When the question of “automation caused job losses” comes up, most people think of industry jobs where production-line or humanoid robots take over jobs whole or at most talk about virtual assistants or other entry level knowledge-worker positions getting replaced by Artificial Intelligence or how Neural Networks can compose news articles from a set of facts; the common theme here is bots taking over jobs whole; while ‘bots taking over jobs at the production-line is old news and vending machines that replace pizza restaurants from the cashier to chef is no longer something to speculate, I think a much more deadly and silent job killer is often overlooked: tools that makes workers more efficient.
As a general rule, efficiencies that reduces work-hours by any given percentage, it also reduces the overall number of position by the same percentage.
To illustrate the point, here is an example from real life industry experience. Going around an upmarket department store fishing for products that users have order online is not the kind of job one expects to get automated anytime soon, as department store shelves design is driven by human psychology around purchasing rather than accessibility for robots, yet one of the projects that I worked on effectively cut the number of said job by more than half, how you ask? the project in question provided a much more efficient means of bundling and finding products around the store that reduced the time it takes for a batch of products to be found by over 60% percent, this directly translated to millions of saving by the company as a result of reducing the number of casual and seasonal workers required.
Another example of this is perhaps somewhat ironically the software development industry, the very workhorse of automation and tools of efficiency is eating itself away. The outward image may look very different with the software development becoming an integral part of more and more industries, yet, the number of developers required for a project of given complexity is reducing at unimaginable rate. People may get excited at new cross-cutting titles like DevOps and Cloud Engineer, Test Engineer but this is effectively new roles that replaces multiple jobs at once, DBA, sysadmin, manual testers, and many of such jobs are going the way of the dodo.
A slightly more different example of death by a thousand efficient is the Flight Engineer, despite the level of automation in civil aviation, no one expects the pilots to be replaced by AI or any kind of automation anytime soon, yet in the process of creating tools and systems that helps pilots fly more safely and efficiently, the role of flight engineer is completely replaced by a thousand little system in majority of modern airplanes.
The takeaway here is that looking at automation risk as something that replaces whole jobs at once is missing the forest for the trees.
Fully agree. I like to compare "AI" to a power tool. Let's say screw gun. You still need to know what to screw in, where and why - but you'll be doing it faster. Suddenly, instead of needing three people to drywall an apartment, you can get by with two.
And it's not only AI. I'm a devops guy. Ten years ago I was managing ~20 servers. Now it's ~200. And the quality and overall value of the work has improved as well. IaC, kubernetes, CI/CD, autoscaling, gitops, what-have-you. It all adds up.