Machines Are Going To Perform More Tasks Than Humans By 2025 (cnbc.com)

In less than a decade, most workplace tasks will be done by machines rather than humans, according to the World Economic Forum's latest AI job forecast. From a report: Machines will overtake humans in terms of performing more tasks at the workplace by 2025 -- but there could still be 58 million net new jobs created in the next five years, the World Economic Forum (WEF) said in a report on Monday. Developments in automation technologies and artificial intelligence could see 75 million jobs displaced, according to the WEF report "The Future of Jobs 2018." However, another 133 million new roles may emerge as companies shake up their division of labor between humans and machines, translating to 58 million net new jobs being created by 2022, it said. At the same time, there would be "significant shifts" in the quality, location and format of new roles, according to the WEF report, which suggested that full-time, permanent employment may potentially fall. Some companies could choose to use temporary workers, freelancers and specialist contractors, while others may automate many of the tasks. New skill sets for employees will be needed as labor between machines and humans continue to evolve, the report pointed out.

Comments

Machines overtook humans years ago.

[Oswald McWeany]

Name one job that doesn't use a machine? Sure, most require a human to operate machines. A computer is a machine, a can opener is a machine, a typewriter is a machine... almost every job already requires a machine.

Now, automated machines is a different thing- they might not have replaced humans yet but even coopers and blacksmiths in ye olde medieval Europe used machines, Machines have been around since man put a stone in a sling.

Re:Machines overtook humans years ago.

[Zorro]

A Knife is a wedge combined with a class 2 lever.

Your point is....

Re:Machines overtook humans years ago.

[The Original CDR]

Name one job that doesn't use a machine?

Signboard person at street corner? Except the Subway store in my neighbor has an automated dummy who looks like a Subway employee with a wig moving the signboard back and forth.

Re:Machines overtook humans years ago.

[Lucas123]

The key word (the dog whistle term) is "overtake". Machines aren't "overtaking" anything. They're helping to automate processes, just as they did in the automotive industry and other manufacturing markets. Some jobs go away, others will be created -- perhaps even more interesting careers will come out of the more mundane processes going away

It's not a zero sum game.

Re:Machines overtook humans years ago.

[Anonymous Coward]

His point is clear, appearing directly in the title of his post: "Machines overtook humans years ago." He is contradicting the OP that states "Machines Are Going To Perform More Tasks Than Humans By 2025".

Captcha: candle

Re:Machines overtook humans years ago.

[ShanghaiBill]

It's not a zero sum game.

Indeed. For many jobs, Jevon's Paradox leads to demand going up as productivity improves.

If you ran a factory, and a machine became available that could make each worker twice as productive, and twice as profitable to employ, would you fire half of them, or hire more?

If automation really displaced workers, then Europe, America, and Japan would be impoverished, while Ethiopia, Afghanistan, and Niger would be prosperous.
 

Re:Machines overtook humans years ago.

[thecatt]

You're missing the difference between automation and tools. What you describe is a tool that make a worker more productive. Automation is more like a machine becomes available that will be twice as productive as a worker and cost half as much, with no worker needed to operate it. How many workers would you hire to work in your factory when you no longer need any? We're not there yet, but it's coming fast.

Re:Machines overtook humans years ago.

[careysub]

Indeed. For many jobs, Jevon's Paradox leads to demand going up as productivity improves.

It is notable that the actual evidence available for supporting Jevon's Paradox (a theoretical claim, not one based on data) in the real world is slim and debatable. The problem is although "rebound effects" do definitely occur, the proposition that they actually increase net utilization instead, of modestly offsetting savings, is very difficult to demonstrate since over time many other factors influence utilization. It should definitely not be treated as any sort of law, that efficiency improvements always/normally/frequently lead to increased utilization. No such findings exist in the literature. It may appear sometimes, to some extent, is a far better statement of the situation.

Re:Machines overtook humans years ago.

[MouseR]

Jerking off.

Re:Machines overtook humans years ago.

[mark-t]

Name one job that doesn't use a machine?

Signboard person at street corner?

Wrong

Re:Machines overtook humans years ago.

[swillden]

A Knife is a wedge combined with a class 2 lever.

A club is a class 3 lever.

Re:Machines overtook humans years ago.

[NicBenjamin]

Consider agriculture.

Back in the 1700s most people's job was farmer. In Europe, villages would be 500-700 acres, and would require dozens of families to farm properly. By the mid-19th century a Homestead (with one family) was 160 acres. These days the cast majority of those have been consolidated. technology has reduced the need for farm work by literal orders of magnitude. We have probably lost more jobs to tractors then any other single technology.

It's possible most office jobs have not been automated, but I doubt it. A lot of functions that used to be done by high school educated administrative types have been automated by the PC. Back then all lawyers had a secretary, and the firm would have employed multiple administrative assistants, law librarians, etc. to keep track of paper. These days the law library is on a computer, and multiple lawyers will share one administrative assistant.

I'm surprised it will be that long

[XXongo]

I'm surprised they predict it will take that long.

The question, of course, is what the humans will do when there half as many jobs. Our economic system is based on the concept that a human who wants to work can find a job.

The obvious solution would be for jobs to work fewer hours. But the economic system seems to have no way to implement that-- what actually happens is that the few who have jobs are overworked and working late every day, while the rest who don't have jobs just stop looking (and hence are removed from the "unemployment" statistics)

Re:I'm surprised it will be that long

[iggymanz]

whatever will we do when the internal combustion eliminates the plowman and the horseshoe. Our agrarian based society will collapse. waaaah!

get real, you have no idea what jobs will exist in 10 years.

automation and IT creates jobs

Re:I'm surprised it will be that long

[fluffernutter]

The internal combustion engine created domestic jobs. Ones that aren't here any more.

Re:I'm surprised it will be that long

[Zorro]

History says get in to trouble.

Re:I'm surprised it will be that long

[iggymanz]

the marketing and selling and maintenance of autos with foreign components in them has created jobs HERE, to say nothing of the use of them to create wealth HERE

Re:I'm surprised it will be that long

[drinkypoo]

whatever will we do when the internal combustion eliminates the plowman and the horseshoe. Our agrarian based society will collapse. waaaah!

Those people got service jobs. But now computers are starting to eliminate the service jobs, and there's nothing for those people to do once they get laid off.

Re:I'm surprised it will be that long

[ThosLives]

Our economic system is based on the concept that a human who wants to work can find a job.

Which economic system is that? I can't think of any economic system based on that premise in all of history. People who want jobs but don't have them have been around for as long as there has been a concept of a "job."

The situation full automation brings into light is that the concept of ownership and the rights to profits of use of capital is breaking down. If human labor is not required to make productive use of capital, then labor cannot be the source of wage. The social upheaval will be because currently only money can be used to gain ownership of capital, and if you have no capital and nobody will give you money for labor, there is no longer a mechanism to gain capital.

This means either forcing a (larger) portion of the productive use of capital distributed to more non-owners (let alone employees!), or reducing the concept of private ownership of capital. Either of those would be a tenuous transition, if for no other reason that people are not used to anything else.

Re:I'm surprised it will be that long

[MightyMartian]

And to some extent the agrarian system did collapse, and beginning with the Industrial Revolution in Britain, you started to see many agrarian workers heading to the cities to work in the factories and related forms of employment. But we're rapidly running out of places where displaced workers can go to find new work. When even service industry jobs like cashiers at department and grocery stores are becoming extraneous, where is it you imagine those people are going to go? For the skilled workers in many fields employment is guaranteed, at least for now, but as we've seen with low-skilled blue collar workers in industrial settings, automation means they end up going from relatively high-paying jobs to more service-oriented jobs, and now that even McDonalds is becoming increasingly automated, even that far less than ideal replacement work is fading away.

At some point, as automation begins to out perform even the cheap labor of developing countries, it isn't just Western low-skilled and service industry workers who are going to find themselves unemployed and unemployable by robots and AI.

The farm hands of the 18th and 19th centuries had options, even if those options were far from ideal. Where precisely do you suggest a single mom with two kids or some 55 year old divorcee getting back on their feet go when all the tills at Walmart are replaced by scanners and two or three floor managers to keep things rolling along smoothly?

This isn't the Industrial Revolution. We've been watching the process of increasing automation for nearly fifty years, and from where I sit, I can see major industrial centers like Detroit basically depopulating. For the moment the system is buoyed by a need for service industry workers, but in part that low unemployment will simply drive automation even faster as businesses give up on finding employees and invest in automation. And then, at some point when unemployment starts to rise those traditional back stops will no longer be there.

Re:I'm surprised it will be that long

[fluffernutter]

No need for marketing when all the cars are in self driving fleets owned by some large conglomerate. Will maintenance still be a thing, or will the trend that puts parts into large foreign made modules continue?

Re:I'm surprised it will be that long

[iggymanz]

large conglomerates have massive marketing and sales departments..and IT too among many other departments.

Re:I'm surprised it will be that long

[The Original CDR]

The question, of course, is what the humans will do when there half as many jobs.

The opposite question is more likely: what will people do when jobs outnumber available workers? Retirees will outnumber workers in the U.S. in 2030. Healthcare and related industries will attract young talent. Good luck in trying to find someone to remodel your kitchen or mow your lawn.

Re:I'm surprised it will be that long

[sheramil]

The question, of course, is what the humans will do when there half as many jobs.

Traditionally, they become homeless, then they starve in an alleyway. Hopefully there will be a machine to dispose of the body.

Re:I'm surprised it will be that long

[bluefoxlucid]

How fast can you breed humans?

Ever wonder why the labor force doesn't exceed job availability? Unemployment keeps going down to about 5% in the US, 2% in Japan. We keep importing labor--300,000 foreign worker per year. Why aren't we just importing millions and millions of workers, having loads of Irish twins, and otherwise building our population?

For that matter, why does population explode suddenly around the Green Revolutions (when food costs dropped by half), computer revolutions, and other technological growth spurts where productivity goes up and the capacity to supply for a larger population increases?

Jobs. Population grows in abundance; and the means to access abundance is jobs. The availability of jobs mediates the moderators of population growth.

The obvious solution would be for jobs to work fewer hours.

This actually increases unemployment.

Imagine you require 40 human labor hours to produce a certain good. On the same technology, cut working hours to 32 (four day work week, 8 hours). Whether weekly wages adapt to the hours or hourly wages adapt to the week, you're going to have 32 hours of income each week instead of 40, and you're going to need 40 hours of income to buy a good.

You see the obvious: we'll need to hire an additional person to fill the last 8 hours--or, rather, we need to employ 25% more people to make the same goods.

You miss the blatant: you only have 80% capacity to buy goods. The same working hours distribute among more workers alright; yet each worker can afford to purchase less. More to the point: our economy has grown to its labor limits under current conditions, and does not have the labor force to fill those working hours; we would need an expansion of the labor force, and thus an expansion of consumption needs, which we could not meet (population would become higher).

Instead, consumption would reduce. Each person would be poorer, unable to buy as much. With reduced consumption comes a reduction in necessary labor: with each need to make five (5) fewer units of that good per week (due to fewer purchases of said good, due to lower purchasing power, due to shorter working hours), we don't replace 5 workers with 6; rather we eliminate 4 workers, with the fifth making up the 8-hour weekly shortfall.

Labor force thus contracts, unemployment goes up, and we wait for the surplus population to die off.

We can achieve shorter working hours by shortening working hours as productivity increases so that we simply create less labor force expansion instead of more poverty. Cutting working hours won't create jobs; it will destroy them.

(We can also shrink the labor force by raising minimum wages.)

Re:I'm surprised it will be that long

[bluefoxlucid]

You know there weren't service jobs at one point because that was rich-people thing? Tipping is a sort of European custom that came to America because European barons and lords were used to tossing some spare change to well-behaved servants. It was, at the time, novel for the less-wealthy to have access to maids and cooks.

Re:I'm surprised it will be that long

[ShanghaiBill]

But now computers are starting to eliminate the service jobs

Computers have been automating service jobs for decades. When was the last time you asked a teller to withdraw $100 from your account? Or a telephone operator to place a call to your mom? Do you pay a laundress to scrub your shirts on a washboard?

... and there's nothing for those people to do once they get laid off.

That's what they said about the tellers, switchboard operators, farmers, hunter-gatherers, etc.

Re:I'm surprised it will be that long

[Anonymous Coward]

Blah, blah, blah, Retirees will outnumber workers in the U.S. in 2030. Blah, blah, blah...

You posted this at least one thousand times already creimer. STFU!

Oh, right, I forgot about your posting strategy and how you chose what you post.

Here is how creimer decides what to post:
https://slashdot.org/comments.... [slashdot.org]
creimer wrote:

All you need to do is find a website with a permissive TOS, say, Slashdot, create a Python script to scrape your own comments, sprinkle Amazon affiliate links in various posts, and then re-post past links whenever possible. Won't be long before you start making "coffee money" each month.

Re:I'm surprised it will be that long

[bluefoxlucid]

This isn't the Industrial Revolution. We've been watching the process of increasing automation for nearly fifty years, and from where I sit, I can see major industrial centers like Detroit basically depopulating.

Well, yes. We don't always create jobs in place; productivity gains come from structural change. That means your farm job goes away and the 300-population small town out west becomes a booming tech megatropolis with 10 million people crammed together in huge apartment complexes.

Detroit didn't depopulate because of lost factory jobs, by the way; it depopulated because the factory jobs moved away. While we have Mexico doing some final assembly, we still employ as many Americans... in the cities surrounding Detroit. The locals of those cities have shorter commutes, and many Detroit workers moved or lost their jobs by some mechanism. Detroit, then, became a collapsed husk.

This is the same as Harley Davidson moving its factory across the country to Pennsylvania. Kansas City is going to lose a huge income inflow, and jobs supported by the spending of now-unemployed factory workers will become unsupportable. A small town in PA will become a booming economy; and people who buy Harleys will spend a little less, since the new factory is closer to the steel mill and the cost of shipping is lower. Not only that, but the new factory uses the latest manufacture processes and technology, so the labor per unit vehicle produced is lower. (That's not the entire factory employment drop: they're also moving some production bound for foreign markets to foreign markets to realize the same efficiency gains and evade Trump tariffs.)

For the moment the system is buoyed by a need for service industry workers, but in part that low unemployment will simply drive automation even faster as businesses give up on finding employees and invest in automation. And then, at some point when unemployment starts to rise those traditional back stops will no longer be there.

A pizza shop needs its input goods (wheat, cheese), its utilities (water, electricity), its mechanism to make pizza (cooks), its mechanism to take orders (cashiers), its mechanism to deliver (drivers), local trucking to deliver its input goods (truckers), people to open and close the store (managers), and its overhead for business operations (inventory, finance, etc.).

Until you eliminate 100% of labor in all of those, pizza stores will employ or cause the employment of somebody. Even self-driving trucks require mechanics. Each step of technology reduces the requirements, which lowers costs. This is, again, structural change: you can buy more pizza--or pizza and other things--and there will be fewer people employed by the same quantity demanded for pizza.

In other words: cheaper goods means more consumption; and the root of all costs is labor--wages paid out of revenues. This balances itself.

Locality does not balance itself. As mentioned above: your town collapses; my town gets all the new jobs; I become rich while you beg for food from neighbors who can't even feed themselves.

Strong social insurances bring things like unemployment insurance, universal healthcare, food and housing assistance, disability assistance, and retirement pensions to these areas. Poverty reduces spending power, which makes it impossible to create jobs (you can't sell things to make revenue to pay workers if your customers are all poor and jobless). Social insurances dump money into those areas, creating an economic stimulus to create jobs.

Means-tested welfares generally cut out as people become employed. This happens too early, cutting off the support for jobs before the economy begins demanding higher-income and broader-reaching jobs. That means you don't couple with the economies around your blown-out factory town, and then you end up with no support when welfare goes away.

Traditionally, we've solved this w

Re:I'm surprised it will be that long

[ThosLives]

Even if there is no "labor" there are still services that can be performed by humans,

Umm... "services performed by humans" is the definition of "labor".

That said - yes, even in a "robot economy" there will still be some demand for human labor (call it "services" if you like). It's a question of scale - in a robot economy, there will not be enough demand for those services go maintain the type of economy we currently have. It's the exact situation of "where will the jobs come from?" You might have knowledge in your head you can sell to others - but the tens of millions of "service economy" laborers today almost assuredly won't. Put a slightly different way: just because you want to offer services, doesn't mean anyone will actually pay you for them.

Also - knowledge is not a means of production. Means of production in the economic sense is agriculture or manufacturing. Knowledge lets you know how to use or create means of production, or maybe trade production around in a way that gives someone a profit*, but it isn't a means of production itself - an important distinction.

*Note there are two types of profit in an economy. There are "accounting" profits, like some stock trader makes a lot of money by buying low and selling high. Then there are "real" profits, which is where an investment in labor and capital increases real productivity (goods and services per unit labor or input material - as opposed to accounting productivity, which is money per unit labor or input material).

Re:I'm surprised it will be that long

[thecatt]

Ask the horses how many jobs internal combustion created for them. So far humans have stayed ahead of automation because there has always been some other job we could shift to we were better at than the machines. That will not always be true. Soon your choices may be show pony or the glue factory.

Re:I'm surprised it will be that long

[bluefoxlucid]

or reducing the concept of private ownership of capital. Either of those would be a tenuous transition, if for no other reason that people are not used to anything else.

That and because this is a fundamental violation of human rights. The core natural rights are the rights to life, liberty, and property; we cut into property all the time (taxes) because governments take away some natural rights under a social contract to protect other natural rights.

I have proposed the theory that governments must actively seek efficiency: if a government taxes 10% and delivers 5% worth of service, the government is unjustly depriving the governed of 5% of their property. This is a human rights violation; thus governments must expend resources to seek efficiency improvements, as they are bound to provide for the welfare of the people and cannot cut spending by simply cutting service.

By the same token, governments have a duty to create and enforce regulation; yet the unnecessary deprivation of private property is a violation of the social contract. A government may, for example, decide that the businessman and the shareholder are entitled to a fair and reasonable profit by their enterprise; and thus the government may tax corporate profits at higher marginal rates when those profit margins are excessive, reasoning that the businessman and shareholders are not simply reaping the benefits of success, but harming society by exercising their market power to excess and creating unnecessary and unusually-large burdens on the common consumer.

It seems excessive and unwarranted, then, for a government to step in and deprive a person or group of property. By taxing them, we place the property in their trust; by taxing them on excessive profits, we reclaim the wealth taken in violation of that trust. By taking the property they have amassed by their own efforts, cunning, and luck, we simply deprive them unjustly.

Many of your assertions base around the idea that labor will no longer be a thing in the future; if that occurs, then people will come together to produce their own means of production. Governments can use tax money to do so. These means of production are magical sources of infinite free things, in your scenario: once created, no human ever need monitor, fuel, or maintain them. It is a one-time expense, and becomes the property of all without depriving it unjustly from anyone.

Should that not happen, it will be because the same thing that has always happened continues: human labor is marginalized, costs fall, and the productive capacity of human labor increases. There is turmoil and displacement, converging toward the same numerical employment rates; there is not a bleeding of employment until full unemployment and mass poverty set in. (Mass poverty at present is a result of certain defective structural policies.)

Re:I'm surprised it will be that long

[Wycliffe]

Ask the horses how many jobs internal combustion created for them. So far humans have stayed ahead of automation because there has always been some other job we could shift to we were better at than the machines. That will not always be true. Soon your choices may be show pony or the glue factory.

There are plenty of service jobs inside the home that are nowhere close to being automated. Cooking, cleaning, even folding clothes is something that AI and robotics still severely lags even people with very low IQs. Worst case scenerio is not massive unemployment but rather going back to a caste system where you have the middle/upper class with dozens of low class servants. This was the norm for the founding fathers of the USA and having household servants is still common in many countries like India. The only reason it is not more common in the USA is because the average middle class person only makes 2-3 multiples of minimum wage.

Re: I'm surprised it will be that long

[ShanghaiBill]

The difference this time is the technology curve is much steeper.

No it isn't. Productivity growth is slowing down, and has been nearly stagnant since 2004.

Previous waves of automation have taken the low hanging fruit in agriculture and manufacturing, and modern service jobs are proving much harder to automate, so it is going slower.

This, of course is bad, because a slower pace of automation means a slower pace of job growth ... which is exactly what we have seen.

Re:I'm surprised it will be that long

[Anonymous Coward]

Just a gentle reminder that industrial revolution was followed by 40 years of misery and hunger to many people. So if you are saying that history will repeat itself, you are saying that after 40 bad years, we will all be better than before.

And while it is relatively easy (just kidding, it is extremely hard) to invent new jobs, it is a bit harder to invent a new job that humans can do better than robots.

Most of the new jobs are just fancy names for engineer/programmer, aka a person who makes technical stuff work. E.g. we didn't have app developers and cloud engineers 20 years ago, but we had programmers doing the same shit even back then. Regarding these jobs, the situation looks pretty good, but the problem is that not everyone can do this job. And to be profitable, these guys have to automate jobs of others. So new jobs here won't actually help the overall situation, just make it worse.

The other half are clown jobs (entertainer) like youtube content creator, social media manager or drone pilot, the problem with these jobs is that they all compete from the time of people and the time of people is limited so either they all get less, or there will be very limited amount of these clown jobs.

Only possible job area where there is no max limit is science. But again, how many can actually do that job and even more important, where could they get funding for their research?

Re: I'm surprised it will be that long

[ShanghaiBill]

The other difference with the jump from horses to engines we could see or predict jobs would be required for the newfangled automobile.

This is false. Car related jobs are "obvious" in hindsight, but were not predicted at the time. The reality is that nobody worried about cars "stealing jobs", because they were such a tiny niche market.

The only near universally accepted prediction was that cars were toys for the rich, and the middle class would never be able to afford them.

It's tough to make predictions, especially about the future.

Re: I'm surprised it will be that long

[ShanghaiBill]

Very very few car-related jobs were predicted. Obviously someone had to make and maintain them, but that is a tiny fraction of the created jobs. The big changes were jobs for pizza deliverers, fast-food workers, and construction workers building the car-induced urban sprawl. Those were not predicted.

Saying "the robots will do everything" is just ignorance of comparative advantage.

what about population growth?

[Anonymous Coward]

Ok, 58M new jobs, for how many people becoming adults worldwide in the same timeframe?

Re:what about population growth?

[Anonymous Coward]

http://www.worldometers.info/world-population/#growthrate
83 Million new people born per year
https://data.worldbank.org/indicator/sl.tlf.cact.zs
62% labor participation rate
51.46 M, call it 45 to account for less births in the past, deaths before age 15.
45 million per year over 5 years is 225 million. Seems we're 167 million short.

IT is going to be hit hard

[MikeRT]

The tools that came out of the DevOps revolution are now going enterprise-wide. It's only a matter of time before most SA jobs become dead-end and then are eliminated. Organizations no longer need a bunch of Unix or Windows admins who are basically power users who can follow instructions. We've circled back a generation to where to be a SA who gets paid well and has job security you'll need to be able to script that system like a boss.

Re:IT is going to be hit hard

[fluffernutter]

DevOps creates such laughable complexity there will be room for all.

Re:IT is going to be hit hard

[phantomfive]

We've circled back a generation to where to be a SA who gets paid well and has job security you'll need to be able to script that system like a boss

There's no such thing as job security in tech (or anywhere, tbh). "Job security" means "I am good at finding jobs." That's the only way.

I'll make a bet with anyone

[DalM]

Here is a $100 bet I'll happily make with any "Robots/AI/Illegal Aliens/Whatever are going to take our jobs" people:

Pick any industry sector you like. Any industry at all. I'll bet a $100 to anyone that there will be more people working in that industry in 10 years than are working in that industry today.

Note: This is not about specific jobs. For example, "Travel Agent" is a job. If you wanted to choose "Travel Industry", I'll happily accept the bet.

Re: I'll make a bet with anyone

[DalM]

$300 then.

Deal.

Re:I'll make a bet with anyone

[ledow]

Though that might hold for most things based on the fact that there'll simply be more people as time goes by for quite a while yet, it won't hold generally.

For a start, depending on how broadly you define your industries, you can always cheat.

The people who made cars are... well... gone.
The people who ran music shops... gone.
The people who took orders at McDonald's... gone.
The people behind my bank counter... gone.

Now that doesn't mean that they can't have hired more bank tellers internally to the bank in other roles, it means that there aren't many bank tellers left. All those people who sat behind a bank branch handling cheques? Gone.

Replaced by a handful of computer technicians and people making machines to accept cheques, and people servicing that infrastructure, sure. But the role has gone. And I very, very much doubt that banks are hiring more people to fulfill that role even if they are hiring more people overall.

If you account for "natural inflation", in that there are just that many more customers, branches and people around, then some industries are indeed dying off. Say, staff-per-customer. That's plummeting in some industries. And there's a reason for that.

Whether the *number* of jobs grows isn't the bet here. It's the proportion (i.e. less humans, more machines, proportionally).

Even IT... I can manage a thousand machines from one desktop. I couldn't do that 20 years before. It simply wasn't possible. But I might have the same team-size as I did back then. The problem is, the other guy is nothing but a keyboard jockey, and proportionally I'm servicing twice as many users as I was back then too.

Thus, though the number may not have changed, the proportion and skills has drastically shifted to the machines instead of the people.

Travel agencies died with the advent of online travel price comparison sites (flights, hotels, etc.). Their replacements may have generated more IT jobs, maybe even more sales jobs, but it didn't make more jobs in the actual travel industry, just the opposite.

Whatever way you look at it, that's a hit. And it's predicated on one problem... that as things get more automated, more of those industry jobs go to IT (whether coding, server support, datacentre rookie, or just plain tech support). And more and more of IT is getting automated. You don't even speak to advisers on websites any more, little AI chatbots cut the simple questions out.

Before long there will be a significant hit to other areas. Big supermarkets and shops probably spend more on their online services now than they do on ground staff... it's ground rent that's killing them and pushing them out of the high-street. Do you know how many big-name high-street retailers have gone out of business in the UK in the last 10 years alone? To be replaced with websites.

And last time I changed car insurer, it was all done online (there's an industry that's almost dead offline), and LITERALLY the backend/company/underwriters that actually insured me for my old and new policies from two different brand names were the same place. Same web interface. Same underwriting clauses. Same technical data access. But two different "brands" offering the same insurance policy at completely different prices.

If anything, that's a perfect example that shows you what will happen - an enormous shift to IT-running of these places, everything operated on the basis of algorithms, no human-face at all (even the customer support lines are way understaffed and refer you to the website more than anything) and then an enormous consolidation of those services from all kinds of places into one place that they all outsource to.

It's slow. There are ALWAYS jobs if you want to go looking for them. But it's inevitable, measurable and inexorable.

I'd go for Insurance as the industry. And I'd say that once things like PPI claims etc. are past their expiry dates, those numbers will plummet. Because, as an industry, they just don't nee

Re:I'll make a bet with anyone

[iggymanz]

utter rubbish, the continued trend on planet earth is for tech to create jobs, create cities, create industries and lift people out of poverty.

your whining and wailing are based on an unreality between your ears. reality is job and wealth creation. the jobs *change*, that is the only truth. jobs become obsolete but even more are create.

For 300 year tech has created jobs, at the present is creating jobs, and will continue to create jobs.

Re:I'll make a bet with anyone

[DalM]

I was going to say the same thing. He was wrong on all four accounts. There are MORE people in the car industry today than there were 50 years ago. There are MORE people in the music industry today then there were 50 years ago. There are MORE people working in consumer facing banking today then there were 50 years ago.

Yes, the jobs people do day-to-day have changed significantly. There aren't nearly as many people working at stand-alone mom-and-pop music stores, but the industry has grown significantly.

Re:I'll make a bet with anyone

[bluefoxlucid]

Bad bet. Structure can change--look at farming, which went from 26 million to as low as 2 million over the span of a century. You want to look at the unemployment rate across all industries.

We may invent new industries. We might conceivably alter the manner in which we set wages to slow labor force growth, drive wage compression, and produce a nation of relatively-wealthy workers (the poor are not that poor, even if they're not exactly middle-class).

Re: I'll make a bet with anyone

[DalM]

It's a fine bet. People's problems is that they tend think that the world as it is right now is in some sort of economic stability. It's not. It's constantly changing and always has been. Look at how much the world has changed today vs 10 years ago. Today there is hardly an industry that isn't heavily affected by mobile internet communications, for example. That whole industry basically didn't exist 10 years ago. And yet, I challenge you to find an industry that is employing fewer people today than 10 years ago. Even the much boohooed travel agent industry hasn't really declined, rather it's actually increased. Right, there are more people working in the travel arrangements industry today than there were 10 years ago. The jobs have just changed from agents to IT. The industry had changed dramatically, but there are no fewer jobs.

Re: I'll make a bet with anyone

[bluefoxlucid]

Oh I know; part of that change is industry dominance as a fraction of total economy, however.

1.2 million agricultural workers, 2008. 856,300 agricultural workers, 2016. Outlook: 0% employment change, 2016-2026, projected to have 400 fewer employed in the industry in 2026 than in 2016.

Total number of employed workers in the United States in 2006 was 145 million, and over 150 million in 2016. Even at a rate of zero additional jobs, farm industry employment is shrinking relative to the economy's total employment growth.

There are more jobs in the United States; there are fewer jobs in the agricultural worker industry. More engineers, chemists, and doctors; fewer people riding tractors.

Agricultural productivity has increased faster than agricultural production. That means for each unit good, we employ fewer labor hours; and the number of unit goods demanded (and produced) has increased at a slower rate. We can make 110% as much per worker-hour, and we buy 105% as much, thus number of employed in the entire industry goes down.

Re: I'll make a bet with anyone

[DalM]

I'm not talking about jobs. There may be more or fewer people actually picking berries in the US in 2028. The number of farm hands is dependent on a LOT of factors, and predicting numbers down to "400" in 8 years is absurdly fake precision. (The number of farm hands in America changes more than than every day.)

"There are more jobs in the United States; there are fewer jobs in the agricultural worker industry. More engineers, chemists, and doctors; fewer people riding tractors."

That's exactly my point. There are more different types of jobs today. There are more people that work in the agricultural industry doing chemistry, engineering, soil science, marketing, import shipping, just-in-time distribution, etc. Ultimately, there are actually MORE people working in agriculture today then at any time in American history, it's just fewer of them are driving tractors.

Re: I'll make a bet with anyone

[bluefoxlucid]

I'm not talking about jobs.

This is an English forum, and employment means jobs.

There are more people that work in the agricultural industry doing chemistry, engineering, soil science, marketing, import shipping, just-in-time distribution, etc.

Not really. The engineering, chemistry, marketing, import shipping, and logistics industries are highly-efficient: they devote a fraction of their resources to agriculture, and each person working in those industries is producing outputs for a large span of industries. Only part of a given worker's output is an input to agriculture; and when you combine those fractions, you get strikingly-few workers.

Your strongest connection is the manufacturing behind farm equipment, and even that's not really well-correlated: John Deere supplies the construction industry, the homeowner, the landscaper, and so forth.

Your front-end, on the other hand, is arguably not even part of the industry: shipping, stocking, and selling--truckers and grocery stores--isn't agricultural work, and distribution comes because we've moved from a small farm a cart-ride away to a continental and global industry. You might as well claim polyester clothing manufacture (which uses oil, not cotton or wool) is part of the agricultural industry because people can't go grocery shopping naked.

By the same token, the bakery industry, fabric industry, and meat industry aren't part of the agricultural industry: they take inputs from the agricultural industry (food and fiber). To include marketing and distribution in the agricultural industry is along the same lines, and is collapsing all industries into a single industry of everything made anywhere ever.

Input industries are arguably separate, too. Maybe your travel agency employs system administrators and network engineers; it also buys CISCO gear, and we don't count CISCO router manufacture as part of the travel industry. Counting John Deere as part of the agriculture work industry is even a stretch; agricultural bioengineers might mesh better; and we absolutely count the contractors who fly spray planes over the fields.

So if you use actual economic definitions and standard measures, you're wrong. If you wave a flag of bullshit, you can pretty much justify that pretty much 100% of American workers work in the agricultural industry.

Re:I'll make a bet with anyone

[iggymanz]

false, how can you say that when massive amounts of the population such as in rural china are being lifted out of poverty into middle class with job.

the world population is growing, yes that means the number of humans who can use tech to create wealth is growing.

people will get money to eat, clothe themselves and sleep by jobs enabled by tech.

Re: Specious as always

[iggymanz]

all cars and trucks have foreign components or are foreign made. Yet their sales, marketing, maintenance and use create jobs domestically.

technology creates jobs, some people can't get that truth into their obtuse skulls.

that truth hasn't changed in 300 years, and will not change

Dependin on your definitions....

[JoeDuncan]

...this has already happened.

I mean are they counting robotic car assembly as "workplace tasks"? What about assembly line QA by image recognition? Is that a "workplace task"?

To vaguely defined to be much more than bullshit click bait.

Re: Specious as always

[iggymanz]

no evidence?? Haha, the graph of percentage of tech enabled jobs has done nothing but go to the sky over 300 years. most the jobs at my employer would not exist without tech. for that matter, the business itself could not exist without tech.

tech creates jobs, history proves it

My experience with DevOps

[rsilvergun]

is that it's an excuse to make the developers pull double duty as sys admins. It also lets companies take jobs that used to require little more than a high school diploma, declare them as "Developer" jobs in need of an advanced degree and bring in H1-Bs to take them. Basically, it's longer hours and lower pay for everyone.

Re: Amish Farmers

[Anonymous Coward]

The Amish use machines ,,, what do you think a plough is? It is irrelevant that it is horse powered. You are the ignorant one.

More shit 'news' to ignore

[Rick Schumann]

All so-called 'news' like this does is spread FUD (Fear, Uncertainty, and Doubt), and as such is just clickbait. While there's this so-called 'news story', there's also this so-called 'news story' that is diametrically opposed to it.

Just IGNORE all of it, none of it means ANYTHING.
Business will continue as per usual. Billions of people will not suddenly be replaced by robots, AI isn't going to cause The Apocalypse, and so on, and so on.
All of this shit is just distractions from the things that will actually fuck up your life: shitty politicians doing shitty things to your government, human-caused climate change making the world less habitable, and shitty corporations progressively more and more screwing you over and using you like a toilet.

Re:More shit 'news' to ignore

[Rick Schumann]

That's all just driven by people with a severe case of myopia and short-sightedness coupled with extreme laziness. They don't want to work at all, they want to sit on their fat asses, eat, drink, party, and otherwise be wastes of oxygen, all on free government money. They're no better really than so-called 'welfare queens'.

Too vague a statement to be meaningful

[sjbe]

In less than a decade, most workplace tasks will be done by machines rather than humans

That was true a long time ago depending on how you parse the word "tasks". If a calculation is a task then computers do untold trillions of tasks every second - far more than what people do. We're a species of tool makers. It's hardly a surprise that our tools make us more productive.

Never as Quick as Predicted

[Only Time Will Tell]

I'd anticipate a growing shift to automation and intelligent business systems over the next 10 years, but I don't think the pace is going to be as fast as predicted. Technology always moves faster than society, and businesses usually adopt new tech when it is convenient for them. In the late 90s, the pundits claimed any business not on the internet was doomed, and then the dot-com bubble burst and people realized only some businesses benefit from e-commerce. Shipping 50lb. bags of dog food over the internet was a bad idea (ask Pets.com). The same thing will occur in this new wave of manufacturing and automation. There will be early adopters who try out the tech and see where reality meets the hype. Once other companies see what those early adopters are able to accomplish (and the money they're making), they'll gradually transition as well. I once thought that by the early 2020s truck drivers and other driver professions would be out of work. It now turns out that self-driving vehicles aren't as far along as once thought, and automated trucks won't likely be out in force until much closer to 2030 or later.

Define "industry"

[sjbe]

Pick any industry sector you like. Any industry at all. I'll bet a $100 to anyone that there will be more people working in that industry in 10 years than are working in that industry today.

You'd win that bet in a lot of industries but not all of them because it depends heavily on what definition of the word "industry" you use to define a particular industry. Industry can be a rather vague and subjective term so you have to be clear about what your definition is. It's part of what makes anti-trust regulation difficult because to determine if a company is a monopoly you first have to define what industry you think they monopolize. Sometimes this is easy but sometimes it's incredibly hard to pin down.

For example is PC operating systems an industry or should we use a definition of industry that applies to all general purpose computing devices? In the former definition Microsoft is arguably a monopoly but in the later it clearly is not. The consequences of these sorts of (sometimes subjective) choices are not trivial.

Re:Define "industry"

[DalM]

I agree, definitions are tricky.

But my point was attacking the "Robots are coming" people. AI and robots will take a lot of jobs. But there will be more jobs created as a result.

Consider farming. Today, very few people are actual farmers. You probably don't actually know a person who has ever plowed or harvested a field. But you probably know lots of people that work in in the agricultural industry. There are people that make and service the machines, sell insurance, find temp employment. There are LOTS of different ag jobs today. And of course there are. It's a multi-hundred billion dollar industry in America. We aren't a farming society, but we still eat and it take a lot of people to make that possible.

Re:c6gunner IMPERSONATING me again?

[Anonymous Coward]

c6gunner please forgive me. OFF MEDS is just not a safe place for me yet

* You're AOK c6gunner & a smart insightful guy.

(PUTTING PILLS IN MY MOUTH NOW fixing what I can't on my own)

APK

P.S.=> Combination of depression, autism, and sexual dysfunction really brings me down sometimes. So sorry... apk

Linear extrapolation and FUD

[sjbe]

AI and robots will take a lot of jobs. But there will be more jobs created as a result.

Agreed. The whole fear of AI and robots seems to have little basis in evidence and seems more like linear extrapolation run amok.

Consider farming. Today, very few people are actual farmers.

Quite a lot of people are actual farmers. About 31% globally or around a billion people globally. Where your statement is correct is in rich industrialized countries which are comparatively automated but even then it is only a relative statement. In the US there are currently several million farm workers which isn't a trivial number even today. The most labor intensive farming has (like other industries) moved to locations with cheap labor when possible while industrialized economies utilize quite a lot of automation very much like manufacturing. The industry is growing even while it's share of total employment falls. The automation is exactly what enables these economies to do something other than mostly just farming.

You probably don't actually know a person who has ever plowed or harvested a field.

Not only do I know people who have done those things I have both friends and family that own farms and I live in an area where farms are common.

Define task...

[skaralic]

Define task. Define machine.

But the capability curves cross soon

[presidenteloco]

Imagine an almost horizontal line, being the average perceptual and intellectual capacity of a person, or if you like, the average intelligence-guided physical action capacity of a person. A complex mishmash of capability that is a little hard to quantify, admittedly, but that's why I said "imagine". It's a thought experiment, with some simplifications.
The line is almost horizontal because human innate capability improves noticeably on evolutionary timescales (100,000s to millions of years).
Now this human capability line actually starts to curve upward a bit, as human cultural memes develop, like education, external information storage, reliably complex communication between people and generations. So human capability, in recent centuries trained by better-informed minds, starts increasing faster than body evolution and brain evolution would seem to support without all the extra information system evolution.

Now imagine a horizontal line that starts well below the human line. This is the "independent-of-human-wielders" capability of artificial tools and machines invented by humans. That line stays really low for a long time, because all such tools/machines needed total human guidance to do their thing, for centuries. Then with the industrial revolution, increasingly automated machines did a lot with less human supervision but still needed a lot of repair, care, and tending by humans. And other humans were needed to manage and calculate about the economic processes supported by those partly automated machines.
So jobs were created as jobs were destroyed, and productivity per person improved. The line of machine capability curved upward a little bit, with the industrial automation.

But now imagine rapidly advancing computer-algorithm-guided machinery, including calculating and modelling machinery, and physical-process machinery. And imagine that the algorithms are increasingly crafted to train themselves and learn and model their environment themselves (still a way to go, but that's clearly the direction and rapid progress is being made.) Imagine that computer perception and cognition, initially in limited-scope areas, but generalizing rapidly with new AI inventions, becomes comparable to human cognition, or in some areas, exceeds it. And imagine that computer cognition guiding increasingly flexible physical-process systems (robotics, automated assembly lines, automated transportation, 3D printers, robot building-assembly systems, etc. There is still a need for human engineering to improve these systems, and manage their orchestration at the top level, and there is still a need for a few repair people for tasks still too variable for the AI and robots to handle, but in general, this whole thing can run by itself and produce stuff and move stuff and people.
So what we see is the line of general capability (intellectual and physical) of the AI and automated systems/processes starts to curve up rapidly, at the speed of the collective human scientific and engineering advances that invent this technology.

At some point, the machine line crosses above the human line, at which point the average job is more cost-effectively, and often just more effectively, done by the AI and automated system.

So no, more interesting careers will not come along with this, except for the odd uber-genius who can keep up. The automated (economic) system (as a whole) will just start to run itself with less and less human intervention.

The lines cross, and then everything is different. This doesn't happen suddenly, but blurred over a 50ish or whatever year period, but we are now well into that period. Our thinking, and economic modelling, about all of this will have to change profoundly. Anywhere your micro-economic or macro-economic model talks about an effect on human labour or income or spending power, cross that out and insert "machine" instead.

Re: Specious as always

[presidenteloco]

Technology creates jobs, until the technology starts becoming more capable by itself, in more and more areas, than people, then technology creates more jobs for other technology, not for people. That's what people can't seem to get into their obtuse skulls.

There seems to be one of those "denial" psychological problems going on. It's too terrible to contemplate, therefore we'll deny it.
Global warming due to human GHG emissions. DENIED. Check.
Job reduction trend due to better-than-human automation. DENIED. Check.

Re: Specious as always

[presidenteloco]

What percentage of all the able adults in your neighborhood, or your city, would be capable of doing those jobs, in your estimation? What percentage of them will be needed, to work on those new jobs you're thinking of?

Re:oh bullshit

[Xenx]

While you're not entirely inaccurate, your assessment of it being a small incremental change is wrong. Yes, a lot of the middle/upper management positions aren't changing. It's all the lower tier jobs, that employ the most people, that are changing. If one or two people can maintain automation for what used to take 10 people... You have 8 people out of work. Even if that efficiency change caused the company to double their output, you're now at 4 people and 6 out of work. Sure, one or two people might then also be hired, higher up in the chain. That's still leaving 4-5 without work. That's half as many workers employed at the lowest level. This is just for a single group of 10. There is an upper bound to the demand of whatever product. It's ultimately going to be a lot less people employed.

Re: Specious as always

[iggymanz]

there are jobs people with such "IQ" can do. you are looking for excuses to turn someone into a dependent adult baby, are you a big city Democrat needing bigger voting bloc?

Re: Specious as always

[iggymanz]

Amazing the huge percentage of people here who believe the lie that technology causes decrease in employment. It causes change in types of jobs, but ALWAYS creates more net jobs.

Bureacracies will take up the slack

[aberglas]

Consider computer automation, which over the last 50 years has been just as widespread as agricultural automation.

In the 1950s, when Parkinson wrote his great paper, bureaucracies like banks and tax offices were almost completely unautomated (yes, there were punch cards). Rows of clerks with mechanical adding machines reconciling banks statements and tax returns.

Yet bureaucracies have grown, not shrunk, as a result of all this automation. The reason is that while the human gut can only consume a certain amount of food, human society has an unbounded appetite for rules and regulations, processes and procedures.

So as automation frees up labor, bureaucracies will grow. Because they can. As predicted long ago by Parkinson.

Re:Bureacracies will take up the slack

[bluefoxlucid]

Of course some industries grow, others shrink, and others appear out of nowhere. They don't always all grow; it depends on if technical progress exceeds consumption.

also surprised

[mschaffer]

Since McDonald's, alone, already serves about 9 million pounds of fries per day, I am surprised that it has taken this long for machines to overtake humans in number of tasks performed.