doing a part-time job for less than €400 per month as employed and the long term unemployed people are not part of the stats, the youths are also not in, etc. They are very good at removing many from the stats.
If you look at the industrial core of Germany, the Rhein/Ruhr region, you see cities after cities cut in two, one side with 35% of population living from social support and the other side with just 5%. When you "start" your life in the wrong side, you are out of the employment stats right from the start.
Note: I am French, living in Gelsenkirchen (such a typical city) and here, we have really hard time finding good engineers even with very good compensation and at the same time we have parts of the city where Doctors without Borders opened a station because the situation is catastrophic.
> at the same time we have parts of the city where Doctors without Borders opened a station because the situation is catastrophic
I can't find any reference to this station online (neither operated nor planned) and have never heard of anything remotely close (I work in the field). On the contrary, currently capacity is being reduced in the region due to permanently vacant hospital beds.
There are some stations for people without health insurance (usually undocumented people that fear detection). The nearest one should be in Duisburg operated by Malteser. Note that those are not because of low capacity but due to documentation requirements at regular hospitals.
> They are very good at removing many from the stats.
This is done by each and every single country, Germany is nothing special here. Very detailed numbers of all those groups usually "excluded" from the main headline statistic are published along with it. There is nothing being hidden. There are also ILO unemployment statistics calculated with the same methodology worldwide, they too look very good (and improving).
Too many people without home who have no health insurance? (Main reason here in Berlin for doctors working in their spare time outside hospitals) Most people - with the exception of failed entrepreneurs - should have health insurance - you must have one if you have a job and there is one if you're on social security payments.
Any links? Doctors without Borders and Gelsenkirchen does not turn up anything in Google.
For the particular case I had in mind it was around 170k€, but in the Rhein/Ruhr Region, you can easily get 100k compensation for good chemical engineers with experience with very good work conditions.
Also in chemical engineering. This is usually for people having around 20 to 50 engineers reporting to them within a big company like Bayer, BASF, Covestro, Linde, etc. or for project managers coordinating large projects (10's to 100's of millions).
If you offer 100k (or 1M or whatever) you just move one engineer from another company to yours. So the other company now has a hard time finding a good engineer. If you increase your salary scale you will get an engineer who is moved away soon by another company which pays more money. In an optimal market, with p% missing engineers you will have p% of your positions open.
You can't plug holes with this strategy. The holes move around while salaries are increasing.
I've worked for a company that payed at the top 10% of salaries and we still couldn't find enough engineers.
I have a number of scientist/mechanical engineer friends who have retrained into software because of how much better the job opportunities are. So it's not just a zero sum game.
Apropos training: Good salaries also attract young people to study computer science, software engineering or a related subject. (I personally think money should not be the sole motivation but it is reasonable to consider the ROI for studying a challenging subject.)
Yes, but with average salaries already higher than most other occupations, the pull from 100k average to 120k average (+20%) shouldn't be that high, but I don't know/have numbers.
Well, with the increasing rents in German cities I think it makes a difference. For me it certainly does. I know at least one company in Berlin that already offers 120k.
But we would need to measure the elasticity of the engineering job market.
E.g. how much more people are becoming engineer if the average salary is going from 150k to 200k, as the salary scale compared to others is already high and should already pull people in.
The people in the final year of high school that are deciding what career to pursue? They will see that there's money to be made in the field and you'll get more people into the CS pipeline, which leads to more engineers 3 years later.
If people aren't choosing the field, no matter how much you are offering new grads, it's an outreach problem and you solve it by making sure people have that information when they are making the decision. Or any other information, really, do you offer relocation assistance? Are your potential hires aware of that? Etc.
It seems to work relatively well for oil companies (else they wouldn't be doing it, right?).
>You can't plug holes with this strategy.
It's too late to be looking for crew when your ship is sinking.
>The holes move around while salaries are increasing.
Yes, of course. But eventually the 'musical chairs' make people move to a higher paying position, and the hole isn't even in your region/country anymore. At least not in your company. The issue with the strategy doesn't seem to be that it doesn't work, just that it takes too long to.
>I've worked for a company that payed at the top 10% of salaries and we still couldn't find enough engineers.
As the other poster mentioned, 10% of what? If the best people from France and Germany are already in Switzerland, you need to compete with Swiss companies, not the local ones that just take what's left.
With salaries in Berlin of $80k and >$100k or $150k in the valley, they currently don't think there is money made in that field? Wow. I do consider $100k a lot of money but YMMV.
"offer relocation assistance?"
Yes everyone I know does this.
"Are your potential hires aware of that?"
Yes.
"It seems to work relatively well for oil companies"
I have no data on oil companies or any experience with oil companies and if they have problems with finding enought good engineers or not. If you have any data, would be interesting, especially on the elasticity of that job market.
I can only speak about software development and hiring people for 20+ years, no clue about other fields.
"you need to compete with Swiss companies"
So I increase the salary to compete with Swiss companies, they increase salaries, holes move around.
Aside from that I assume Swiss companies already have their own holes - at least considering all the executive search calls I get.
You only can plug the holes with more engineers, not higher salaries (above a certain point). And it's questionable you can pull in more people into CS when salaries are already very high compared to most other occupations (but I have no numbers on the elasticity of the engineering job market).
Also: If most ads do not show salary information, how can I get more good engineers interviewing with higher salaries? The problem was not getting good engineers when they showed up for an interview, the problem was getting enough resumes - a problem everyone has outside Google,FB,...
We did some A/B testing on job ads, including salary information did not help. I'm a proponent for
But yes, I would double the salary and let go of half of the people if joining a company as CTO, I'm not against increasing salaries, but they don't close holes from my experience.
Your concept of "holes" is just the state of the current job market for engineers. There are more open positions than engineers available. Hence, companies have to compete more for the available engineers. Salary is not the only but certainly a major factor to make people switch jobs. At the same time the profession gets more attractive for "outsiders". We had this the other way around a couple of years ago in Germany: lots of electrical engineers and to few jobs.
Offering 100k (or more) would also attract more people from abroad. We (I'm German too) have to compete with other regions in the world where the pay is much higher. Higher does not mean better of course (cost of living), but people seem to prefer higher numbers on the paycheck.
We've hired extensively from Poland, but the argument holds: This is creating holes in Polish companies that can't find enough good engineers (as some of my Polish clients tell me).
Of course Polish companies have to increase the salaries too. Seems a lot of developers in Poland are considerable underpaid ... or they really like to work in Berlin :)
Looking for qualified people is the issue being described.
You can find plenty of people with a high school education but training one of them is less efficient then spending more time looking and more money on a qualified person. Which translates to nobody using a large swath of the labor force.
You find it extremely hard to find qualified people - for the price you want to pay, the environment you want to provide, etc. Expect that to continue. Until you start looking for people to work exclusively remotely, you'll have a hard time.
>The problem is that we are losing the simple jobs. You have a lot of people with a basic education, who could do good simple work. The push for higher level of education kind of work is removing them from the market.
The obvious solution to this is to make more people qualified, not to prop up "simple jobs" that aren't economically viable.
As this is an interesting topic to me, how many people have you qualified and what was the ratio of people failing? When I've qualified people, I could make around 25% of people good coders, others I've talked to have similar numbers.
If you have higher numbers, I would be highly interested in the "how".
My only experience is talking to my friends who are either software engineers, or other STEM people, law people or people who study or have studied rather demanding programmes like medicine. Think roughly the top 20% of the population in terms of academic achievement. In my opinion, most of them would be able to learn software development, with proper motivation - i.e. the raw problem solving potential is there, but it's not used often enough.
Funnily enough, those people are exactly the group that doesn't need to pivot their careers.
Even if you can't make a former manufacturing worker into a programmer, you can still try to have more programmers in the next generation.
There's an interesting talk that you might be able to find[1] about abstract reasoning, and how people weren't able to do it just a couple hundred years ago - "what's the difference between a fish and a bird?": "you can eat the fish". No reason why the next generation couldn't be more suited to software engineering jobs than we are now.
1 - I vaguely remember it was something about an old survey of Russian far east villagers (who live(d) as they did hundreds of years ago, for all intents and purposes).
I have been teaching CS classes for 20+ years ... everybody can learn to code, it's just that it takes a different amount of effort for everybody (in a way the same argument as 'offer more pay', in this case, give them more time :)
Interesting, completely different from my teaching and CTO experience. Also interesting it looks very different from German universities, where e.g. in Berlin the majority of students I've interviewed can't do FizzBuzz in an online interview.
How would you assess "good programmer"?
From a very practical point of view, mine would be something like can write a feature on his/her own, gets transactions correct, doesn't create performance bugs that show in live systems, good maintainable, readable, understandable architecture. We definetly could not get everybody on that level even with a lot of mentoring.
We had a CTO meetup recently and mostly everyone shared the same experience.
Coming back, would be interested in "How would you assess good programmer"?
The point is everyone has the potential to learn to code. You don't have to be born a genius or super special to do it. With enough effort and time anyone can pick it up. There's a lot of people who think "Oh, I am not smart enough to learn that" but anyone can provided they put in the time and effort.
There's things that can make it take longer of course. If you're not familiar with computers at all, or have never even owned a computer in your life of course you will take longer to learn. If you are not an analytical person you will have to train your brain. If you're bad at even basic math you'll have to learn that too. You may have to do more learning than other people to learn to code but you can still do it.
Specifically we were talking about "good engineers" and my experience is sadly different from you. Also from my experience >50% of developers are cargo cult developers.
It's more like playing piano, some have the ear for music and some don't (I haven't although I love Mahler ;-) And putting a lot of effort into learning to play piano, many people are not becoming the equivalent of "good engineers".
I'm holding workshops "Coding for managers" to C-level and VCs. I agree with "You don't have to be born a genius or super special to do it", many people are astonished when they write their first code.
Incidently I will give a workshop here inhouse today to non-programmers.
I mean what makes a good engineer that doesn't use cargo cult code? Actual deep understanding of concepts. A lot of people try to fly through the learning process without good study habits so they never really understand it. But proper study habits can be learned too. Obviously someone has to be a motivated learner though, you can't force someone to learn if they don't want to. I -do- see plenty of people without proper motivation do poorly. But anyone who wants to can.
They need the temp workers during peak seasons - they can't really afford to not hire them, nor can they afford to keep on that many longer-term employees.
"Afford" is a very relative term. They don't want to be responsible for that many FTEs but with $48 billion in gross margin and $2.5 billion in net income, they can certainly afford to employ a few thousand seasonal workers year round.
You're vastly underestimating what sort of effects "peak season" has. UPS hires 95,000 temporary employees each year; FedEx hires 50,000. Amazon hires 120,000. Retail scales up similarly, but I'm less familiar with the industry.
"A few thousand" is how much single distribution centers scale up employment.
I think you're vastly underestimating how much money Amazon makes.
So Amazon currently hires 120,000 people at $12/hr for 10 weeks for a total cost of somewhere near $576M. This counts as COGS and is included in their gross margin. If we just multiply that figure by (52 / 10) and subtract out the $576M, we can see that it would decrease their annual Gross Margin by $2.5B. Add on an additional 30% of the combined $3B in salary expense since I doubt the temp workers get benefits, and you'd be looking at a total of ~$3.4 billion.
So in 2016, this would've changed their gross income from $47.7 (35% GM) billion to $44.3 billion (32.5% GM).
Right, but net income is a fiction to begin with. It has almost nothing to do with a company's actual health. Last year was the first year that Amazon ever showed a NI over a few hundred million (for most of their life, it's been negative).
In my experience, this doesn't make them that different from virtually any corporation. I don't think I've ever worked for one where getting money for literally anything wasn't a constant uphill battle (which incidentally is why so many consumer electronics feel cheap as shit; they are).
> In my experience, this doesn't make them that different from virtually any corporation.
For that matter, not any different from a typical person. Expenses that you would be better off not incurring are, by definition, frivolous. In the case of Amazon, frivolous spending at any level changes the price at which they can achieve a fair margin and reduces customer satisfaction.
Except Corporations take it to a uniquely stupid level; things like refusing to pay adequately for internal support, security, auditing, i.e. anything that doesn't have a direct impact on customers.
Every year I need to have usually a 2 hour meeting with my superiors to argue for renewing our antivirus for our Windows infrastructure.
Their constant argument every time is "we have no viruses, why do we spend money on this" and I have to tell them every single time "that you're paying for it is WHY we don't have viruses." They have no idea what viruses are, what malware is, what ransomware is and despite being hired to maintain infrastructure, I must continuously explain my decisions to people who do not maintain infrastructure and wouldn't know how with a gun to their head. This, to me, is stupid.
> They have no idea what viruses are, what malware is, what ransomware is and despite being hired to maintain infrastructure, I must continuously explain my decisions to people who do not maintain infrastructure and wouldn't know how with a gun to their head
Nobody hires someone outside their domain of expertise and then leaves them to do whatever. Your constantly justifying your recommendations is, like it or not, a critical part of your job.
I don't disagree in principle, but it would be nice if I could just do it once, instead of yearly, to the same people, with the same responses. Or at the least have it socially acceptable to script the conversation with a chat bot.
Are you suggesting they should be strong-armed into employing underutilized employees? Perhaps we could do that for the finance industry too. Really, by that logic, any company making money should be strong armed into excess hiring. I suppose they could also push for increased wages instead, which seems more tenable once you get close to full employment.
I wasn't saying they should employ them for moral reasons, I was saying that they absolutely could afford to do so.
Side note; I do hold stock in the company, and because I'm not an asshole, I do think they should rely less on temp labor and they should improve working conditions for their existing warehouse workers. Even if it reduces my capital gains by some fraction of a percent.
> Trying to sell it like no jobs are lost is quite naive at best, and purposefully misleading at worst.
I agree that it comes off as very cynical and feels like a PR piece for amazon and automation.
Also, the problem isn't just that current amzn employees aren't being fired. The problem is also that hiring of NEW employee hiring could be significantly curtailed and wages could stagnate. Not to mention how this would affect local businesses.
AMZN is a major employee during holiday season. Would be interesting how things work out going forward.
Right! This calls for immediate discussion!
Completely new motion, eh, that, ah-- that there be, ah, immediate action--
Ah, once the vote has been taken.
Well, obviously once the vote's been taken. You can't act another resolution till you've voted on it..
In the-- in the light of fresh information from, ahh, the article--
> Trying to sell it like no jobs are lost is quite naive at best, and purposefully misleading at worst.
Well, the article does literally say "No people were laid off when the robots were installed, and Amazon found new roles for the displaced workers, Mr. Clark said.". Either Mr. Clark is lying, or this point is moot.
I'm not following. Suppose they "normally" ran 150 full time employees and 150 temp workers for holiday peak. Suppose they have a 25% attrition rate. (37 of those 150 employees quit voluntarily.)
After the robots, they could have 115 employees working and still not laid anyone off. Yet, in terms of full time employment, they went from 150 to 115 and in terms of temp labor went from 150 to 0.
Did the robots destroy jobs in such a case? I'd think "yes". (I still support the deployment of robots in such cases; I'm only responding to the logical argument here, not the philosophical implications thereof.)
They also remove the need to make future hires, expanding the workforce to match growth. There is a lost "opportunity cost" of unrealized employment caused by massive automation.
The reallocation of labor during the throes of the industrial revolution worked because the new technologies needed constant care and babysitting. Reliable machines that need infrequent service will not need much maintenance staff.
> Suppose they have a 25% attrition rate. (37 of those 125 workers quit voluntarily.)
Why would they retain employees which they would not rehire (minus wages in place of hiring cost)? They have no obligation to retain these workers, so clearly Amazon's base (non-seasonal) demand for labour increases fast enough to justify retaining them. At the current rate you would expect them to hire more labour, not less. The human jobs described here are still so far beyond current robotic technology that it's likely there will be no easy gains in automation here. Meanwhile, as Amazon's efficiency increases, throughput will also increase, which means demand for labour will continue to increase even if there are marginal increases in automation.
> Why would they retain employees which they would not rehire (minus wages in place of hiring cost)?
For good relations with the Government which has a goal of decreasing unemployment, and avoiding bad PR in popular media? If you can wait a couple of months to achieve exactly the same outcome with fewer risks, why not.
> Meanwhile, as Amazon's efficiency increases, throughput will also increase
No, throughput is a function of demand, not efficiency, unless Amazon is currently backlogged.
> For good relations with the Government which has a goal of decreasing unemployment, and avoiding bad PR in popular media? If you can wait a couple of months to achieve exactly the same outcome with fewer risks, why not.
This. Technically, centralisation has given amazon the scale to automate. Socially, that same scale of centralisation exposes them to more critical eyes. Their main vulnerability I guess...
Maybe no people were laid off, and they are hiring like crazy. But they are growing like hyper-crazy. The question is: would they've hired more people if it wasn't for automation? Are they hiring at a proportional rate of their growth? This is something that, unless I've missed it, it's not discussed in the article.
It all sounds like wishful thinking or, indeed, a political PR spin.
It looks like they've used the automation to densify the shelving in the warehouse, and they've streamlined the picker and placer jobs. So the same square footage of warehouse, and same headcount, now has substantially more storage, and higher throughput.
Which probably allows savings on the number of facilities they need to operate, or at least the rate at which they build them.
And i'm sure the story for the bottom 50% of employees is even worse.
The trend is downwards, and the internet accelerated it. Does anyone think that robots and a massive technology boom that may soon happen won't accelerate it ? Why ?
And from that trend, given minimum wage, it seems that at some point in time, we'll have large unemployment.
Why concentrate on share of GDP? Isn't it more relevant how many people are starving? How many people are literate? How long people are living and what the quality of their life is? I don't think share of GDP is a proxy for any of those. Perhaps median wage is though at PPP. The trends there are not great, but not as dire as this (IMHO false) picture implies.
Because money isn't just for sustenance and shelter - it's also a way to buy power, freedom, and autonomy. As the shares become more uneven, people's ability to act to better their own circumstances without a handout from the ultra-rich (or without the permission of the ultra rich) is curtailed.
Look at businesses being dominated by companies that can afford high fixed costs in order to gain huge economies of scale. Good luck opening a mom-and-pop grocery next to a Wal-Mart, or even an independent gas station within two miles of a Sheetz.
If all else remains equal the yes, if you are happy with what you get already you'd be fine.
However we do not live in a society with static rules. Corporations and to a lesser extent individuals are constantly competing and lobbying to get rules changes in their favor, look at the current net neutrality fight. If workers have an ever decreasing share of GDP then they have an ever decreasing amount of power with which to counter that lobbying or lobby for themselves, and it's the sort of thing that can snowball
Poverty has been defined in relative terms for decades, because all the data shows that this has a better correlation with various measures of the effects of poverty, such as poor health and early death, criminality, extreme political opinions, risk of homelessness, life satisfaction etc.
Example: even when corrected for purchasing power, the vast majority of people on earth are poorer than the bottom 10% in the US. Yet they tend to do much better on those measures above than absolute definition of poverty would suggest.
I'd keep in mind that the causality might be the other direction - it might not be relative poverty causing those things, it might be those things causing relative poverty.
The real excat numbers aren't really important but the point is that there's a trend(also showing in median PPP like you say), a long term trend that shows employees get less of the pie, and I think it's reasonable to assume that automation will push this much further and has a chance to create unemployment.
Wouldn't you expect the wage share to decline in aging society? Labor force participation has been on the decline https://data.bls.gov/timeseries/LNS11300000 the number of Americans receiving government disability benefits has been hitting the new records https://www.cnsnews.com/news/article/terence-p-jeffrey/10996... and the only massive new opportunities are in self-employment sector (Uber, Doordash, AirBnB, app store), whom wage share explicitly excludes.
That does not preclude GDP from growing as iPhones are still selling and Google searches are still returning ads.
Why would these demographic changes impact the labor share? A declining labor force participation would (assuming constant productivity) lower GDP, but it wouldn't have an impact on the distribution of that GDP between labor and capital.
Why is that necessarily bad? Does everyone need ongoing employment? Why can't more people build and own a business (or a robot, or other productive capital) instead?
You're right! More people should open independent bookstores, groceries, and gas stations like they used to. Well, maybe not those things. Maybe a car dealership? Only if Elon Musk doesn't succeed in lobbying them out of existence.
Concentration of wealth leads to the ability to outcompete any small business, increasing the risk of operating independently. It's like playing blackjack in a casino. Even if you could eke out a small statistical advantage, the house can stay solvent longer than you can through volatility (and that assumes an otherwise level legal/regulatory/lobbying framework).
Even Silicon Valley's startup culture implicitly acknowledges this - bootstrapped companies are a footnote. Only those companies with the financial resources of insanely wealthy investors stand a chance at mainstream success. That means the (decreasing proportion) of insanely wealthy people have a growing amount of say in what companies are allowed to succeed.
There's a reason why "I started my company with only myself, a garage, and a $3 million dollar inheritance" is an apt description of so many founders' stories.
Perhaps seizing the means of production is the way to go ;)
In all seriousness though, Even starting a small business requires a decently large buffer that most people simply don't have. Hard to build one if you're surviving on mimimum wage + food stamps. Living paycheck to paycheck is a real thing. Especially if you have family
Well, how many super wealthy people's stories are described by "I worked really hard for 45 years and got every promotion and now I have a pension and SS check"?
Only a very tiny portion of the global population is in a position to own productive capital (or, at least, enough to support themselves). Even if the portion were higher, though, you are basically just advocating rent-seeking as the solution to job losses caused by automation, and that strikes me as dubious at best.
Because Amazon, due to economy of scale, low wage employees, automation, and sheer capital can out-compete anyone - and the consumer at large will go for the best deal.
It's very common, we've seen it with e.g. internet providers here; big international ISP takes over one of the major cable companies here, they invest heavily in marketing and go under the price of competitors - and thanks to billions in investment money they can keep that up for years -, the competitors lose money and customers until eventually the big one takes over. Monopoly, and then they start increasing the price again. 5-10 year plans. Amazon can do the same with businesses. Walmart, Target, etc all took local businesses and the middle class out of business when they landed in towns.
That would be great but our society, as it currently stands, is built around having a job and so people losing their jobs in large numbers is a big problem
I did that. My income's really closely tracked the labor share of the economy: follows all the dips and swerves and stuff. You cannot build an entrepreneurial business without a customer base.
I don't think India will be affected at all. Because of the 2 things. Most of the households in middle class only one person works. Otherwise, the 60 to 70 percent of the population is in farming. +Govt kind of already provides minimum wage to live. There are 5 Rs break fast, 108 days of worth govt provided job(Actually no job only money gets distributed about 100 Rs a day), 20 kg rice for 20 Rs, Free bus pass for school kids, laptop, tv, fan etc free for kids,
So in short, robots do the boring and physically demanding tasks, making the humans more efficient. No workers were laid off.
I see the benefits of "relieving" humans of boring jobs, but most of the article seems like a weird PR spin. Nobody got fired because Amazon is expanding its warehouses rapidly enough that those people have work to do despite the robots. Without this modernization Amazon would have needed to hire more workers, meaning this is still a net job loss.
Very few parents say: "I hope my child can get a shipping center job stacking bins, day after day, and hold that job for 40 years."
Shipping-center automation is going to move people out of rote jobs in the same way that farm automation sent a lot of workers off the land. From this story, it sounds as if the people working at Amazon shipping centers have found mildly more palatable jobs that could lead (with more training) to better paying careers in industrial operations, design, etc.
We can (and should!) argue about how fast "better" jobs are taking shape. We can explore ways of making that happen more systematically so that children of pallet stackers have a brighter future. But it's a dim world if we think rote labor like pallet stacking ought to be preserved as long as possible.
>Very few parents say: "I hope my child can get a shipping center job stacking bins, day after day, and hold that job for 40 years."
For a lot of parents, the possibility of their children having a secure full-time job is a lofty aspiration. Millions of Americans are trapped in cycles of unemployment or in low-wage jobs with no security and no benefits. Vast swathes of the US are blighted by long-term unemployment, crumbling infrastructure and a lack of inward investment.
Amazon aren't buying these robots out of some altruistic desire to lighten the burden of their workers. Their treatment of warehouse staff is notoriously dreadful. If Bezos could fire every single worker in his warehouses, I have absolutely no doubt that he would.
Better jobs would be great, but at the moment we just need jobs.
If you work in a software company, why don't you stop using calculators and hire a team of 20 women to do computations on slide rules? That's 20 jobs for salt of the earth, hard workin' families.
Your on a PROGRAMMING forum and your advocating for doing things less efficiently just to keep people employed
It's a big enough economy that we can both find lots of examples to support our different world views. Probably not worth arguing that at length.
Still, the U.S. economy has created about 12 million net new jobs since 2012. (BLS data.) The unemployment rate is at 4.3% now, down from about 8%. For people willing to hunt for something better, the hunt is getting easier.
nobody says they want their kid to have a job stacking bins, but there is a need for low-skill, low-training jobs. "Industrial operations, design, etc" are jobs for skilled employees who want a career. Not everybody wants a career, or wants to go to school. Some people are just looking to stack bins for a few months to fund their travel or surfing or whatever.
I can certainly agree to that sentiment; we are not moving forward as a society if we artificially preserve jobs for the sake of the job itself.
However, the theme of this article seems to be "Automation came to Amazon and no jobs were lost, so don't worry about that pesky automation issue!", which seems blatantly misleading. While new jobs like robot maintenance will be created, I can imagine they will only be a small portion of the jobs that were replaced, and that is going to leave a lot of people unemployed and even potentially unemployable. That is an issue we certainly need to acknowledge and discuss.
They do some of the boring and physically demanding tasks. Robots bring the bins, but humans still stand there putting things into them or taking things out of them, and that is still incredibly boring, albeit necessary because those taks require visual acuity and fine motor skills, which are currently kind of hard problems in robotics and AI.
And even though those human jobs are standardized as much as possible, there are still inefficiencies in the system, many of which are (in my inexpert opinion) caused by the design of the bins and UI deficiencies in the stowing app. If Amazon really wants to improve employee efficiency, I think they could do well to invest more in their internal tools.
That assumes a zero sum game, which is not how the world actually works.
Which is why China could go from $1.x trillion to $11.x trillion in GDP in just ~17 years, while the US - the world's largest economy - simultaneously still managed to add $8+ trillion to its GDP.
Global and local economic growth makes it possible to absorb slack labor. Which is why we have 7+ billion people on earth, while simultaneously the global median income has been rising for decades nearly non-stop, the global median living standard is at an all-time high, and famines are at an all-time low.
If it worked in any regard as a zero sum game, most of those people would be dead or never would have existed in the first place.
The fact that nobody was laid off does not mean that no jobs were lost. If efficiency increased, and nobody was laid off, what does that mean? More product/service was produced. That marginal additional product/service would have required more workers, had they not had this efficiency gain from automation, and I think it's only fair to consider those jobs to be 'lost'.
That being said, I think this is all wonderful, in spite of the misleading framing.
I'd say it's more like declining birth rates in wealthy countries due to the collective decision to have fewer children. From the perspective of social accounting and governance, it doesn't matter why we aren't having fewer children, simply that we are. The same with jobs.
Losing jobs due to lack of job growth concomitant with GDP growth (as is the case here - i.e. more efficiency, same job count) creates more inequality - a larger share of the value produced flows to capital. Depending on your politics, this is either a bad thing, or just a benign fact. But it is definitely a thing.
I don't think that's an apt comparison. Generally the assumption is that when the economy grows, the amount of jobs/employment rate grows. This scenario runs counter to that assumption.
I think it's more like opening a new McDonalds down the street, but it's automated so you don't need a bunch of minimum wage workers to staff it. Or in Amazon's case, maybe more like doubling the size of a warehouse without doubling the staff.
If u ever been to a warehouse there are plenty of jobs nobody here on HN wants, they are mindless and joyless labor and often time temporary. Should we really mourn over the jobs that consume people and treat them like drones? The survived workers are indeed better off doing something that suits their long term interests.
Lots of people want to be drones so that they can feed their families. Ideally, they could feed their families without being drones, but your saying that they should be glad to lose their drone jobs without gaining a different means of support is simply cruel.
Isn't it equally cruel to talk like people want to be drones?
Nobody want to be drones. They are forced into being drones.
Simply keeping those jobs around, merely as an exchange of blood labor for basic living support, is not a mercy to the people you are trying to help. No, it is slaver's talk. It signals a greater failure of the social security network, or the dysfunction of redistribution of wealth to benefit the general public, that people cannot keep basic living standard without such compromise.
> Isn't it equally cruel to talk like people want to be drones?
Well, I didn't append all the qualifiers I could have because I assumed that my comment would be read in a reasonable way.
Right now, those drone jobs are the difference between eating and not eating for a lot of people. I obviously want to see the eating problem solved before the drone job problem is solved. Until then, saying that these people should be happy to lose their drone jobs is cruel. I encourage you to go present your theory to some of those drones. Tell them about how they can cast off the chains of their slavery and be free to starve.
Talk about failures of the social safety net is nice (and I agree, as far as that goes), but it's just talk. Talk is not going to fix the social safety net. These are real lives that depend on this issue.
Difference is, in this case, nobody loses their jobs and no one is starving. What you are saying, is the loss of future drone jobs. Should it be that case, we should never automate agriculture ever, since it had been the major sector of employment for many many centuries.
I think in this article's case, it is good automation, not the aggressive kind, so I want to know why people would lament over the loss of the those non-existed shitty jobs.
First of all, the "no layoffs" claim can easily be achieved through attrition, as covered elsewhere in this thread. That means that someone would have been hired to replace a departing worker but now isn't getting that job.
Second, even aside from attrition, there are absolutely people who would have wanted the jobs that the robots are doing. Some of those people would have wanted those jobs because it would have been their only option. I would rather they had other options (via a social safety net, presumably), but that's not the world we live in right now.
At this point, I suspect you are intentionally reading me in the worst sense you can manage. I think I have adequately explained my position, and if you really want to continue this discussion, I encourage you to go back and re-read my comments. Engage the statements I made or ask questions about the points you don't understand. Otherwise, have a good one.
Some people really do just want to be drones and are perfectly happy going in and doing a simple mundane task over and over.
If anything, this last election should have exposed this. There is a whole cohort of people out there that feel they've been slighted and left behind because of the shift in manufacturing. They feel like they should still be able to go in and be a drone and get paid well enough to raise a family.
Not necessarily. If the costs drop as a result of increased efficiency and that's what leads to more sales, it would be unfair to say that not hiring extra workers is like lost jobs.
That's fair-ish, though I think if the costs drop, it at least part of the time means the demand is not being met by some other competitor, who will likely get forced out of the market, which will cost jobs.
> though I think if the costs drop, it at least part of the time means the demand is not being met by some other competitor
Massively false, in fact you've got it exactly backwards. See: very well recorded 50 years of industrial revolution era history.
The price of consumer goods, and the cost of producing them, dropped aggressively for decades during that time. It was not due to the lack of competition (demand not being met by another company); competition was the reason for it. Specifically, competition competes the margin away.
Input costs and end consumer prices would rise if the demand were being left unfilled by other companies (ie if the market were, for some reason, totally or partially devoid of competition), until competition arrived again. The lack of competition tends to inflate costs across the entire supply chain.
That's just moving the needle though. Demand for lower prices from a competitor would likely come through some efficiency and humans are the most inefficient part of most enterprises - so more than likely there would be less and less job growth over time anyway. What you describe is creative destruction, except Amazon is doing it to themselves.
I totally agree. Like I said in my first post, I think this is great. But I think the way it's being framed as not killing jobs is disingenuous. It is killing jobs, and that's ok. That's the way progress works.
Robots are cool. Even though amazon is hiring, they employ robots for a reason, it means that fewer employees are needed per a unit of work. In the absence of robots, more employees would be hired.
I like efficiency. I think abundance is great. Our economic system ensures that those things will not help the great mass of people in our society.
Whoever said it, the quote about digging with spoons instead of shovels and backhoes comes to mind.
Productivity is the root of economic surplus. It's the lodestone of the economy. If we stop increasing productivity, the pie stops growing. Productivity is the reason the poor today live better than a king or emperor of old.
It's patently obvious that the freshly baked pie is being hoarded by an elite. Average wages and compensation have been almost entirely stagnant since the 1970s. At best, this unequal growth is irrelevant to most people; in all likelihood, it's driving rampant inflation of house prices, rents and an array of essentials like healthcare and education.
If trickle-down economics ever worked, it certainly doesn't work now.
> Our economic system ensures that those things will not help the great mass of people in our society.
If that were actually true, why do the most productive economies nearly universally have the highest median incomes? That includes the US, which has nearly the world's highest median & disposable income levels. If it worked as you're implying, the US would have a terrible median income.
In fact, our economic system - and variations of free market economics broadly - is precisely the reason so many do so well. That's true in all developed nations on the planet, from the US to Canada to Scandinavian countries to Japan to Australia to Germany to Switzerland. To the extent a nation lacks such, is the extent to which they suffer poverty at the median vs developed nations.
Most of those same high median income nations that have taken aggressive advantage of free'ish market economics, are able to provide robust healthcare systems and social safety nets, precisely because of the glorious abundance the system makes possible.
It kinda worked after world war two in the wake of the New Deal (which incidentally was instituted in a depression and a >90% top tax rate). In the 1970s, deregulation took over and worker wages began to stagnate and decline. Tax rates fell. Since 2007 when the banks imploded and were bailed out on the people's dime (the people weren't bailed out), we've been unable to fix the economy with monetary policy.
Demand has been slack for a decade and economists are freaking out. What's going on? Why is civil society disintegrating? Why is the fascist right rising? Why are white people's life expectancy declining?
Neoliberal reforms (deregulation, private capital, sale of nationalized industries, etc) that were instituted in central america, south america, Russia, and others basically destroyed them. In particular, the wave of desperate people arriving at our borders are in significant part due to US support of neoliberal policy and the support of violent regimes in El Salvador, Guatamala, Chile, Argentina, and on and on.
You have to ask yourself, why are we facing problems now? We're the global superpower! Even more so, why are the economic maladies present throughout the western world at a time when we should be at the peak of our civilization? This was something that stumped me for a long time. Shouldn't problems be regional? How is it possible for Donald Trump, Brexit, Geert Wilders, and Marine le Penn to occur at the same time?
>>In the 1970s, deregulation took over and worker wages began to stagnate and decline.
Labour productivity growth has been very tepid since the 1970s, concurrent with an expansion of regulations and social spending.
To give just a couple of examples: the EPA and OSHA were both created in the early 1970s, and have had a documented negative impact on manufacturing productivity. The Empire State Building was built in 18 months. With today's regulations, that would be totally impossible.
This idea that the US saw a rise in neoliberalism in the 1970s is a myth. The opposite is true. Only at the international level is it true, with the market reforms of China, India and numerous other developing countries, and the international situation is looking better than it ever has, with average wages doubling over the last 20 years.
While I should look up a chart, I'm not able to at the moment, but I'd point out that computerization should have added a large constant factor to labor productivity starting in the 90s, you'd expect wages to rise proportionally if that's the case (and for the jobs displaced to be replaced with high wage jobs too).
I can point out banking deregulation, airline deregulation, oil extraction deregulation.... it hasn't been a uniform march in a single direction, but there has been significant deregulation in core industries. One might point out that the construction industry is now safer, as in fewer people die per building. It's not really possible to justify any deaths for construction in reasonable circumstances.
The question is has a doubling of wages kept pace with the increase in wealth in those countries? I suspect 2x is far below the wealth that was produced. It would be helpful to discuss a particular example.
>>but I'd point out that computerization should have added a large constant factor to labor productivity starting in the 90s, you'd expect wages to rise proportionally if that's the case (and for the jobs displaced to be replaced with high wage jobs too).
Computerisation provided a brief boost to labour productivity in the late the 1990s, but that quickly vanished. Labour productivity growth has been at/near century-long lows since the early 2000s, and from the early 1970s to mid 1990s.
I would dispute that we've seen an overall reduction in banking regulations. From what I've heard from those inside the financial industry, the costs of regulatory compliance have risen enormously. Less anecdotally, filing fees for initial public offerings have also increased significantly. All classic signs of rent-seeking.
And then there's KYC/AML/ATF regulations, that have seen enormous growth since the early 1970s, and especially since 911.
On top of all of this a new class of affirmative action type regulations and regulatory pressure that encourages banks to increase lending in particular neighbourhoods where disadvantaged groups are overrepresented has emerged over the last 40 years and steadily become more pervasive.
Beyond the financial sector, any broad measure of the scope of regulations shows that it has increased significantly since the 1970s, at rates far exceeding economic or population growth.
>It's not really possible to justify any deaths for construction in reasonable circumstances.
Numerous people die everyday from numerous causes. A more developed economy generally lowers the risk of death from many causes. So you can definitely justify voluntarily assumed risks in employment, that are associated with more workplace deaths, with a more rapid rate of construction specifically, and economic development generally.
But whether fewer regulations is desirable somewhat beside my point. My main point is that by no reasonable standard has the US become more 'neoliberal' (free market) economically.
>>The question is has a doubling of wages kept pace with the increase in wealth in those countries? I suspect 2x is far below the wealth that was produced.
Our primary concern should be increasing median wages and reducing poverty, not ensuring equally distributed gains. In any case, it's been pretty equally divided. One third of developing countries have seen income equality improve, one-third have seen it remain the same, and one third have seen it worsen.
The difference is primarily in investment. If you stabilise the political situation and provide enough money to invest in infrastructure you're practically guaranteed to turn a developing nation into a developed nation.
The problem is you need a reason to give them money. If robots are more effective than human labor there is no reason to give them any money at all because they lack the education needed for more advanced jobs which also again requires money.
The only alternative to solve this problem is by letting them come to us and immigrate to a developed nation to send money back to their home country. Unfortunately Brexit has shown that this is not an accepted solution.
The article prints Amazon's claims and vision - even their animated VR presentation about future robots - but doesn't tell us the warehouse workers' reality or their point of view. I think it has one short quote from one worker. So we know the former, which is interesting, but we have no idea how it really works out for the employees.
If someone asked your boss about your experience of your job, or someone a couple of levels above your boss (much of the article quoted "Dave Clark, the top executive in charge of operations at Amazon") and didn't ask you, how accurate a picture would they get? Can Amazon's (effective) COO really tell us what it's like to be a factory worker there?
In my experience of being in the position of the interviewer, as a consultant, the boss and even the immediate supervisor usually have little idea about their subordinates' actual work, challenges, frustrations, etc. Unless you ask the person doing the work, you really have no idea what's going on. A common experience: The boss says, 'the new system works great!'; the subordinate says, 'it's awful and I just have to work around it'. But I shouldn't have to talk about consulting fundamentals here; I thought that was a journalism fundamental too.
I don't really have an answer, but you may be interested to know that the retail half of Amazon has a program in place whereby all employees (up to a certain level) are required to visit a one of the warehouses. This is a 3 day affair with a mix of "classroom" time and floor time. While this is obviously far from the unfiltered reality of associates, we (corporate) are at least not totally oblivious.
I don't know about Clark, but I'd imagine he has a rough idea, at least of the general state of affairs. The quote about wanting associates to have less monotonous jobs rings true- I've heard that notion before, and it's good to know that it comes from that high.
Thanks for responding. IME, the types of things you discuss are things "corporate" and executives talk about and sometimes believe, but often are not reality on the ground.
Without evidence, why do you believe it? If you were given that sales pitch about something core to your job, would you buy it at face value? If corporate actually cares, why isn't this issue taken as seriously as other Amazon projects? Wouldn't they be collecting data to evaluate the results? Why is there no regular, quality, anonymous objective survey of the warehouse employees?
For all I know, there is, but there's no evidence of it in the article (the NY Times dropped the ball on it too - and they are supposed to be the skeptical journalists). If not, how do I, the NYT reader and Amazon customer, know if it's not just feel-good stuff that eases hearts and minds not on the factory floor, but in the corporate offices and in the customers' homes.
Does anyone know why Amazon has such visually striking facilities? I'm not used to seeing a warehouse like that. I imagine there is some sort of reason for it (reduces workplace accidents by making certain things stand out), but I suppose it's also possible that Jeff Bezos just thought it looked pretty.
A ton of modern-day manufacturing and warehouse facilities in the U.S. are brightly lit and have safety colors everywhere. This isn't that unusual, except for maybe the yellow bins everywhere.
I work at Amazon in operations, so I'm a little more able to answer than most, but to me it's just a warehouse. The buildings are built as huge boxes. The only thing I can think of is that they're brightly lit. There is an eye toward safety.
Everything is the exact same shade of yellow with a splash of red for danger and occasionally blue. However, after seeing the video (I read this on mobile initially), there is more photoshop going on than I gave it credit for.
However, there is still a high degree of color-coordination, even when you're not looking at shots emphasizing the orderliness of the place with the colors brightened.
There are lots of very clear, systematic horizontal lines. Part of that is how all the equipment is the same, and those self-moving crates are all the exact same height, but they went to the trouble of painting the bottom of poles yellow and the rest white. When you do see red, it's all on the same horizontal line, too. The bottom of the self-moving crates are red, then nothing until you get the tops of the poles, which are all at the exact same height. If I didn't know anything about the company, say it was Walmart's warehouse, I'd expect it to look more like this: https://imgur.com/a/A8ALC
The photos are definitely unusually bright. That said, the 360 video in there seems strangely "dusty" as well. From my recollection, reality is somewhere in between. The photo composition is also having an effect on things- they are crafted to show, as you say, order.
It's also worth considering that there may well be an intentional effort towards aesthetics. These are frequently hard and unglamorous jobs. If white poles and good lighting help make the place feel less oppressive, then I think we'd consider those merits.
Side note: I liked "self-moving crates". Those are Kiva robots carrying pods around. They're fascinating and oddly adorable.
That's a complex question. I get two conflicting perspectives on their jobs: one from corporate, and one from the internet.
Corporate wants to believe that we're trying to do right by them, by making their jobs easier where we can, and by being a decent company to work for. We supposedly pay a bit above average (certainly better than minimum wage), with benefits, and provide some other perks, at least for the full time associates. A lot of effort goes into making their jobs easier and more reliable. It's understood that these are, as I said, not fantastic jobs and most of them won't stay long term, and that's ok. Some move up into management.
The internet tells me that we drive people like mad, that management is brutal and unforgiving, that there's no such thing as full time associates, and even if there were, you'll never become one. I have seen some utterly grueling jobs in person. (The one I have in mind I believe has very short shift times and I think was meant to be automated this year.)
There is some variance where and when we employ FTEs and PTEs, and I think that's a lot of the difference. And the hard truth is that when we bring on more work for Q4, it's seasonal, and most of them will simply not be needed afterwards.
The internal tone could colorfully be described as a controlled panic, that we must scale our operations, faster every year, because Amazon is not slowing down, and we have to keep up. I don't doubt that this manifests itself as a scary downward pressure on the floor, especially during peak. It is understood that manpower alone is failing us and we will shortly be simply unable to hire enough people to meet demand. That we can barely do this now. Automation is crucial.
I suppose my view is: they work incredibly hard and we try to do right by them, and probably don't do enough. We occasionally make the news for the latest horror story, but I find it hard to believe it would ever be intentional. And I have serious doubts about our part time associates.
I can elaborate if you had a particular question in mind.
There was a lot of polish, cleaning and color correction done in those photos. Amazon warehouses look a lot dingier in real life, especially once they've been used.
Automation is good. It enables greater productivity.
The kind of automation here doesn't seem actually much different from the 19th century when hand weavers were replaced by power looms. Except we're talking about warehousing and fulfillment instead of weaving thread. But same case of workers now moving from doing something themselves (weaving or stacking) to overseeing (multiple) machines doing the previous task faster. But this doesn't mean the old jobs required less skill; oftentimes automation actually REDUCES the required skill level.
...and so overall, more stuff is made. And this can reduce worker leverage (especially if combined with devaluing of skill... such as skilled spray booth technicians replaced with people tending automated spraying machines). And THIS is where the negative part comes from.
So we need some way to give leverage back to workers. Whether minimum wage, labor unions, tech cooperatives, etc.
Automation pays for the $15 or $20/hour minimum wage. $20 minimum wage also drives automation. If employment is maintained and innovation in automation continues, this is a virtuous cycle that COULD benefit everyone.
> “For me, it’s the most mentally challenging thing we have here,” Ms. Scott said of her new job. “It’s not repetitive.”
This is a really important statement. If you ever feel like your job is boring, unchallenging and repetitive, that is a sign that it could be done by a robot. Either you need to figure out how to automate it and get a promotion, or you're eventually going to be replaced and let go.
HS dropout. I used to build windows and doors when i was nineteen at an industrial park for wage. I wrote poetry and prose, d&d adventures and designed games + drew maps. Later I fixed cars on the side but wasn't very good at it. Studied philosophy in comm college and on my own.
Eventually I started coding and building/modding PCs and became a TKD instructor and international FC fighter. I've since become an experienced SA and systems programmer.
We tried to create this system in Turkey with AltaTech technology and we can't find a solution about virtual routing.Workers can't find a job in the future because even the backward countries try this system.
The captions "Human pickers followed instructions on computer screens" and "A robotic arm in Florence" are swapped, depicting a robot and a human respectively.
Whether you read article or not it is quite clear the Robots will eradicate a lot of working force though not very soon for delivering products but even that after some time considering the development of drones.
At a quicker rate than they already have been for decades now? New tech has always displaced workers but we've survived it before. Unless you foresee major unprecedented advancements in AI, I don't think it's going to be catastrophic.
If I remember correctly (don't quote me on that) Amazon used to employ quite a lot of temporary workers and had a high turnover (hire/quit) rate. Then they would no need to purposefully fire anyone to reduce their headcount. Just not temp hiring those people would be enough. Or not hiring as many people as befoe as they grow. Or ... My point: there are many ways of making it look good on paper; but when a robot does a human job, that job is not needed anymore.
What IMO we need to do is to keep on discussing structural unemployment from all points of view.