In this week’s episode, we (Gary Gensler and Simon Johnson) discuss the potential impact of AI on jobs. This is a very important question as AI rolls out around the world. For now, though, it remains largely unanswered.
Both of us teach (different but related) courses on the economics of AI at MIT — and have been doing so for many years. In these classroom discussions, as in the broader public discourse, there are multiple issues to consider. What tasks within jobs will be most readily replaced by AI? Will AI largely automate generic tasks or also replace expert skills for which workers currently receive premium wages? How will AI affect demand for labor in specific sectors, at individual companies, and between countries? What new jobs or sectors will be created? What will happen to public finances? Many of these issues come down to what tasks will be replaced by algorithms and who will keep a job (or get a better job).
With only early evidence and hints about AI’s effects on the overall labor markets, a set of hard questions arise for policy makers: are there ways to put AI development onto a path that would create more good jobs for more people?
This is a matter for serious debate across the US and other industrial nations, such as Western Europe. It may be even more challenging for middle- and lower-income countries which have less scientific basis for innovation and fewer fiscal resources to use to buffer whatever labor market transition lies ahead. Further, poorer countries may find themselves hard pressed if they lose good jobs at the same time as their corporate sectors come under intense pressure — for example, if the US and China maintain their leads in AI development and deployment.
Gary has spoken and written extensively about how AI, including its impact on finance, investing, and work: “It really took Henry Ford about 25 years to figure out: ‘Let’s re-envision the factory floor.’ ... I think there’s a lot of re-envisioning that’s going to need to go on.”
In Power and Progress: Our 1000-Year Struggle Over Technology and Prosperity, Simon and Daron Acemoglu wrote about the history of technology and why it does not always deliver benefits for everyone. While AI could create a bonanza that supports shared prosperity, there are reasons to be concerned that this is not the path we are on. To address these concerns, Simon, Daron Acemoglu and David Autor have proposed what they call “Pro-Worker AI.”
The British government is currently running a Pro-Worker AI Adoption prize, “recognising organisations using AI that benefits workers.” Simon is chair of the prize committee.
