Skip to content

Can computer technology enable work without jobs?

A newsletter from the IEEE Society on Social Implications of Technology put me on to a collection of essays Perspectives on Digital Humanism (2021), one of which has Daniel Samaan pontificating on Work Without Jobs. Samaan begins with the claim that “a society might run out of jobs but it can never run out of work”.

He gives three reasons for thinking that computer technology might destroy the “job system”, the term he adopts from Frithjof Bergmann for the present system of assigning individuals to their day-to-day work: artificial intelligence is a “general purpose technology” that affects the entire economy and not just one particular industry; “big data” eliminates the need for mediators (such as middle management) who currently connect decision-makers to individual workers; and automation of tasks. He furthermore thinks that these radical changes mean “it is very well possible that we will see more than a mere reshuffling of a few lost jobs in some sectors and emerging new jobs other ones” and woe betide policy-makers who attempt to cling to the job system (as, he notes, pretty much all current policy-makers and aspirants to the role do).

Even if one accepts Samaan’s claims for general-purpose technologies, big data, and automation, it’s not obvious to me that these will of necessity destroy the job system: after all, electricity is a general-purpose technology, “disintermediation” was a buzzword in the 1990s, and any number of activities have been automated since the Industrial Revolution. But I think the more important point is that these phenomena could end the job system, if we so chose. Samaan thinks we should, because “almost everybody hates it and we all know that it is at most partially fair” (others disagree).

Plenty has been said and written about how automation might destroy jobs—but also about why it won’t. If one takes “work” to mean the activities that people carry out in “jobs”, it’s far from obvious why eliminating the latter shouldn’t also eliminate the former, or why failing to eliminate the former shouldn’t also mean failing to eliminate the latter. But if one takes a broader view of “work”, allowing for developing one’s skills, carrying out projects for the pleasure of doing them, and so on, it becomes clear why we might never run out of work: a society whose material needs have been met might conceivably run out of jobs in satisfying those material needs, but no individual could ever run out of skills to learn or personal projects to carry out, unless that individual had only very meagre ambitions.

On the face of it, destroying the job system requires either a different system to be put in its place, or the need for it to disappear altogether. I take Samaan’s discussion of big data to be searching for the former: one can easily imagine an all-knowing big data system farming out tasks to available workers a la Amazon Mechanical Turk or TaskRabbit, for example. But the latter examples aren’t exactly notorious for the wealth and comfort they generate for their participants, and the system in charge of it sounds rather like the kind of central planner that would outrage economists and politicians who fancy themselves to be liberal democrats. So maybe that’s not the right system either.

Leave a Reply