This article is based on a post originally sourced from opensource.com, written from the perspective of a working software engineer reflecting on the broader economic and social implications of automation.
The Scale of Automation-Driven Job Loss
A study titled Myth and the Reality of Manufacturing in America, conducted by the Center for Business and Economic Research at Ball State University, concluded that nearly 88% of manufacturing job losses in recent years can be attributed to productivity growth—not competition from foreign workers. Translate “productivity growth” into “automation” and the implication becomes clear: it is machines, not people from other places, that are displacing workers.
This is not a new phenomenon. Mechanisation has been displacing human labour since before the Industrial Revolution. What is new is the rate of displacement. Technology now compresses into a few years what once took generations.
Sectors Already Feeling the Pressure
Projections from the U.S. Bureau of Labor Statistics (figures cited in the original 2017 source, looking ahead to 2024) pointed to significant workforce reductions across several categories:
- Electronics installation and repair in vehicles: ~50% fewer workers needed as diagnostics and components become more reliable and self-correcting.
- Telephone operators: ~42% reduction as natural-language call-routing technology improves.
- Postal clerks: ~26% fewer, as self-service kiosks handle routine transactions.
- Parking enforcement officers: ~20% reduction as smart meters and digital monitoring take over ticketing.
These are not edge cases. They represent entire occupational categories that are shrinking faster than new roles appear to absorb displaced workers.
Self-Driving Vehicles and the Ripple Effect
Consider long-haul trucking. As autonomous vehicle technology matures, self-driving trucks on highways shift from experimental to economically inevitable. The number of professional truck drivers potentially displaced in that transition alone would represent one of the largest single-sector workforce disruptions in modern history.
The ride-share scenario compounds this further. If a company like Uber or Lyft replaces its human drivers with fully autonomous fleets, the economics change dramatically for consumers. A flat-fee, unlimited-mileage monthly subscription could easily undercut the combined cost of vehicle lease payments, insurance, fuel, and maintenance for private car ownership. When that threshold is crossed at scale, privately owned vehicles parked on streets become an anachronism—and the downstream effect is fewer cars manufactured, which means fewer workers in automotive factories and supply chains.
Automation Inside Software Itself
Automation is not only displacing blue-collar roles. Consider how logging infrastructure has evolved within software engineering:
- Stage 1: A person manually rotated log files on disk when they grew too large.
- Stage 2: A script automated file rotation based on size thresholds—one manual task eliminated.
- Stage 3: Cloud-based log aggregation services (e.g., centralised logging platforms) absorbed the need to manage archive hardware entirely. No more server room visits, no capacity planning scripts, no manual intervention.
Each stage removed human effort that had commercial value. Multiply this pattern across CI/CD pipelines, infrastructure provisioning, monitoring and alerting, and security scanning, and it becomes apparent that automation is consuming significant portions of what used to be skilled engineering work—not eliminating it entirely, but reducing the headcount required to operate the same infrastructure.
“Just Retrain Workers” Is Not Enough
The most common policy reflex when automation displaces workers is to call for retraining programmes. For knowledge workers in technology, this argument has an obvious flaw: continuous learning is already a baseline survival requirement, not a one-time intervention. The skills needed to write an infrastructure-as-code deployment against a cloud provider are non-trivial, and they are often obsolete by the time a worker finishes learning them. Retraining assumes a stable target skill set. In fast-moving technology domains, no such target exists.
For workers in manual or semi-skilled roles, the challenge is even steeper. The cognitive and educational prerequisites for replacing an eliminated role with a technology role are significant—and expecting a workforce of millions to make that transition quickly is not a realistic plan.
Economic Responses: Universal Basic Income
One structural response to near-total automation is Universal Basic Income (UBI)—a programme in which citizens receive a baseline income regardless of employment status. The concept is not new; economists including Milton Friedman advocated versions of it in the 1960s under the label of a “negative income tax.” Finland launched a limited pilot programme exploring UBI principles, and other countries have conducted experiments in this space.
The logic is straightforward: if the economy structurally cannot provide enough employment for the population, tying individual survival to employment becomes unsustainable. UBI attempts to decouple basic economic security from labour market participation.
Human Dignity in a Post-Labour Economy
Perhaps the least-discussed dimension of this transition is psychological. Philippe Kahn, founder of Borland, Starfish Software, LightSurf Technologies, and Fullpower Technologies, framed the paradox clearly when asked about automation’s social impact:
“I remember being in India and watching a construction site: lots of people, elephants, activity. I asked, ‘Why don’t you all use bulldozers and tractors and prefabricated concrete parts?’ The answer was enlightening: ‘How would this community make a living with dignity?’ That’s the whole paradox that we are dealing with: productivity versus dignity for communities that could be unemployed for generations.”
— Philippe Kahn
The distinction between work and labour matters here. Labour is what people do to survive economically. Work, in the broader sense, is activity that gives life meaning—creative output, caregiving, community contribution, artistic expression, scientific inquiry. In an economy where machines handle most labour, humans do not automatically lose the need for meaningful work. They lose the income that labour provided.
Where Human Activity Can Still Excel
A highly automated economy will, by definition, generate significant productivity and wealth. The critical question is how that wealth is distributed. Activities that carry high human value but low commercial value—rare-disease research, basic science, space exploration, arts and culture, education, social care—are precisely the domains where human contribution is most defensible and most needed. These are also the domains that market mechanisms consistently underfund.
Funding these activities at scale, through whatever mechanism society chooses, would allow humans to engage in meaningful work that machines cannot easily replicate or commercially justify. This is not a utopian proposal—it is a recognition that the old relationship between productivity, employment, and income will require deliberate restructuring rather than passive market adaptation.
What DevOps and Engineering Teams Should Take Away
For those of us building the automation infrastructure that is driving these changes, the engineering challenge is inseparable from the social one. Pipelines that eliminate manual deployment steps, monitoring systems that replace human operators, AI-assisted code review that reduces the need for junior reviewers—these are not neutral technical achievements. They have labour market consequences that accumulate.
This does not mean engineers should slow down innovation. It means the profession should engage honestly with the downstream effects of what it builds, advocate for policies that distribute automation’s benefits broadly, and resist the comfortable assumption that retraining alone is a sufficient social response to structural displacement.
The question—what will we do when everything is automated?—is no longer speculative. It is an engineering and policy problem that is already in progress.