Why we should stop panicking about robots stealing our jobs

[Monirul Islam]

Ideology is what determines how you think when you don't know you're thinking. Neoliberalism is a prime example. Less well-known but equally insidious is technological determinism, which is a theory about how technology affects development. It comes in two flavours. One says that there is an inexorable internal logic in how technologies evolve. So, for example, when we got to the point where massive processing power and large quantities of data became easily available, machine-learning was an inevitable next step.

The second flavour of determinism – the most influential one – takes the form of an unshakable conviction that technology is what really drives history. And it turns out that most of us are infected with this version.

It manifests itself in many ways. A prime example is the way the political earthquakes of 2016 – Brexit and Trump's election – are being attributed to technology: if only Cambridge Analytica and other dubious actors hadn't weaponised social media, normal life would have continued. Hillary Clinton would be bombing Syria, David Cameron would still be prime minister and Jacob Rees-Mogg would just be muttering into his Veuve Clicquot.

While there's no doubt that social media played some – as yet unquantified – role in the upheavals of 2016, it seems implausible that the technology was the key element. Far more important were populist rage against the 2008 banking crisis – in which the wages of bankers' sin were paid for by austerity imposed on ordinary citizens – and the social carnage wrought in some regions of western societies by decades of neoliberal economic policies, globalisation and outsourcing.

But technological determinism not only colours the way we explain the past, it also affects the way we see the future. Take our current concerns about the impact of the next wave of automation on middle-class employment. In 2013, for example, two Oxford researchers, Carl Benedikt Frey and Michael A Osborne, caused a stir when they published a report arguing that "about 47% of total US employment" was at risk from the kinds of "weak AI" technology then available.

The "hollowing out" of the middle classes is alarming: a stable middle class seems a prerequisite for a stable democracy
The main reason the Frey/Osborne study generated such interest was that many of the job categories they identified as vulnerable were middle-class or white-collar. And the prospect of the "hollowing out" of the middle classes is alarming because, at least up to now, the presence of a stable middle class seems to be a prerequisite for a stable democracy.

Closer examination of the study, however, suggested that panic might be premature. What the researchers had done was to take the 702 job categories employed by the US Bureau of Labor and used machine-learning techniques to estimate the vulnerability of each to automation. Within its own terms of reference, it was a pretty good piece of research, and it had the useful side-effect of making policymakers and others sit up and take notice. But as a guide to what might actually happen, it wasn't much use. Why? Because there's more than a whiff of technological determinism about it.


Although the granularity of the research was an advance on much of what had gone before (702 occupations), it couldn't address the reality that decisions about what gets automated, and why, are usually local and contingent on lots of factors which have nothing to do with technology. The fact that a machine could in principle do a particular job doesn't mean that an entrepreneur or a company will make the necessary investment and embark on the disruption implicit in its deployment. Many jobs are in fact bundles of tasks, some of which are easily automated, while others are not. And so on.

So we need analyses of the potential impact of technology that are less deterministic. As luck would have it, one such study has recently emerged from the Organisation for Economic Co-operation and Development (OECD). It finds that across 32 of the 35 countries in the OECD, close to 50% of jobs are "likely to be significantly affected" by automation. But, the report says, "the degree of risk varies" and only 14% are "highly automatable" – ie the probability of automation in these jobs is over 70%. And there are significant regional differences: 33% of jobs in Slovakia are at risk, for example, but only 6% of jobs in Norway.

This difference highlights the inadequacy of technological determinism in helping us predict the future: in every case, one needs to delve deeper. The explanation for the differences between countries, for example, is probably that across the OECD, jobs that are superficially similar actually involve different occupational mixes.

The future of employment will, of course, be influenced by technology. But it's never the only force that shapes our destinies.


Source: https://www.theguardian.com/commentisfree/2018/apr/15/stop-panicking-about-robots-stealing-all-our-jobs