White-collar jobs, interpersonal jobs, and exposure to A.I. displacement
The Brookings Institute published a new analysis of jobs that are most likely to be disrupted by artificial intelligence in November 2019. They used a new method for modeling exposure to displacement through automation developed by Stanford Ph.d. candidate Michael Webb, which ironically used machine learning and natural language processing (an AI technique), to identify professional occupations that are most associated with AI development. This novel method analyzed keywords in patent filings like "AI" and "automation" and modeled them to match with keywords in job descriptions and gave them a score of most exposure to displacement via automation.
In the report, you can see which occupations are most likely to be displaced and the estimated number of jobs in that field. What's surprising is that this way of looking at it suggests that analytical white-collar jobs have more exposure to displacement. Furthermore, these jobs are currently occupied predominately by white and Asian males living in bigger cities.
In contrast, roles in education, health care support, and personal care, which show more involvement from women, have less exposure to displacement from automation. It also suggests that rural communities have less exposure; however, that is mostly because their existing occupations per capita are less weighted by white-collar jobs.
An important question to ask is are jobs most exposed to displacement because they have a high-proximity to investment in A.I. or are they most likely to be displaced because they are the highest value occupations worth automating?
My take on this is that the intellectual property being developed right now is weighted more towards disrupting white-collar jobs because the firms driven by those occupations are the ones investing in the development of automation directly or they are buying it from the originators of the IP. This might have more to do with the capability to develop such technology and less to do with marginal value created by the automation itself. Another way of looking at it is that the reason why white-collar jobs might have more exposure is that they have a less marginal value associated with human-driven value.
In the short-term, it seems obvious that if higher-paying professions cost businesses more and they consist of repeatable processes that automating them would enable those businesses to offer similar services at a lower cost, thus creating more value to consumers. Long-term, however, society might gain more value out of automating processes involved with things like healthcare because it has a higher impact on the quality of life on average than say automating investment banking processes.
The final question I'm left with is, will AI driven-automation encourage white-collar workers in bigger cities to find more interpersonal opportunities in smaller communities or will they stay in bigger cities and transition to working in fields that build scalable automation that will eventually displace the less exposed occupations? I think part of the answer will have to do with the importance of human to human connection and the empathy required to truly understand what problems they're solving.
What are your takes on this study? What do you see in your community? And how are you personally thinking about your future with respect to these findings?