So You Automated Your Coworkers Out of a Job (gizmodo.com)

merbs writes: Automation is too often presented as a faceless, monolithic phenomenon -- but it's a human finger that ultimately pulls the trigger. Someone has to initiate the process that automates a task or mechanizes a production line. To write or procure the program that makes a department or a job redundant. And that's not always an executive, or upper-, or even middle management -- in fact, it's very often not. Sometimes it's a junior employee, or a developer, even an intern.

In a series of interviews with coders, technicians, and engineers who've automated their colleagues out of work -- or, in one case, been put in a position where they'd have to do so and decided to quit instead -- I've attempted to produce a snapshot of life on the messy front lines of modern automation. (Some names have been changed to protect the identities of the automators.) We've heard plenty of forecasting about the many jobs slated to be erased, and we've seen the impacts on the communities that have lost livelihoods at the hands of automation, but we haven't had many close up looks at how all this unfolds in the office or the factory floor.

Comments

any job that can be automated

[pezpunk]

should be.

Re:any job that can be automated

[Austerity Empowers]

Any job that can be automated will be . End of discussion.

Re: any job that can be automated

[Opportunist]

Give it time. A few more rounds of automation and the average minimum wage person will be able to do the job.

Re:any job that can be automated

[DarkOx]

^^THIS^^ its the fundamental problem with the 'you are a collaborator' in the labor vs capital argument. If *I* don't do it management will find someone who will (and probably with little difficulty). There is not resisting this from the front lines anyway.

There really is no resisting this from the political lines either. One way or another is going to happen because even if we outlawed certain types of automation or chose to forbid certain industries from automating, some other nation would choose not to do so and our industry would simply get wiped out.

There is no choosing people over productivity. If you don't chose productivity you get no products and the people suffer anyway. We must find solutions that allow people to retain their value by moving into new roles.

Only management can eliminte jobs

[Anonymous Coward]

Blaming job elimination on non-management workers is like blaming 9/11 the jet passengers.

A new way to get an employee to quit?

[xxxJonBoyxxx]

>> been put in a position where they'd have to (automated their colleagues out of work) and decided to quit instead

I wonder if this would work on a overly righteous but inept employee. Hmmmm.

>> To write or procure the program that makes a department or a job redundant.

I don't know about you, but automating work that people manually previously had to perform is one of the main reasons I enjoy what I do.

Evolution

[chrpai]

I've spent a career automating processes. My first such innovation came in my first year and I remember having these feelings when I realized the consequences of my proposal. I spoke to my manager and she said it was our duty to make things more efficient for our customer and that if we didn't someone else would. There is always someone paying the bill whether it's customers, shareholders, private investors or tax payers or maybe in a more abstract way the environment. We always have an obligation to use those resources wisely. In the end these people will retrain and do something else as evidenced by our current unemployment rate.

Re:Evolution

[Opportunist]

This is the beauty of corporations. There's no personal responsibility. It's always someone else to blame.

As a group leader you get ordered to let one of your employees go. It's tough, but you don't have a choice, your department head decided that one of your employees has to be axed and you do it, justifying it thusly that you don't really have a choice.

The department head has to cut his expenses by 10%, so he tries to split the burden, knowing that if he doesn't make his group leaders fire one or two of their personnel, there's a pretty good chance that corporate decides the department is not profitable and they get closed down altogether. The justification is that it's either a few that get fired or the whole department, resulting in many, many more layoffs.

This goes on upwards the corporate ladder until you're at the C-Level. The CEO doesn't enjoy firing people, but he knows that the stock of the company has to perform well or investment bankers would instantly drop them like they're poisoned. He has a responsibility towards the shareholders, and a failing stock could mean that even more people lose their job. That's his justification for firing the aforementioned 10%.

The investment banker isn't to blame either. He was entrusted with money from people, potentially hard working people who want to retire at some point in the future, his justification for buying and selling stocks is simply that his responsibility is to those that trusted him with their money, their hopes, wishes and dreams.

Which leads back to the person who just got fired and his pension fund...

So if you want to find the greedy asshole that just cost you your job, you might not have to look far.

Re:Evolution

[Zmobie]

This is a fantastic point to bring up actually. Automation has been around for 30+ years, you would think that if it were going to destroy the working world and economy it might have made some progress by now right?

Things are about to get interesting

[CastrTroy]

Things are about to get pretty interesting in this respect. For the last century we've been focusing on automating physical labour and we've made a lot of headway, but automating data oriented tasks has been kind of ignored. Sure we've introduced computers into the workplace, but we haven't done a lot of work to make sure we are using them efficiently.

I've seen countless organizations who had 2 systems that didn't communicate with one another so they just employed a bunch of people to copy and paste data between them, and never thought of whether it could be done better and/or cheaper if they just did a little bit of programming to glue the systems together.

Very few companies realize how much time they are wasting when they don't have a good system that is tailored to their needs. There are so many companies working in an Outlook + Excel + Word culture where they don't have any real processes, nobody knows what anybody else is doing, and they aren't really taking full advantage of the computers sitting on their desks.

Re:Things are about to get interesting

[apoc.famine]

There are so many companies working in an Outlook + Excel + Word culture where they don't have any real processes, nobody knows what anybody else is doing, and they aren't really taking full advantage of the computers sitting on their desks.

I've helped a couple of them, with varying levels of success. It really takes management to be not only on board but exemplars of using any new process, or the organization doesn't break that cycle.

One where I was able to mostly break that up (fuck marketing) had some visionary leaders, and they were willing to engage in some workplace disruption for the sake of change. We had gone to a centralized knowledge bank over a sea of personal MS Office docs and personal knowledge, documenting processes and workflows, data structures and filetypes, etc. This was pitched partly as a business continuity process, and partly as an efficiency process - that's what got most of the management very engaged. Some had recently been burned by key staff leaving with an excessive amount of institutional knowledge, while others were feeling the squeeze to become more efficient. As a family owned business, the family part of management was largely on board with it which helped a lot.

Once we had organizational visibility of the business processes, people found that other people were largely doing the same thing, or that they were just a useless cog in a chain. That drove a lot of efficiency changes, as people questioned, "why do I spend X hrs every week doing this, when the person before me could just give it to me in a better format?" Or, "Six people are addressing data quality problems in this process chain. Can we address this further upstream?" A particularly hilarious thing we found was a linear chain which needed to be a fork instead. One person got some data, manipulated it in a necessary but rather extreme way to do their job, then passed it to someone else who subsequently spent a lot of time getting it back largely into its original form. Lightbulb moment when everyone realized, "Oh, we should probably just send the data to both of them, and not have the second guy spend all that time fixing what the first one broke."

Making those changes did eliminate or radically change some jobs, but they made everyone happier. (We found more productive things for most people to do, other than a couple pieces of utterly dead wood.) However, the only way these things could really come up was by documenting, in detail, how the business worked, and letting everyone see those processes and examine their role in them. Most of the people who saw radical job changes drove that change themselves saying, "Why am I spending all my time doing X, when it really looks like we could use Y?"

One job where I wasn't able to break that up had management split or ambivalent on whether or not it would be worth it. After one department pitched a fit because their workflow was changing, and they didn't like it, it got shut down as an organizational change. Some departments still embraced a move to well documented business processes and flourished, but they always had some cruft around the edges where they interacted with the departments who refused. It became a game of insulating the sensible departments from the crazy ones, and building in processes to handle their inconsistencies and incoherent processes and systems.

There's a ton more efficiency to be had in a lot of businesses. I just don't know what percent will be institutionally able to make that change before going under. It always surprises me how few people it takes to utterly derail progress.

Re:bad for dumb people

[Shaitan]

"The ones losing their job due to automation are dumb people, factory workers, people pushing shelves. They truly desrve it."

Big myth, the people losing their jobs are tech workers and most of them automate themselves and their co-workers out of the job.

Automation is the goal, not the enemy

[Anonymous Coward]

We all should strive to have a higher standard of living and work less. Automation is not the enemy, bad allocation of resources is.

Good employees are gold

[grasshoppa]

I know we're all supposed to fear automation, but the fact of the matter is humans respond best to other humans; there will always be work for people to do.

On that note; if I have a good employee, and I write some code that deprives them of anything with which to pay them for, a few things are happening, and will happen:

1) I was grossly underutilizing the good employee to begin with
2) I will find something else for this good employee to do.

Good employees are like gold; you never throw one away, or waste them in such a manner that they'll go looking for someone to better appreciate them. I realize a lot of managers don't grasp this concept, but enough do that good employees will find one if they keep looking.

Mind you; if I automate someone out of a job, and that's all they're capable of doing, they aren't a good employee. At best, mediocre, but probably lower than that. My payroll is more important to me than their want to waste my money.

Re:Good employees are gold

[bill_mcgonigle]

Right. Every instance of automation I've seen for a knowledge worker (we're not talking factory jobs here) has made them delighted and increased productivity. Smart people use better tools to improve their environment and become smarter knowledge workers. If what they are actually doing is knowledge work.

For those who disagree, please head down to your local library and find a counterargument in an index of journals, pull the article from microfiche when they get it from ILL, and post a photostat to my work address.

take away $10/hr jobs and create $20/hr jobs

[Anonymous Coward]

I automated a bunch of tasks that resulted in us hiring less $10/hr positions, but it created $20/hr positions. I was moved to new projects but we needed jr devs on board to support the scripts I left behind. We need more $20/hr jobs, not more $10/hr jobs. The latter isn't going to pull anyone out of poverty.

That's right.

[Anonymous Coward]

Any job that can be automated will be . End of discussion.

And where do all those displaced workers go? Rely on the retraining fairy tale?

For one, everyone cannot be retrained for a marketable profession. And who knows what will be marketable by the time they're done.
And only so many people can work in any profession before it gets saturated.
And it's one thing if someone is in say their 20s, but sending a middle aged person for retraining? Even if they could do very well, employers don't like hiring old (over 40) people.

And in the past when workers lost their job because of automation, they were SOL. If they were LUCKY they got a job lower on the socioeconomic scale. Those weavers who were displaced in the Industrial Revolution became unskilled laborers when they could. Supervisors? Nope - unless they knew someone. Machine operators? Nope - they trained and hired kids to do that.

We really need to think on what to do with those displaced workers because they're all not going to crawl away and drink or do opioids until they die; which is exactly what's happening in much of parts of the country that is being decimated because of our changing economy.

Re:That's right.

[Sarten-X]

Rely on the retraining fairy tale?

Yep.

For one, everyone cannot be retrained for a marketable profession. And who knows what will be marketable by the time they're done. And...

And those are all strawman arguments.

Automation doesn't happen overnight. If I automate a production process, it's just one small part of the industry. There are still jobs available for the displaced workers on other lines production lines or in other departments. As an example, they might not be holding the meter, but they would be reading the test report. In the time it takes to automate a whole industry, the oldest workers usually retire, the mid-career ones head toward management, and the youngest (who started their career when automation was starting) are easily able to move, because they're grown alongside the new automated processes.

It's a common fallacy to think that someone like me (occasionally an automation specialist) will come into a factory in the morning, and put a thousand people out of work by evening. The reality is it takes about 20-30 years to fully automate an industry. Automation just shows such promise that most industries (even those that were reluctant in the 1980s) are about halfway down that road now, so people look around and see automation everywhere, and get worried, even though the unemployment rate has actually dropped, and workforce size has stayed relatively flat.

Now, I'm not saying automation isn't disruptive, and in the short term and small scale it can indeed be devastating to a local economy, but at a national scale it isn't going to lead to any major economic collapse.

Re:That's right.

[Kaenneth]

More art and handcrafted luxury goods is the answer.

Hollywood, Broadway, Netflix, Video Games, Etsy, Books, etc. are not 'productive' yet employ large numbers of people; and people appreciate things being custom crafted instead of cranked out automatically.

The other way is sad too

[Anonymous Coward]

The other way is sad too. Having employees do busy work that could easily be automated is kind of soul-crushing in its own way. It's like watching someone dig a tunnel with a spoon when you're standing next to them with a shovel.

It's sad to see someone lose their job, especially if they haven't built skills to get their next job. But we would never go the other direction and purposely remove automation and modern tools from people so that we could hire additional folks. The fact that we never choose to go backwards means that we should be wary of being too critical of moving forward.

Re:The other way is sad too

[turp182]

When considering IT efforts I like to ask people: "What do you do that makes you feel like a robot?" and/or "What sucks the most?"

That's where efforts should mostly be focused when not purely focused on revenue generation.

There is no shortage of answers to these questions, in any organization.

Involve workers to increase productivity, wages

[raymorris]

Multiple times I've automated much of the work that a person or department has been doing. In each case, it made the workers more valuable.

I talked to workers and watched them work. Together we looked at what tasks they spent a lot of time on, tasks that could be automated to help them achieve department goals more efficiently, while removing human error from that task. We talked about what their workflow would be after the automation and what additional value they could add after they didn't have to spend time in $menialtask.

Being part of the planning, they were able to think about how they could more effectively accomplish department and organization goals when they were freed up from the time-consuming task we were automating. There are ALWAYS more things the company or department wants to do, worthwhile things for people to work on, that they don't currently have time to do (unless perhaps the company is headed for bankruptcy).

The people I "automated away" didn't sit there and say "well now that I don't have to copy/paste from system A to system B, I'm useless". They said "now that I don't have to copy/paste from system A to system B, I can do these other important things to move the organization forward".

Re:Involve workers to increase productivity, wages

[drinkypoo]

"The people I "automated away" didn't sit there and say "well now that I don't have to copy/paste from system A to system B, I'm useless". They said "now that I don't have to copy/paste from system A to system B, I can do these other important things to move the organization forward"."

That's great when there is opportunity to do so. Most managers are shit though, and they not only don't know what to do with that employer, or outright won't let them change things either because they don't understand the change, or don't want any changes they didn't invent to occur. They flail for a while, then they blame the problem on an underling, and fire them. In my last big job I got trained to do a bunch of marketing stuff which we then never did because the marketing manager spent her time sniffing markers, literally. I had mostly automated my job and needed other stuff to do, but it wasn't my responsibility so I wasn't allowed to go outside the box. Then they fired me because it seemed like I had little to do, and finally hired two people to do what I had been doing alone because it wasn't as easy as I made it look.

Most managers I've worked for have been just as incompetent. They don't understand the business, they don't understand what their subordinates actually do, and they have no idea how to improve things. They're just there for a paycheck, like everyone else. On the rare occasion that a competent manager comes along, they are usually hamstrung by management above them.

Re:bad for dumb people

[iamgnat]

"The ones losing their job due to automation are dumb people, factory workers, people pushing shelves. They truly desrve it."

Big myth, the people losing their jobs are tech workers and most of them automate themselves and their co-workers out of the job.

I disagree with who they disparaged, but I also disagree with you too.

I have been doing process automation for 24 years now and no one has ever lost their job because of my work. That was because they were smart enough to not be tied to dead end monotonous jobs and when I presented them with the ability to do a tedious job faster (or not to have to do it manually at all) they used that extra time to perform other tasks.

There is simply always more that needs to be done and automation frees us to work on other things. If someone isn't willing to invest in themselves to learn new skills and find new things to do to stay relevant, it's not the fault of the automation if they lose their job.

What do you do with the people out of work?

[rsilvergun]

I keep hearing that the only solution is to come up with new jobs for them, but there doesn't seem to be much of anything. When I was a kid it was coding and then the H1-Bs and outsourcing took those jobs. Then it was biotech, but those jobs never really materialized in mass (and you need a 4-6 year degree to get them).

I keep saying this on Automation threads, but there was close to 80 years of strife and unemployment following the industrial revolution before WWI & II came alone (the largest government backed guaranteed jobs programs in history, which I could take the credit for that observation but it was Rob Reich who made it). We blew up most of Europe & Asia and killed tens of millions of working age males. The 20th century equivalent of Aztec sacrifice to cull the population.

Are we gonna do that gain? If not what are we doing to do with all these people? Look at the American Indian reservations before the Casinos if you want to see what life is like for people who aren't needed by anyone. Do we want large masses living like that? If not do we have a solution besides "Wait 80 years for a technological revolution to employ everyone"?

Re:What do you do with the people out of work?

[RobinH]

What are you talking about? The unemployment rate is at an all time low. We can't hire unskilled laborers fast enough.

Re:What do you do with the people out of work?

[barc0001]

> There's also other professions like the traditional trades (carpentry, plumbing, electricians, etc) that can't be easily automated away.

Sure, but how *many* carpenters, plumbers, etc are needed? Just one sector that's going to be automated to hell in the next 5-10 years is self driving vehicles. In the US alone that is projected to put 4 million drivers out of work. Truck drivers, taxi drivers, delivery drivers. Is there really enough work in trades that 1 in 50 working age adults becomes another one? To put that in perspective there are ~430,000 people in the US with a job that matches the category plumber/pipe fitter/steamfitter, 600,000 electricians and just under a million carpenters. So about 2 million of those 3 trades. If all the displaced drivers took up those trades, there'd be 3x as many of those trades. Do you think the current demand could support that kind of explosion of the workforce? Not from where I'm sitting...

Re:What do you do with the people out of work?

[atrex]

By definition a UBI program doesn't have any requirements to be eligible for the receipt of the money. It's Universal. Everyone would get a stipend, _everyone_. So getting a job or doing work wouldn't make a person ineligible for it. Now, would someone be able to live comfortably off of it alone? Almost assuredly not. But they wouldn't need to starve to death or live out of a cardboard box (although their living opportunities on UBI alone would probably amount more to some form of community house). But by being universal with no strings attached, they wouldn't be penalized by finding any employment that would supplement and make their lives more comfortable.

Time to lower the full time to 30 hours and up OT

[Joe_Dragon]

Time to lower the full time to 30 hours and up OT pay. So that 80 hour weeks come with an BIG OT payout.

Well, Great employees are gold

[rsilvergun]

folks who can do cryptography, AI math, complex mechanical and electrical engineering, etc.

The key here is these are highly skilled and creative people. They're not just workers, they're creators. When you've got people literally making new things for your business yeah, they're gonna be worth while.

This is not to say you can't make money off good employees. But you're going to run into margins at some point. Like the classic pizza example of economics. That first slice is great, and second might even be better, by the third you're pretty much done and you're probably not gonna make it to the crust on #4. Diminishing returns.

The key here is your good employees are "doers". They aren't making new things for you and opening up new markets, they're just servicing the existing markets.

Most of us are "Doers". Some of them are even very, very good at it. But there just aren't that many "creators". Especially in STEM fields. If there were we'd already have flying electric cars and no disease. You're expectations are too high, which sadly is pretty common among small business employers. You want the world, but you don't want to have to pay for it.

Re:Must be tough for prospective parents

[ichimunki]

If all the jobs are automated, and no one has any jobs, who will buy the stuff the machines produce?

Eliminating an Entire Section of a Factory

[Anonymous Coward]

Some years ago I worked in the sawmill industry using machine vision and automation tooling. There was a sawmill I visited in Arkansas that was eliminating a huge portion of their workforce because of a piece of equipment they had purchased that sorts lumber automatically. Before they purchased the equipment they had 3 shifts of people picking up lumber as it came out of the mill and sorting it/placing it in the appropriate bins - by width and length and grade. Each shift had about 10 people, so over 30 people lost their jobs. This was a big deal because sawmills aren't usually in large towns and this one was particularly rural, so the impact was huge. I asked the plant manager what they do in such cases and he told me that it isn't their responsibility to retrain workers for other jobs if those jobs aren't in their plant (which I understood, but the local government wasn't doing anything for them either).

Do I think that humans should manually sort lumber? Of course not. But there's no denying the impact in cases like this one, which are common in heavy industries, and are now coming into lighter industries and so-called white collar jobs. The people on the factory floor may or may not be rocket scientists (probably not, but one never knows about unrealized potential), but you have to acknowledge a wide variation in abilities and skills across a population distribution. What happens when the line that defines "automatable" vs. "done by human" keeps moving to the right across the distribution curve? What will those people do? This is the reality of things - we either get to the point of a Butlerian Jihad against all "thinking machines" or Star Trek where people just do whatever they want for "the greater good".

Luddite didn't start out as an insult

[rsilvergun]

it was a movement started by textile workers put out of work by automation. Their aversion to industry was because they didn't get anything out of it except unemployment.

What new jobs? Be specific? How will anyone buy the things those new jobs product if they don't have money from jobs now? It doesn't take much to get humans to stop progressing. Remember the Dark Ages? 1200 years of no progress and abject poverty for 99% of the population.

It won't be apocalyptic. The world isn't coming to an end. But we're going to have anywhere from 50-100 years of mass unemployment, poverty, social strife and war. This is exactly what happened the last time we had a major industrial revolution. Eventually new tech caught up and employed people, but in the meantime folks suffered. We have history. We know this happened and we know it's happening now. Why not do something about it?

Put another way: When in your life has the solution to a complex problem (mass technology unemployment) been to ignore it and hope for the best (laissez faire)? Because right now that's all I see us doing.

Not necessarily

[Solandri]

Your reasoning works for jobs whose entirety can be automated - e.g. assembly line worker. The problem occurs when some idiot manager starts trying to apply it to jobs which mostly can be automated. If you decide to eliminate a job because statistically 99% of it can be automated, but fail to account for how to accomplish the 1% of the time which requires human intervention, you're just setting up your business for failure.

IT is a good example 95%-99% of it can be automated by writing a bunch of scripts. But woe unto the manager who decides that once the scripts have been written, there's no more need to pay for an IT department to sit around waiting for a new problem to show up. Or replace the IT guy with his 14 year old nephew who "plays around with computers a lot." You pay IT guys to sit around waiting for a problem when the cost of downtime would exceed their salary.

Save the buggy-whipmakers

[nospam007]

Or embrace the future.

I automated two people out of a job years ago

[mschuyler]

Literally. This is not an apocryphal story. Their manual job was completely taken over by computer. I then hired them to run the computer system at about twice the pay.

At the moment high paying factory jobs

[rsilvergun]

are being replaced with low paid service sector jobs. That can't keep up. People won't have the money to shop at those service industries.

More importantly those low skill jobs (cashier, driver, data entry, back office worker, etc) are what's being automated. So there's going to be nowhere left to go.