Innovation

Why automation could lead to mass unemployment and how a Universal Basic Income might be the only solution

As editor of a technology title, I receive hundreds of pitches each day on virtually every single conceivable topic. The hype surrounding certain technologies tends to come and go. However one area has been consistently topping the coverage charts for years: artificial intelligence.

April 28, 2017

Given AI systems have beaten humans at chess, the ancient Chinese game Go and even poker, it’s easy to see why people are so keen to keep up-to-date on how we’re measuring up next to computers intelligence-wise.

In particular one topic we regularly return to is the ‘singularity’, also known as artificial general intelligence: the point at which AI has intelligence comparable or beyond that of a human mind.

However, focusing all our fears on this prospect fundamentally misunderstands the risk AI poses.

The immediate threat is not that AI becomes cleverer than us and takes over the world. Experts within the field tend to say this prospect is not plausible or that it is very remote.

The real fear is that AI and robots become good enough at doing so many basic human work tasks that many existing jobs are automated out of existence. Estimates vary regarding jobs lost from 47 percent by 2034 (Oxford University) to 38 percent by 2032 (PwC) to six percent by 2021 (Forrester).

AI

One thing is clear: huge job losses are on their way. Anything that’s easy to automate will be eventually. After all, machines tend to be cheaper than wage bills.

Some tech CEOs, especially startup founders, like to insist that this is not true. Although some, like HP’s Meg Whitman or Microsoft’s Bill Gates, have admitted it’s a concern, many in Silicon Valley still appear to be brightly optimistic.

The standard script goes like this: AI or robots will ease the burden of repetitive or manual work, freeing those people up to focus on more useful, value-added tasks. It sounds like an appealing argument. You can have your cake and eat it, basically.

People also like to point to previous examples of automation throughout history: haven’t we coped with industrialisation, which was equally disruptive?

However it seems blasé at best to just say we don’t need to worry about it. While automation might be a great way to get rid of some of the small, specific, boring bits of your job (as a journalist I’d love to automate transcription), what if your entire job is automated? It’s not clear that we’ve ever experienced quite the same level and rapidity of change previously in history.

It might be appealing for a lawyer analysing contracts, but if you are a truck driver, it’s counter-intuitive to suggest driverless trucks are anything but a disaster for your job prospects. And office workers should not be too complacent: plenty of the work they do is likely to be in line for automation too.

Indeed, the International Monetary Fund (hardly a hotbed of left-wing thinking) has said technological advances are the number one thing responsible for why workers are earning less and less of the income pie.

President Trump famously blames trade deals for workers’ woes – but in reality adjusting to new technology is a much bigger issue when it comes to workers’ ever declining fortunes.

It’s a situation we are sleepwalking into. Research by bot platform LivePerson found that 87 percent of UK respondents were either ‘not worried at all’ or just ‘a little worried’ about losing their job due to automation.

1401x788-AI-opener

More concerningly, the UK government shows little sign of grappling with the issues either, distracted as they are with more immediate concerns like the general election and Brexit (despite the fact automation and job losses could potentially have a bigger impact than either of these).

If Theresa May has any interest in automation, or indeed in technology at all (beyond surveillance), she has yet to demonstrate it. Jeremy Corbyn has expressed interest in one increasingly popular solution: a Universal Basic Income.

UBI is a set amount of money paid without any conditions to every citizen in society, no matter their circumstance. It’s an attractive idea but not a silver bullet – depending on the way it is implemented and the context, it could even be used to cut benefits.

However, given the Conservatives 20-point poll lead and the upcoming June election, it seems unlikely we’ll see it put into practice in the UK any time soon.

Vast retraining programmes to cope with technology have been suggested as one way to cope but firm commitments from ministers have not followed, and it’s hard to imagine the money coming forward for it under continuing austerity.

We cannot rely on technology CEOs to act. Many of them they can’t see the problem with automation, think it’s a good thing or profit from not tackling it. Some have been upfront and rang the warning bell, but it’s not in their power to change the law.

So what could make us start taking the risks of automation seriously? It’s already leading to lower wages and fewer jobs, according to the IMF plus academics at Stanford and MIT.

So when do we start to act? How many millions need to be unemployed or underemployed before we start to take this problem seriously?

There are 1.6 million unemployed people in the UK (a historic low, to be fair) and almost a million people in the UK on zero hours contracts. However, wages have dropped by 10 percent since 2008. So people are in work, but earning less – and if Brexit is anything to go by, many of them are angry.

o-ARTIFICIAL-INTELLIGENCE-facebook

As someone who is fascinated with new technology and innovation, I am no Luddite and I am never going to argue we go back to a mythical happier, simpler time. However, we need to have a good, hard think about the consequences of automation and how we will cope with them.

So often it’s impossible to raise these sorts of questions within technology circles without being shouted down. But it’s something we have got to start thinking about and acting on, before it is too late.