Should Workers Be Able To Manage The AI Transition?
In a recent article, I looked at the process of transitioning from one career to another, which is a major topic when we’re in the midst of a technological revolution, as we currently are with AI. The challenge is, that while transitions are technically possible, in reality, they’re pretty hard, not least because people tend to get punished for doing so and seldom manage to transition into a similar level they enjoyed previously.
Pause for thought
This should make us pause for thought as we read the latest paper from the Brookings Institute, which confidently asserts that most American workers can easily transition from jobs at risk from AI to those that are not.
The paper suggests that there are over 37 million workers who are highly exposed to AI-related job loss, but that 70% of them should be able to transition to less risky roles. That leaves 6 million or so struggling to adapt, with those workers predominantly in administrative roles.
One’s ability to successfully transition was largely based on things like savings, age, the availability of local opportunities, and skills.
“We know from history that the adoption of these technologies and full integration of what they can do takes a while, on the order of multiple decades, to realize their full productivity advantages,” the researchers explain. “There are lots of reasons for that. There’s sluggishness in terms of institutional change and people being able to adapt their workflows and try and experiment with new technologies, learn when they’re reliable; when they’re not.”
Quick to adapt
The report argues that those most well placed to adapt include lawyers, software developers, and financial managers. Those in these professions typically have a good income, a diverse skillset, and the kind of financial buffer that will allow them to weather any transition. They also often have strong professional networks to help them hop into new disciplines. These things aren’t generally the case for those in clerical or administrative roles.
The authors acknowledge that any transition, whether for those better placed or not, will be difficult, and that governments should do more to help. For instance, they could make unemployment insurance more generous and long-lasting. They could also provide the unemployed with better career coaching to help them transition into a new role.
“English language skills could be a thing that is really valuable for one worker, whereas for another worker, they actually need to move physically to a new labor market,” the authors say. “Or they need to learn how to use AI much more in their workflow, to be competitive and productive in their occupation. This diversity of needs is a challenge, but is maybe something to try to embrace with investment in different active labor market policies.”
Shoulda, coulda
Of course, this sounds very nice in theory, but the reality is often very different. For instance, in a recent article, I explored the challenges that come when the skills we have become fundamentally mismatched to the new needs of the labor market. Harvard research finds a consistent loss of earnings of up to 16.5% a decade after people lose their jobs to automation. This was compared to their earnings trajectory had they managed to keep their job in the first place.
A second Harvard study shows that while retraining after displacement is generally better for people than not training at all, it’s still not really enough to allow them to continue on the same trajectory as before.
So it’s fine to say that people can transition to different roles should they find their work automated, but the reality is that even if they do so, their income will quite probably be lower than had they continued in their original career. And this impact isn’t a short-term adjustment, but a long-term decline of significant proportions.