People see automation as a threat to their jobs. According to a recent survey by the Northeastern University, three out of four Americans believe that artificial intelligence will destroy more jobs than it creates, and one out of four were worried about losing their own. Education could be an answer to this danger by upskilling students and workers, so they meet the new skills required in the job market, and thus, avoiding technological unemployment. In a study on the impact of AI by the former President’s Council of Economic Advisers of the White House, education was proposed to be the only solution to unemployment brought by automation. In this article, we will explore that option.
Some may argue that automation will bring new employment opportunities – although jobs will be replaced, new needs surrounding this technology will lead to employment creation. The “Luddite Fallacy” discusses that technology does not lead to unemployment, but simply redistributes the composition of jobs within economies. Moreover, a study by Deloitte after analysing 140 years of data concluded that technology had created more jobs than destroyed.
On the other hand, there is an established point of view defending that the number of jobs in the market is finite. If jobs are limited, and automation takes some of them, those replaced workers will lose their job – hence, the existence of technological unemployment. A report by MGI in 2017 estimated that half of all the current work activities could be automated by 2055, with a 20-year margin of error in both directions. Although this datum can be misleading, as some jobs will be fully automated while others receive no impact, it can give us a sense of scale on the urgency of the issue.
Balancing these two opposed positions might bring us to the solution of the problem. Even if the job market will balance after some jobs being automated, there is a time factor that may lead to immediate unemployment for a while until those new jobs are created. That unemployment will be the result of a mismatch between the skills required in the market for the new jobs, and the old skills that the substituted workers have – a skill gap. Therefore, the key to prevent technological unemployment will be to provide workers and students with the new required knowledge prior to these events, so they can fill that skill gap as soon as possible once they are replaced by machines.
After breaking down the reasons why education will be required to overcome this potential situation, we need to define what set of skills the job market will demand, what that future education will look like. Even though it would require an exhaustive analysis to detail these skills, we can intuitively categorise them into two subsets: technical and critical ones.
The first set is the most evident one. It refers to the knowledge that is required to technically adapt new technologies arriving at the workplace. In the case of artificial intelligence, for example, new-skilled developers and programmers will be needed to improve the software, statisticians that will develop new statistical methods, data analysts, and scientists, cybersecurity experts or engineers researching to find new areas of applications. Automation will open the door to new opportunities that can be exploited, and there will be a need for technical knowledge in this particular field to maximise these favorable circumstances – more employment opportunities.
Additionally, critical skills will be required — indeed, because machines lack of these. This is another opportunity that we could take advantage of to employ replaced workers, based on all the work that machines are not able to do. Pablo Picasso once stated that computers are useless, due to the fact that they only provide answers. And he was very right – without a critical human mind behind, computers are not of any use. There is always the need to have a person asking the right questions. Society will require critical thinkers — for example, lawyers that are able to deal with compliance regarding these machines, policy-makers who will be able to establish ethical protocols in the workplace or HR employee relation managers that will be able to understand human-machine interactions.
The most fascinating roles, however, will be created from the merger between these two skill sets. In fact, we get to the point where these two opposite areas of expertise overlap. Using the previous examples, even if a data scientist decides to focus on developing her/his technical skills, in order to provide better insights from the data results, she/he will need to have a critical mind. On the other hand, a lawyer who focuses on developing her/his critical skills, she/he will still need to have the required technical knowledge in order to understand the technology so that compliance requirements can be established or met. Instead of specialisation, the job market will look for multidisciplinary individuals. The education of the future is diverse, against our current model based upon limiting areas of knowledge.
But who should be receiving this multidisciplinary education? In the article we have referred to both workers and students – following the previous estimation by McKinsey, we can assume that 50% of work activities will be substituted by 2035 the earliest. Therefore, it would be a sensible choice to immediately start with lifelong learning education for those whose jobs are at risk of full-automation (all tasks automated) and are the most vulnerable (the easiest to replace) – the jobs most likely to being automated the earliest. Education should gradually spread towards more complex jobs (harder to replace by AI) and that cannot be fully automated (from jobs with the largest possible number of tasks automated to the least). On the other hand, it would also be sensible for educational institutions to shift towards a curriculum that will provide the workers of the future the multidisciplinary skills that the market will require. A change like that could save us from future automation.