https://www.globalpolicyjournal.com/blog/13/02/2019/jobs-future-protecting-labour-market-face-ai-revolution
Technological progress, particularly the development of artificial intelligence (AI), has the potential to drastically change our societies and our labour markets. Whilst AI could make a large proportion of jobs redundant, it also offers opportunities to reshape labour markets and improve the ways in which we work. However, we still know very little about the impact AI will actually have on labour markets – predictions in the academic literature vary wildly, with some claiming that 47% of US jobs are at risk of automation (1), while other work arrives at a prediction of a comparatively small 9% (2).
This is why our research set out to review the literature on labour market effects of automation and AI, critically evaluating the assumptions behind it and the paradigms that shape different predictions. We found that while much of the literature uses highly rigorous methods to estimate the labour market impact of AI, authors end up with drastically different conclusions because there are simply many things we do not yet know about how automation will affect labour markets. These include how the process of automation will proceed in countries with different labour market structures, how long it will take for AI to overcome certain "bottlenecks" such as social intelligence and creativity, and whether AI will be able to replace some occupations entirely, or merely certain tasks within occupations. Moreover, a holistic assessment should estimate not only rates of job replacement, but also the potential for job creation in new sectors and occupations made possible by technological progress, further complicating matters.
While we conclude that cautious estimates may overall be more appropriate, there is no denying that AI has the potential to fundamentally transform our labour markets, and a number of valuable conclusions can be drawn from the literature. AI could disrupt labour markets significantly, and the burden of adjustment will fall mainly on workers in low-skilled, low-wage jobs that will be most easily replaced by technology. Yet the focus on job destruction should not stop us from recognizing the possibility of job creation through new technologies, and it is imperative for policymakers to harness this potential. Finally, a focus on the employment impact of AI as a "numbers game" is too narrow – technology will have large consequences for the ways in which we work, and if the challenges it poses are met, could inject greater flexibility and opportunity into labour markets.
Focusing on the UK policy context, we argue that so far AI has been largely seen as an investment opportunity, with comparatively little attention paid to its potentially disruptive socioeconomic impact. Drawing on international examples, our white paper identifies a number of policy recommendations to address the challenges posed by automation, focusing on labour market and education policy:
1) Assigning Responsibility: Dealing with AI requires coordinated government action across a range of different departments affected, necessitating the establishment of an "AI Tsar" who can draw together different government bodies for an effective and coordinated response.
2) Education and Training: While the UK government has identified investment in STEM education as a priority, more can be done to encourage take-up of STEM subjects among pupils, such as closer embedding of career advice in schools. Moreover, more investment is needed to address the inequitable socio-economic take-up of STEM education. Crucially, however, an exclusive focus on STEM education neglects the fact that collaborative and creative skills remain one of the fields in which humans are expected to retain a comparative advantage over technology in the medium term, necessitating a focus on take-up of these types of skills. We recommend a variety of measures, including the bridging of gaps between the arts and sciences in schools, and raising the profile of caring work as a career through improving its economic and social status.
3) Job Quality: The disruption to labour markets caused by technological progress raises concerns over job quality, and it is imperative that the government smooth increasingly frequent transitions between jobs and rethink a social security system built on the assumption of continuous careers over people's working lives. This necessitates revisiting the incentive structures linked to unemployment support and anticipating effects of longer transitions on pensions and savings, as well as a particular focus on raising job quality in growth sectors such as care work. Moreover, we recommend wider consultation on and greater employee engagement in the improvement of working conditions.
4) Skills and Retraining: The government needs to take action to ensure that workers can update their skills throughout their working lives. Measures that can be taken to make this possible include greater investment in lifelong learning, creating incentives for employers and private contractors to run training programmes, ensuring that necessary infrastructure such as broadband coverage is in place to smooth the transition, and increasing the focus of sanctioning systems on matching people to long term employment.
5) Sharing the Gains Fairly: While the measures outlined can mitigate the potential negative socio-economic consequences of automation, low-skilled and low-waged workers remain most likely to suffer during the transition. Several measures to ease this burden could be explored such as a "robot tax" or, through small-scale trials, a universal basic income.
-- via my feedly newsfeed