In depth: Will robots like Honda's Asimo steal our jobs

Introduction

Everyone in IT is scared of the same thing – becoming obsolete. Just like those floppy discs, the Windows 98 machines and even those USB thumb drives kicking around the office, everything gets replaced. It gets upgraded. It gets forgotten. Is your job about to be automated, perhaps by an artificially intelligent robot?

Our survey says

The word 'robot' is, of course, rather emotive, and conjures images ranging from the cute Wall-E and NASA's Curiosity rover to evil forces like the Terminator. But robots is shorthand for machinery and for artificial intelligence – and both are a real threat to jobs.

A recent report called Agiletown: the relentless march of technology and London's response from Deloitte brings together a survey of the plans of 100 London-based companies with research from Oxford Martin School academics Carl Benedikt Frey and Michael A Osborne on the potential impact of automation on jobs in the UK and London over the next 20 years.

Its main finding – that 35% of existing UK jobs are at high risk of replacement in the next 20 years from technology, automation and robotics – is truly shocking. Even worse, lower-paid jobs (defined as below £30,000 a year, which is around $45,000, or AU$55,000) appear to be more than five times as likely to be replaced compared to higher-paid jobs.

"Unless these changes coming in the next two decades are fully understood and anticipated by businesses, policy makers and educators, there will be a risk of avoidable unemployment and under-employment," warns Deloitte's senior partner Angus Knowles-Cutler.

The good news

There are, however, reasons for the IT industry in the UK to be cheerful. A UK government report in March 2014 called Working Futures stated that the country's IT workforce – currently reckoned to be in the region of 100,000 – will rise by a healthy 15% come 2022. Even better, the Deloitte report included evidence that 73% of London businesses plan to increase their headcount in the next five years.

In the short-term, robots will not be taking your job. Phew

Who is under threat from computerisation or automation?

All this isn't inevitable, say the researchers Frey and Osborne, who identified three 'bottlenecks' to automation and computerisation that reduce the risks to jobs. No robot yet has skills of perception and manipulation, of creativity, or of social intelligence. If you're in senior management/financial services, computers/engineering/science, education, legal services, community services, the arts/media and healthcare, you're safe – for now.

However, if you're in sales, services, office and admin, transport, construction and production, you should keep an eye out for robots. They're already in our supermarkets, and could be coming to our cars.

Even if you're safe for now, it may not be long until your job is under threat from robots. Microsoft predicts that a small percentage of skilled tech workers will drive innovation for a highly internet dependent world.

"In 2025 there will be nearly five billion people online and more than 50 billion connected devices," reads Cyberspace 2025: Today's Decisions, Tomorrow's Terrain. However, the report's bombshell is skills shaped – by 2025 emerging economies will produce 16 million science, technology, engineering, and mathematics graduates annually, nearly five times as many as from developed countries.

However, it's not just in the West, and not just a 'skills gap' that is encouraging more and more robots. China's Foxconn – the country's largest private employer – is likely to be a role model in the coming era of increased automation. Its CEO Terry Gou has promised to augment his million-person workforce with a million robots in the coming years. Foxconn, which assembles Apple products among many others, wants its 'Foxbots' to take over assembly line duties from humans, according to Business Insider.

"We have over one million workers. In the future we will add one million robotic workers," said Gou in June. "Our [human] workers will then become technicians and engineers."

Foxconn's plans are about combatting the rising cost of labour. The initial boom in robot production is likely to come with an ironic sting – the displaced workers at places like Foxconn could get short-term assembly line jobs in the robot-maker factories. After the initial boom is over, there would be huge unemployment unless the local population either skilled-up, or were 'swapped' for more skilled workers.

True AI

Where do robots work?

In dirty, dull, dangerous places in industrial counties like the US, China, Japan and Korea, though countries like India, Thailand and Central/Eastern European nations have recently been adopting robotics, too.

Factories making everything from cars, planes and trains to TVs, cameras and most household appliances use robots to permit both dust-free assembly, and increased accuracy. Critical industries like consumer electronics, food, solar and wind power, and advanced battery manufacturing are all heavily robotised.

Robots are present elsewhere, too. Drones are robots, while tracking and surveillance robots like the Samsung SGR-1 claims to have no blind spots and can – in theory – identify and 'dispatch' an enemy soldier without human intervention. About 300 iRobot PackBot Tactical Mobile Robots remove unexploded bombs and mines from war zones and collect forensic evidence in Iraq and Afghanistan, and have worked in the tsunami-damaged Fukushima nuclear power plant in Japan. The technology comes from a prototype Martian rover named Rocky-7, a test bed for NASA's Mars rovers Spirit and Opportunity.

iRobot also has robot cleaners like the Roomba, and even a prototype for a fully mobile robotic helper around the home, called Eva, which responds to human instruction and starts conversations. The next step is the 'robo-nurse', home help robots that can dispense medicine, diagnose medical conditions, and can even keep patients company with small-talk.

Could robots actually create jobs?

"One million industrial robots currently in operation have been directly responsible for the creation of close to three million jobs," claims a study by the International Federation of Robotics. It's reckoned that the growth in robot use over the next five years will create another million jobs, too.

Freeing up workers from menial tasks to concentrate on the 'complicated stuff' might threaten jobs, but it's hard not to equate it with some kind of progress. However, what happens when engineers achieve human-level machine intelligence?

Swedish philosopher Nick Bostrom – Professor at the Faculty of Philosophy and Oxford Martin School at Oxford University (and Director of the Programme on the Impacts of Future Technology) – mulls over the implications in his new book Superintelligence: Paths, Dangers, Strategies. They're not good.

When will we have human-level machine intelligence?

"There's a big chance that we'll have it [human-level machine intelligence] in this century, but a lot of uncertainty as to the exact date," Bostrum told New Republic. "But what I'm interested in trying to figure out is, what happens when we reach that point?"

Wages could fall to catastrophically low levels as 'digital minds' – which need little or no sustenance compared to 'biological minds' – mushroom in number.

How close is true AI?

About twenty years away – just as it's always been. Bostrom makes the point that two decades is a "sweet-spot for prognosticators of radical change", but that the past over-predicting of the birth of true AI doesn't mean that it will never be developed. After all, a computer able to play chess was once thought to be the pinnacle of human-like AI.

"Chess playing expertise turned out to be achievable by means of a surprisingly simple algorithm," writes Bostrom, who speculates that general reasoning ability, or some key ability in programming, could also be found to be just a simple algorithm.

Bostrom is not worried about jobs for humans per se, but about the biggest issue of all; our dominant position on the planet. Ultimately, he compares it to the current situation of gorillas, whose future now depends on humans, not on gorillas themselves. Could we one day lose – to super-intelligent computers and robots – the capability to determine our future as a species?

How we can stop the robots

Although the long-term outlook is grim, Bostrum has a word of advice for would-be robot-builders. "We do have one advantage," he writes. "We get to build the stuff."

However, that's where the good news ends. "The control problem – the problem of how to control what the super-intelligence would do – looks quite difficult. It also looks like we'll only get one chance. Once unfriendly super-intelligence exists, it would prevent us from replacing it or changing its preferences. Our fate would have been sealed."

from TechRadar: All latest feeds http://ift.tt/1BYUGhO

via IFTTT