Here Come The AIs To Make Office Workers Superfluous
from the are-you-next? dept
Stories about robots and their impressive capabilities are starting to crop up fairly often these days. It’s no secret that they will soon be capable of replacing humans for many manual jobs, as they already do in some manufacturing industries. But so far, artificial intelligence (AI) has been viewed as more of a blue-sky area — fascinating and exciting, but still the realm of research rather than the real world. Although AI certainly raises important questions for the future, not least philosophical and ethical ones, its impact on job security has not been at the forefront of concerns. But a recent decision by a Japanese insurance company to replace several dozen of its employees with an AI system suggests maybe it should be:
Fukoku Mutual Life Insurance believes [its move] will increase productivity by 30% and see a return on its investment in less than two years. The firm said it would save about 140m yen (?1m) a year after the 200m yen (?1.4m) AI system is installed this month. Maintaining it will cost about 15m yen (?100k) a year.
The Guardian article quoted above gives a few more details:
The system is based on IBM’s Watson Explorer, which, according to the tech firm, possesses “cognitive technology that can think like a human?, enabling it to ?analyse and interpret all of your data, including unstructured text, images, audio and video”.
The technology will be able to read tens of thousands of medical certificates and factor in the length of hospital stays, medical histories and any surgical procedures before calculating payouts
It’s noteworthy that IBM’s Watson Explorer is being used by the insurance company in this way barely a year after the head of the Watson project stated flatly that his system wouldn’t be replacing humans any time soon. That’s a reflection of just how fast this sector is moving. Now would be a good time to check whether your job might be next.
Follow me @glynmoody on Twitter or identi.ca, and +glynmoody on Google+
Filed Under: ai, future of work, japan, jobs, robots, watson, white collar workers
Comments on “Here Come The AIs To Make Office Workers Superfluous”
Hasn’t Mike already been replaced by a google shill bot?
Re: Re:
Based on what?
Re: Re:
What’s wrong? Your employer not like someone challenging their stranglehold on cultural influence.
Don’t worry – you’ll be automated out of a job soon enough. Then you won’t have to worry about your employer’s views at all!
Re: Re:
If he has, it’s a Google shill bot that mysteriously criticises Google on a regular basis. They should get a refund.
Re: Re: Re:
That’s the problem with AI. You think you’re creating Skynet and what you actually get is Marvin.
Re: Re: Re: Re:
I’d prefer Max Headroom
Replacing insurance company adjusters with a computer would be simple. Just hardcode “claim rejected” and there ya go.
Re: Human obsolescence
I used to work for an insurance company. We’ve had programs for this kind of thing for ages, so no biggie. AI’s can make decisions based on cold, hard facts, but they can’t do nuance. Basically, they make your job easier, so you can get more done more quickly, but they can’t do your job for you.
Even if it’s ultimately possible to replace human workers in insurance, good luck with replacing us in facilities maintenance. You could program an AI to book staff to attend, to run up quotes, and to bill the client, but I don’t think you could program it to do the human interaction that makes our industry work.
There’s no real value in making humans obsolescent: sooner or later there will either be a catastrophic backlash against this kind of thing or we will find a way around it. I’m just about optimistic enough to believe we will find a way around it.
Re: Re: Human obsolescence
Of course there’s real value in making humans obsolete in insurance… it’s much harder to subpoena an AI about falsified documents in denied claims…
Re: Re: Re: Human obsolescence
It may be easier to subpoena the records of everything that went into the AI’s decisions. Especially if the programming is rigged. Reproducible inputs, and outputs using the same software and learned training data.
In fact, computerized systems may make it impossible to hide things that were once deliberately concealed in corrupt human brains behind claims of “I don’t seem to recall”, “I have no memory of that”.
Re: Re: Re:2 Human obsolescence
Sorry prosecutor … a hacker wiped out our entire archive
HR
I say start with the HR department.
Re: HR
Well that would make it easy to bring in the Robots because without HR, how will anyone be hired?
Japan is way above on the curve when it comes to using Robots for all kinds of things. Time will tell if this works out for them. I don’t think people in the U.S would put up with a Robot telling them NO!!! They would request to talk to a real person.
At least I have a job, no robot could ever replace me with. But there are a ton of jobs that a Robot could easily move on in. Someone is going to have to Fix all these robots. That’s going to required a skilled person making good money. Still will be far more jobs lost then gained.
Re: Re: HR
GMail and many other services and companies have spent over a decade teaching people that they can’t talk to a real person. Insurance companies have been teaching them that a real person might as well be a robot.
Nope. I did board level repairs in a computer store in the mid ’80s. That was gone by the late ’80s. Boards and power supplies were simply replaced; never repaired. Modern televisions might get a circuit board swap, but otherwise they’re simply replaced altogether at the service depot.
When the arm on your robot breaks it’ll be easily detached and replaced by relatively unskilled labor. Or possibly by a robot, autonomous or remotely piloted by third-world labor. The old arm might get shipped back to the overseas factory to be refurbished, but probably it’ll be thrown out.
Re: Re: Re: HR
Sure, you can talk to a real person about your grievances .. thing is, that person will be a reporter not a PR rep.
Re: Re: Re:2 HR
But, I DON’T WANT to talk to a CALL CENTER. I want to speak to a HUMAN BEING!!!
Re: HR
This is less funny than it seems. When you look at the classes of jobs AI’s are predicted to replace they tend to be jobs where all the upside rests with the company and all the downside with the customer. These are things like helpdesks, delivery jobs (taxi, courier etc), HR/Payroll etc.
HR departments suck not because of the humans but because companies hate spending a cent more on them than is strictly necessary and heck, no-one in the C-suite ever has a problem getting HR to jump when they say jump so how bad can it be? HR won’t be better if you replace them with an AI
Re: Re: HR
Once the public becomes upset, you won’t believe the ingenious ways that people will find to screw with or just plain sabotage the robots.
Well, maybe not pizza delivery robots that cook the pizza during the delivery. Timed so that it just pops out of the oven as the robot reaches your front door.
Costs?
The dollar figures in this article seem a little optimistic – IBM hardware/Software + Large Corporate IT department + new technology implementation = $1.4 million + $100k ongoing. It’s not clear what those dollar figures are paying for but that seems pretty cheap.
Re: Costs?
It also diffuses blame and risk.
If an employee messes up, rejects a claim that they obviously shouldn’t and it makes the news, the company gets the blame. If the computer messes up, well, blame IBM.
Re: Re: Costs?
IBM, or other supplier will probably point blame squarely back at the company, with facts and data to prove it.
On one level, it would be like blaming DELL. Hey, the machine just does what you program it to do.
It’s a similar problem, just a different level of abstraction. If the company is using AI to game the system, there will be data to prove this.
Re: Re: Re: Costs?
You might prove it to a qualified analyst. But not necessarily to a juror.
I’m not talking about blaming IBM for the hardware. We’re talking about IBM’s AI software. With no-one at the insurance company truly understanding how it works, blaming IBM would be easy.
Sure, IBM would point the blame back at the insurance company. That’s not the point.
An AI that can read a doctor's handwriting?
Wow, I’m impressed.
Re: An AI that can read a doctor's handwriting?
What if the doctor is connected to a printer that prints out whatever prescription the doctor calculates?
It's that time again.
Time to break out the sabots.
Re: It's that time again.
Yep. I think companies will underestimate the ways people will screw with AIs and robots.
Most jobs will be automated over the next few decades. As long as tech keeps improving it is inevitable. Even before that happens we will have to have some way of dealing with the unemployment. What does a country do when 20percent of its working age citizens aren’t qualified and/or needed for anything?
The whole “your spending is my income and my spending is your income” breaks down real quick. Tech advancements can be a bitch.
Power companies are starting to get nervous based on similar economic issues. What happens to rates when 10 or 20 percent of your customers provide their own power or never even connect to your service? Rates go up which drives more people to provide their own power. The death spiral in action.
Now imagine the two things I mentioned above working together to really screw people. People with resources/money decide to provide their own power thus indirectly raising rates on those who cannot afford to do the same. These people who now have to pay higher utility costs are probably going to be the ones who also lose jobs due to automation and AI, double whammy.
Capitalism as we know it cannot deal with the tech changes we are seeing or about to see without a lot of people becoming “losers”.
Re: Re:
“Most jobs will be automated over the next few decades”
Many pipe dreams end in confusion and misery.
Re: Re:
I think the answer depends on which political party is in power.
And what if it is more like 80 % instead of 20 % ?
As the Finland minimum income experience has already resulted in such discussion, we are already contemplating the scenario where humans are not needed to do most jobs. The solution will be to actually tax the companies and the rich and introduce the Finnish experiment everywhere. Who pays? The ones profiting the most by eliminating humans from the productive chain. And if you disagree then tell me: who is going to buy the products when most of humanity is unemployed due to AI/Robots?
Interesting times those will be.
Re: Re:
Anti-socialist panic in 3…2…1…
Re: Re: Re:
If you were to believe the conspiracy theorists…
Almost nobody will be left buy those products.
As a secondary consequence all but the top-most rich would gradually go out gradually as the economy contracts (fewer and fewer jobs).
Worst case, the super-rich build robotic enclaves with personal armies & minimal amount of maintenance staff and we move to some sort of neo-feudalism.
All those who “aren’t needed” end up starving or revolt and clash with the private armies of the rich.
Re: Re: Re: Re:
I wouldn’t go as far as these theories. There will be jobs. Just not nearly as many as the world needs (we already don’t have jobs for everybody though this could be reduced, not eliminated, by fixing some problems). I also don’t think the private armies will be widespread among the rich but I do think there will be private armies for some. My friend worked in a hotel where the top head of Exxon Mobil stayed. He booked the whole floor and his private security strictly controlled access to that floor. So private armies already exist.
What can’t be denied is that money is getting more and more concentrated in the hands of very few. A few years ago 288 people had half of the money circulating in the world. Today it’s only 8. This isn’t sustainable.
Re: Re: Re:2 Re:
I should clarify that I’m not talking about total world money as in half of the GDP or something, if memory serves the figure excludes money not readily available. Still the concentration exists.
Re: Re: Re:3 Re:
Aha, found the article!
https://www.oxfam.org/en/pressroom/pressreleases/2017-01-16/just-8-men-own-same-wealth-half-world
Re: Re: Re:2 Re:
Whilst I agree in general about this trend – I don’t see it as being particularly tech-driven.
As technology has improved this particular “tide” has gone in and out. The 18th and early 19th centuries saw increased concentration of wealth based on new technolgies – but the late 19th and early 20th centuries reversed the trend and brought (to the US and W. Europe) the universal prosperity that we still enjoy the remnants of today.
It has been globalisation and the replacement of local workers with cheap third world labour that has done the most damage – not the advent of machines. In fact cheap labour has actually held up the advance of technology – because it is easier to deploy than robots.
Re: Re: Re:3 Ludd was early.
Mechanical automation replaces human muscles. After the industrial revolution humans were still far more efficient and capable at mental tasks, and so were employed to manage the machines. Computational automation replaces human brains. When a machine can perform the mental / computational parts of a job as well as the physical / mechanical parts, and perform all of those better, faster, and more efficiently than any human, what do you have left to offer? Your argument amounts to “We’ve never had this problem in the past therefore we can never have it in the future.” This time is different.
Re: Re: Re:2 Re:
” There will be jobs. Just not nearly as many as the world needs “
It is not inevitable, the problem is choice. The human race can decide to self destruct or not. Our “leaders” are not interested in leading us in the direction required for the latter. They seem to have tunnel vision or simply do not care. Like squirrels stashing their nuts, the rich are building bomb shelters and other walled garden type fortresses. Much to their dismay, those things they want so badly will not help.
Re: Re: Re:3 Re:
And the anti-government anarcho-capitalists are helping this along in the name of saving us from socialism.
We need to stop labeling things and think logically about how to ensure the best possible deal for as many of us as possible. Taxing the rich fairly isn’t wealth confiscation: they’re confiscating ours by not paying us to live on. It’s very hard to look for another job when you have to keep yourself available on a zero-hours contract in case you’re wanted. Thankfully I’m not in that situation but my husband is and it sucks.
Oh Dear - that isn't what actually happens at all!
Here Come The AIs To Make Office Workers Superfluous
Sorry – that isn’t what happens – for 2 reasons:
1) Alongside the plan to use a new technology to replace workers come another plan to use the technology to DO MORE than the previous workers did. (The quoted piece more or less admits that.) In the end that tendency will not quite restore all the jobs – BUT – it will do more than that when combined with
2) Empire building. New departments will be created to manage and procure the AI and because humans will still be in charge and their prestige depends on "having people working for them" these departments will not stay small long and will spawn subdepartments where the same processes will happen.
My old schoolteacher used to say that no machine had ever been invented that had actually reduced the amount of work that needed to be done – and nothing I’ve seen so far will change that.
So – AI will not cut jobs until it takes over the CEO’s job and reason prevails in the company structure.
(Unless of course the AI is such a good impersonation of a human that it carries all the human flaws too!)
I for one wont mind losing my job to a bot. I can then sue companies for messing with my business plan of working until I die.
Bender – 2020
You meatbags had your chance
Kiss my shiny metal ass.
over concerns of the next administration, the US military decided it was time to turn on Skynet
Re: Re:
There will be a built in safety factor.
The President will have to send a five digit security code over twitter in order to activate Skynet.
That code probably won’t match what most sensible people use for their luggage.
Re: Re: Re:
12345
Already happening in "FinTech"
AI is now being used to invest your retirement savings.
I’m waiting for the explosion when people find out that the AI ‘bot “invested” all of your retirement savings in the penny stock market & lost it all.
Or the ‘bot invested all of your retirement savings in cigarette companies, because they appeared to have the best return on investment.
Or the ‘bot invested all of your retirement savings in coal companies.
But the execs at these FinTech companies will themselves retire with billions.