In January 2017, a group of 34 insurance brokers at Japan’s Fukoku Mutual Life Insurance Company were told they’d no longer be needed at the firm: they were to be dismissed. Not because the role or department or branch was becoming obsolete, but because they were. Their jobs would henceforth be conducted faster, more efficiently and more cheaply by IBM’s Watson Explorer AI system.
If you can’t imagine the deadening blow of being told a machine can do your job better than you, perhaps you should try. Because we are softly treading at the tipping point of the automation age, and the chances are, they’ll be coming for you too.
The IPPR thinktank estimates up to 44 percent of jobs in the UK could be automated over the coming decades, equating to over 13.7 million displaced workers. The transport, manufacturing and retail sectors are forecast to be the first to be hollowed of human presence. But soon to follow are industries we never guessed would fall prey to our robotic usurpers: law, insurance, medicine, finance.
Some cry fear mongering and assert that it’s likely to be discrete tasks that will be automated, leaving workers free to focus on the most stimulating and challenging aspects of their roles. There will be new jobs created, say others, holding up 'app developer' or 'social media manager' as examples of positions no one could have imagined 20 years ago.
While it’s undoubtedly true that new jobs will be created, questions remain. At which end of the job spectrum and on what scale? Even more seriously, how long until these are automated too? Because it’s widely agreed that in a timescale of 50 to 100 years, robots will be superhuman. They’ll be able to do everything you and I can do but better and without requiring those pesky little things like wages or breaks or sick pay.
Optimists claim it will simply be different skills, such as social and creative skills, that will emerge as most valuable in the automated economy. But what about the (many) people who have neither? While it’s not yet certain who will lose the hardest, that there will be losers in this brave new world is indisputable.
To counter growing anxiety, one idea gaining traction within the EU is the so-called ‘robot tax’, which states that companies would have to pay a higher burden of tax in line with their rate of culling of human staff members.
On the surface, this doesn’t seem too bad an idea - if the funds were directed towards the retraining of employees or the funding of wider programmes aimed at easing social pressure created by unemployment, for example. However, there is the troubling question of how high to set the tax, because despite all of the risks, is AI-driven automation something we should actively resist?
“The UK economy is mired in a productivity crisis, where productivity has basically flat lined in the last decade,” says Alex Williams, co-author of Inventing the Future: Postcapitalism and a world without work and Lecturer in Digital Media & Society at the University of East Anglia.
He points out that despite record numbers of people being employed in the UK, the average output per hour is relatively low, arguing that AI could prove vital for boosting the UK’s position on a global scale.
In fact, at the moment there are whole swathes of the jobs market that could be automated but aren't, because it makes more financial sense for companies to pay humans a pittance to do them instead. Is the possibility of unburdening the workforce of the most repetitive, and mind numbing tasks like working in warehouses or call centres really something worth fighting?
Automation has the power to transform a number of industries for the better. Take autonomous vehicles as one example. Automation has the potential to push the uptake of electric vehicles, reduce car ownership and inner-city congestion. One million people in the UK currently rely on driving for their livelihood, but could the huge societal gains be worth rendering these jobs obsolete?
Who holds the reins
The transition towards an automated world could deliver huge benefits, but whether or not AI is leveraged for social good depends on who dictates the terms of this societal shift. “A lot depends on how the process of automation occurs and who gets to make the decisions about what gets automated,” says Williams, “what are the priorities? What are things we definitely don't want to be automated?
"Even more seriously, who is it that owns the machines or the intellectual property rights to the algorithms? Who is it that gets control and, hence, profit from this process?”
It’s clear that right now, those with the power to sculpt the narrative of an automated, post-work world are the tech behemoths currently racing to be at the forefront of the AI revolution. And this sets a dangerous precedent for the future of society.
To explain why, let's take a look at a trend currently troubling economists. It's most clearly encapsulated in a graph illustrating the steady decline in the share of total national income being channelled towards workers, and the share instead being directed towards corporate profits, shareholders and finance. In other words, the share of income going to ‘labour’ and the share going to ‘capital’.
While in the 1970s, an unwavering 75 percent of wealth was directed towards labour, since the 1980s, this share has been in decline. It hit its lowest point just before the crash of 2008 and hasn’t recovered since. This trend is depicted in a 2017 IMF report that attributes half of this decline to the impact of technology, and the other half to globalisation and finance. In fact, in economics parlance, this steady leeching of cash away from workers and towards business has been dubbed 'capital-biased technological change' (CBTC).
Because 'capital' is in the hands of investors, shareholders, asset holders and business owners, this means that more cash is accruing in the pockets of the already wealthy. In the UK, this effect is clear in the ever-growing disparity between the richest and poorest members of society. While the wealthiest 10 percent of households have amassed 45 percent of total private national wealth, the least wealthy half holds a paltry nine percent.
Technology is already contributing to the undervaluing of labour and the shifting of funds away from workers. Clearly, as automation reduces corporations' reliance on labour even further, there will be an even bigger shift in power towards the founders and shareholders of companies. It's already happening, and without any intervention, there's no reason it won't continue in the same way.
“In the first Industrial Revolution, we made some quite large structural changes to society to make sure the benefits were spread around,” says Malcolm Torry, Director of the Citizen’s Basic Income Trust, pointing to the creation of unions and labour laws in the UK during the post-war period.
“We introduced universal education, so people were educated, and had the skills for the new jobs that came along.”
Re-skilling and education is the solution the IMF report settles upon as a means to counter the impact of CBTC. But for the robot revolution, there are greater limitations to what these could achieve. The industrial revolution took over manual jobs from humans, but what we have on our hands now is a knowledge revolution, leaving the question of what will be left for humans to do.
Admonishing workers to be more adaptable simply puts the onus on citizens to render themselves somehow more employable in a world that is fast changing due to global forces entirely beyond their control.
At the same time, the protective initiatives that were introduced after the first industrial revolution are based on principles that are fast being eroded today. The rise of the gig economy, the redefining of employees as self-employed ‘contractors’, the abundance of zero-hour contracts and the requirement for in-work benefits all combine to form an unsteady foundation upon which to build a fair automated world.
It's clear that we can't expect big business to suddenly grow a conscience and begin behaving altruistically. But would leaving a large portion of the workforce jobless and wageless be a desirable outcome for them?
“Historically, the economy is run on a consumerist model based on giving people the resources with which to buy products," says Williams. "It's how the whole system keeps ticking over. If you have lots of people without a job and no way of buying anything, the economy won't work.”
Clearly, if the power tips too far in favour of business, unemployed people wouldn’t be able to afford to engage as consumers in this new society, meaning automation could precipitate the implosion of the capitalism itself. Perhaps that won’t matter for a technocratic superclass that can live in self-contained luxury communes patrolled by robot guards. But what about the rest of us?
The robot revolution: it may happen quietly, incrementally, but what's certain is that it's coming. So what are some of the most influential ideas for shaping a fair automated future?
Ideas for a post-work world
Universal basic income (UBI) is a much-lauded idea, often put forward as a salve for the deep gashes automation is set to tear into the jobs market. In its simplest form, this would be a basic wage paid to every citizen from birth (or the age of 18, depending whom you ask) until death. Beyond this, the specifications are the subject of intense debate. While some believe that this should be enough to live on, others argue only for a supplementary income to top up your salary, and while some argue that it should present a means to tear away the tangled mesh of bureaucratic welfare benefits, others argue that it couldn’t possibly work without incorporating some element of means testing.
“In the UK and various other countries, we've always added to a system rather than replaced bits of it. For instance, we started off in the UK with a fairly chaotic system of privately and charitably run schemes before the second World War along with some government benefits,” says Torry.
“We then also added a universal child benefit to that mixture, and I suspect that if in this country we were to implement a basic income, it would be implemented alongside the structures we already have.”
This system would have some obvious benefits. One of which being that in the current system, if people accept some form of work their benefits immediately reduce, often creating a situation where people are reluctant to take certain types of work at the risk of sacrificing benefits.
Another advantage would have deeper societal implications. “In our society at the moment, there's a great deal that doesn't unite us. There are some people who are on means-tested benefits and some people who are not which can cause stigmatisation of those who are on them," says Torry. "Whereas a citizen's or basic income would mean everybody would get it and there would be no stigma attached to it.”
But clearly, in an automated world, the feasibility of UBI would require a rethinking of how the government creates wealth. Currently, income tax is levied on working people, but with growing automation, the sum gathered through this channel will decrease, as the requirement for welfare ramps up. This will of course necessitate a shift in the tax burden from workers to companies. From labour to capital.
Some indication that tech companies are readying themselves for this eventuality comes in the form of tech billionaires like Facebook CEO Mark Zuckerberg and Richard Branson advocating for UBI.
“Automation might create problems for the ability of society to hold itself together and the overlords of Silicon Valley are very much aware of this as a problem," says Williams. "They understand that automation, machine learning, the lower end of AI that's coming in now - this has the potential to create serious social dislocation.”
Given Facebook’s notorious record on tax evasion, this could indicate a willingness to support citizens as long as they are the ones holding the reins of power. While they’re reluctant to contribute to government in its current form, as they pivot into quasi-governmental bodies themselves, they’re happy to provide for the proletariat and perhaps instigate a return to a pre-state system where a serf class subsisted on handouts from generous aristocrats.
Whether we should welcome the toying of tech companies with government-like capacities or not is a separate question. But as it stands, UBI has been shown to function well in the places it has been trialled.
Research from the Citizen's Basic Income Trust shows that a basic wage for all Britons would be feasible with the current restrictions on expenditure and shadow chancellor John McDonnell has said that there will be a universal basic income trial in the next Labour manifesto.
Citizen’s wealth fund
But some still see inherent flaws within a UBI model. “Most people frame it as, ‘what we're going to do is tax the income of these robot owning companies, and use that to distribute to the masses’,” says Duncan McCann, researcher at the New Economics Foundation.
“I have some problems with that conceptualisation of UBI, and what I think is, if we are going to move to a world of less work, what we actually need to do is broaden the ownership of the economic capacity of the country.”
McCann backs an alternative idea that could complement UBI, which is that of a citizen’s wealth fund. In practice, this fund would take the form of asking public UK companies over a certain size to issue extra shares which would be claimed by the citizen’s fund. Over time, this fund would amass 10-20 percent ownership of UK companies and be able to pay dividends to UK citizens as well as being reinvested in various initiatives. Unlike an extra tax, it would not act as a direct drain on a company’s cash flow.
Alaska currently runs a similar scheme, where wealth from the country’s oil industry is partially owned by all of its citizens and pays an annual dividend to each. But we can look closer to home for successful examples of sovereign wealth funds. The Shetland Islands made a deal with oil companies back in the ‘70s which agreed on annual disturbance payments in return for access to the North Sea. The proceeds have been channelled into a wealth fund managed by the Shetland Charitable Trust which is now worth £200 million.
Research estimates that if the UK had set up a similar oil fund (which had also swallowed the proceeds from privatising vast swathes of public infrastructure since the ‘80s), it would now be worth upwards of £600 billion.
While this may seem radical, similar ideas have been floated by mainstream political parties, even from the unlikeliest of ideological strains. The 2017 Conservative manifesto laid out a suggestion for a sovereign wealth fund, named ‘Future Britain Funds’. But while these were proposed to receive funds from ‘the sale of public assets’ or revenue squeezed from national resources, it’s not too far an ideological step to apply this more widely to public companies with the UK as their home address.
These funds are conceptualised to be distinct from the government, and instead controlled by the collective governance of citizens, diverging both from state-driven models decried as ‘old-hat’ and from the neoliberal inequality machine. Instead, it ties the interests of big business and citizens together - surely a huge imperative for the coming automation age.
“I think it's really important that it's a universal basic income based on some pool of assets that we collectively own, rather than a tax the rich and give to the poor," says McCann. "We need a universal basic income that really tackles some of the structural problems with the economy and doesn't just allow a few billionaires or ultra-billionaires to amass all the power, all the productive capacity, all the capital, and then basically drop some crumbs so that the rest of us can continue to live.”
A similar argument pivots on the ownership of mass data collections. Data has been hailed as ‘the new oil’, but although we all contribute to these vast databases, we are not remunerated and cede ownership at the point of creation. If we are feeding into the huge data collections that make intelligent automation possible, then why shouldn’t we claim a portion of the returns?
Whatever form the eventual solution takes, it’s becoming increasingly obvious that our current system is teetering at the edge of tenability and that automation may well herald the implosion of the current economic model. This means the undertaking of the ideological groundwork for shaping the world of tomorrow, must begin today.
The groundwork will involve a deep self-analysis of what kind of world we want to live in. There are two broad possibilities - one a dystopian technocracy where a tiny reining elite enjoys dominion over a reinstated serf class. The other possibility reimagines working life, with more flexible days spread between different activities and responsibilities, supported by some blend of UBI and communal ownership.
It’s not simply a question of what will we do after work, but what do we want humanity to be? It may seem counterintuitive, but the rising of the robots is set to force a reckoning on the fundamentals of what it means to be human.