Technology has come a long way since Peter Drucker, having witnessed the first attempts to automate knowledge work in 1967, declared of the computer: ‘…It’s a total moron[i].’ Indeed, one widely cited paper by Carl Frey and Michael Osborne at Oxford University found that as many as 47% of jobs will be highly susceptible to automation over the next two decades.
Questions have been raised over the methodology used in this paper, with the OECD suggesting the method was too blunt and that, in fact, only 9% of jobs in the OECD were at high risk of automation[ii]. The differences between the two papers make for uncomfortable policy prescriptions since high degrees of uncertainty are not generally embraced by politicians. The extent to which cognitive AI will augment rather than replace humans is still to be settled. However, even 9% equates to millions of jobs lost and wider changes to the content of job typologies that do survive. Combined with a rise in freelance work, automation could easily reduce a wider percentage of jobs to a series of mini tasks that would easily be ‘auctioned’ off. The need to redesign economic and social systems is still pressing under such a scenario; the diffusion of work benefits from health insurance to pension provision would not work as intended, for example. Although the Swiss have rejected the universal basic income proposal, we are likely to see similar proposals elsewhere, perhaps even tied together with educational credits – which will surely be needed in the face of the need to reskill and retrain millions of individuals.
Although technology allows both more efficient ways of dissemination educational content (and indeed of creating more relevant curricula), there remain several structural issues in retraining such a wide swathe of people in a timely fashion – especially given the global backdrop of high youth unemployment and underemployment. Whether or not traditional educational models are agile or flexible enough to retrain and reskill at both a personalised level and at scale remains to be seen.
Lifelong accounts that comprise learning credits, benefit credits and other provisions have been floated and could better reflect the needs and wants of an individual than the currently strained welfare state. Issues of provision, tax and eligibility are complex and potentially have far reaching consequences – most notably on the notion of freedom of movement and immigration. Some, no doubt, view such proposals as a solution without a current problem. Given the rise of populism in key geographies and the acceptance by many of the idea that the current system is stacked against them, thinking about the nuts and bolts of a new safety net now would seem prudent. Automation, at either end of the forecasts, holds the potential to destroy the contemporary welfare state and/or result in costly and regressive policy action to preserve jobs. Automation holds great promise for the world, yet in order to benefit from it, we must actively explore a new social contract that doesn’t confine millions of workers to technological unemployment.