Automation versus jobs: a new model

How will automation impact tomorrow’s jobs?  I and others have been exploring this increasingly urgent question, with an eye towards the intersection of education and society.

The Foresight Alliance team just added to the discussion with a new publication, combining simulation modeling and scenario development.  I recommend the whole article, but will also pull out some key points here.

(NB: I’m only going to discuss automation’s impact on education through society.  No time now for writing about automating education itself.)

One is the pair of trends they selected for scenario creation, which point to a wider range of possibilities than we normally consider.  These are degrees of technological innovation and adoption pace.  For the former, consider how much of AI and robotics is still tentative or flawed, and could well take a long time to become operational.  For the latter, every educator knows organizations don’t necessarily pounce on new technology.

Using the classic two-by-two matrix model, the Alliance offers this set of possible futures:

scenarios matrix: technological innovation and automation adoption

We can readily imagine different social outcomes for each of these.  SAG and ATO generate substantial un- and underemployment.  RAH is closest to our present day, while HMC might involve the deepest types of change, as jobs are transformed along with people’s sense of their working life (and selfhood, I think).

Note the huge range of potential job loss, from under 2% to over 34%:

automation and job loss to 2035_Foresight Alliance

Note, too, the divergent timelines.  The report sees 9 years out at the closer horizon, while pushing forward a further decade – 19 years ahead – to see further impact.  This isn’t happening overnight.

I’m also struck by the wildly skewed impact on different economic strata.  Automation is classist, no matter which scenario bears fruit.  And widespread: “The jobs threatened by automation are among the most ubiquitous jobs in the US workforce.”

Automation and unemployment, different levels for scenarios

This suggests income and wealth inequality are likely to grow, at least when considering the impact of automation.

These findings also point to post-secondary education participating in that widening class division, as university degrees should remain a ticket to higher income – and avoiding the robots.  “With only a few exceptions (accounting clerks, paralegals, etc.) college educated business professionals face minimal risks from job automation across all scenarios.”  What does this mean for educators and campus leaders?  Should we accept this division and our role in it, or organize to shape a different, more egalitarian outcome?

At a meta-level, let me commend the Alliance team for some nice data dashboard work.  Here’s a too-cramped screenshot:

Foresight Alliance dashboard


Liked it? Take a second to support Bryan Alexander on Patreon!
Become a patron at Patreon!
This entry was posted in scenarios, Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *