The Disruption Narrative Hides Who Gets Hurt

"Disruption" is the most successful euphemism in modern business. It takes a process that destroys livelihoods, hollows out communities, and concentrates wealth, and recasts it as the natural metabolism of progress. Painful, sure — but ultimately beneficial. The old gives way to the new. It all works out.

Except it doesn't all work out. Not for everyone. And "it all works out" is doing enormous political work by obscuring who, specifically, it works out for. The good news: once you see the narrative's blind spots, better frameworks become visible — ones that take seriously both innovation's genuine benefits and its uneven costs.

The Origin Story

Clayton Christensen coined "disruptive innovation" in The Innovator's Dilemma (1997) to describe a specific phenomenon: how incumbent companies fail when smaller competitors introduce cheaper, initially inferior products that improve over time. It was a theory about corporate strategy — why big companies lose to small ones.

Silicon Valley took this narrow observation and turned it into a worldview. Disruption became inherently good. Industries should be disrupted. Resistance was futile, adaptation the only option. The word migrated from business school case study to civilizational imperative in under a decade.

Christensen himself objected. In a 2014 interview with The New Yorker, he argued the concept was being applied far too broadly, to situations it was never meant to describe. But by then the narrative had escaped its author. "Disruption" was no longer an analytical framework. It was an ideology — one that happened to serve the interests of those doing the disrupting.

Where the Narrative Breaks Down

The disruption story has five structural problems that become visible the moment you ask who benefits and who pays.

1. The Distribution Problem

Disruption may create aggregate wealth. But aggregate wealth is a misleading metric when gains and losses are distributed across entirely different populations.

When ride-hailing platforms disrupted the taxi industry, investors and founders captured billions. Taxi drivers — many of whom had purchased expensive medallions as their primary asset — watched their livelihoods evaporate. Some new jobs were created, yes, but they were lower-paid, benefits-free, algorithmically managed gig work. The total economic pie may have grown. The people holding the old slices got crumbs.

Daron Acemoglu's work on automation and inequality makes this precise. He distinguishes between technologies that complement workers — making them more productive and valuable — and "so-so automation" that replaces workers without meaningful productivity gains (Acemoglu, 2021). Much of what passes for disruption is so-so automation: it shifts surplus from labor to capital without creating new value. That's not creative destruction. That's redistribution upward. (For a deeper treatment of this distinction, see The Automation Debate and the analysis in Power and Progress.)

The complementary path is not hypothetical. Technologies that augment rather than replace workers — like AI Coding Assistants that help programmers handle routine tasks while freeing them for design work — demonstrate that disruption and worker benefit are not mutually exclusive. The question is which path gets funded and deployed.

2. The Timeline Problem

"New jobs will be created" is the standard reassurance. And over decades, it has often been true. But individuals don't live on the timescale of economic history. They live on the timescale of rent payments.

Tell a 50-year-old journalist that journalism will evolve into exciting new forms over the next twenty years. It doesn't help them pay rent this month. Tell a truck driver that autonomous vehicles will eventually create logistics management roles. They need to know how to feed their family during the transition.

The timeline problem is not a footnote to disruption. It is the central moral question. As Aaron Benanav argues in Automation and the Future of Work, the question is not whether the economy will eventually adjust, but what happens to people during the adjustment — and whether "adjustment" is even the right frame when the losses are permanent for many who experience them (Benanav, 2020). A 55-year-old factory worker does not get retrained into a machine learning engineer. They get early retirement, disability, or poverty.

Economists call these "transition costs," a phrase that makes human suffering sound like a line item. The disruption narrative treats transition costs as temporary friction. For the people living them, they are the whole story. The Shape of AI Displacement maps how these costs distribute across industries and demographics.

3. The Agency Problem

The disruption narrative presents technological change as a force of nature — like weather, or gravity. "AI is coming whether we like it or not." "You can't stop progress." "Adapt or die."

This framing is politically convenient and factually wrong. Technology deployment is a series of choices made by specific people and institutions. Which technologies get funded. Where they get deployed. How fast. With what safeguards. These are decisions, not natural phenomena.

Langdon Winner's foundational question — "Do artifacts have politics?" — applies directly (Winner, 1980). The decision to automate customer service is not an act of nature. It's a choice by executives to reduce labor costs, made possible by specific AI capabilities, enabled by specific investment decisions, permitted by specific regulatory environments. Every link in that chain involves human agency.

When disruption is presented as inevitable, the political space for shaping technology collapses. If AI-driven layoffs are simply what happens, there is nothing to govern, negotiate, or resist. The inevitability frame is not a description of reality. It is a strategy for preempting opposition. The Autonomy Illusion explores how this same pattern operates at the product level — attributing agency to systems rather than their makers.

4. The Historical Cherry-Picking

Disruption advocates love the automobile. It "disrupted" horse-drawn carriages and created vastly more jobs — in manufacturing, road construction, fuel production, suburban development. Progress in action.

But the full ledger of the automobile is considerably darker. Robert Moses's highways destroyed Black and working-class neighborhoods. Suburban sprawl produced car dependency, social isolation, and one of the largest contributors to climate change. The automobile didn't just replace the horse. It restructured American society around a technology that made oil companies rich, hollowed out cities, and created environmental crisis we are still trying to solve.

The disruption narrative cherry-picks the job numbers and ignores the externalities. It counts the auto workers hired and skips the communities demolished for highway on-ramps. It tallies the new industries created and omits the century of carbon emissions. A full accounting of "creative destruction" requires counting what was actually destroyed — not just the jobs, but the social fabric, the environment, the communities, the ways of life.

5. The Power Asymmetry

There is a reliable pattern in disruption discourse: the people calling for it are never the people experiencing it.

Tech executives champion creative destruction from positions of total personal security. Venture capitalists celebrate "move fast and break things" while their diversified portfolios protect them from any individual breakage. Consultants prescribe "transformation" to industries they will never work in.

As Kate Crawford documents in Atlas of AI, the material costs of technological systems — the mining, the manufacturing, the precarious labor — are consistently borne by those with the least power to resist (Crawford, 2021). The disruption narrative extends this pattern. Those with capital to invest capture the upside. Those whose labor is being automated absorb the downside. And the narrative itself — this is how progress works — insulates the winners from any obligation to the losers.

The AI Disruption Narrative

All five problems are amplified in the current AI moment.

"AI will create more jobs than it destroys." This claim remains unproven at the current scale and speed of deployment. Previous technology transitions unfolded over decades. Large language models went from research curiosity to workplace deployment in roughly two years. The historical analogies that disruption advocates rely on may not apply cleanly to a technology that can automate cognitive work across many white-collar professions simultaneously. This does not mean net job creation is impossible — but it means the optimistic case needs to be argued with evidence, not assumed from historical precedent. The Productivity Paradox explores why the expected gains often fail to materialize as promised.

"Workers just need to reskill." This formulation reveals the power asymmetry with unusual clarity. The burden of adaptation falls entirely on the displaced. Learn to code. Learn to prompt. Learn to manage AI. Meanwhile, the companies deploying AI and capturing the productivity gains face no corresponding obligation — no requirement to fund retraining, no mandate to share profits, no responsibility for the communities they hollow out. Reskilling can be part of the answer, but only when paired with institutional support — Collective Bargaining and AI and proposals like UBI and AI represent efforts to spread the burden more equitably.

"Resistance is Luddism." This is meant as a dismissal, but the historical Luddites deserve better. They were not anti-technology. They were skilled textile workers who objected to the use of machinery to circumvent labor standards and depress wages. Their argument was not "machines are bad" but "machines should not be used to make workers powerless." That distinction matters. Calling critics of unregulated AI deployment "Luddites" is more accurate than the name-callers intend — the Luddites were right that technology deployed without regard for workers produces immiseration, and two centuries of labor history proved them so.

The Redistribution Angle

The disruption narrative serves capital by naturalizing inequality. When disruption is a force of nature, the resulting distribution of gains and losses is also natural — and therefore not a political question. Redistribution becomes "interference" with beneficial processes. Regulation becomes "stifling innovation." Labor protections become "rigidity."

This is the narrative's deepest function. It does not merely describe technological change. It forecloses the political response to technological change.

When we accept that disruption is inevitable, we surrender the space for asking who decides what gets disrupted. When we accept that transition costs are temporary, we abandon the people living through permanent displacement. When we accept that the market will sort it out, we hand the sorting to those who already hold the most power.

The better question is never "how do we adapt to disruption?" It is: Who decides what gets disrupted? Who captures the value? Who bears the cost? And what institutional arrangements could distribute those costs and benefits more fairly?

The Better Frame

Technology is a choice, not a force of nature. Deployment can be shaped. Benefits and costs can be distributed deliberately rather than left to accumulate where power already concentrates.

This is not utopian. It is what every functioning society already does with other disruptive forces. We don't let pharmaceutical companies sell untested drugs because "innovation." We don't let factories dump waste in rivers because "progress." We regulate, tax, redistribute, and require that those who profit from disruption bear some responsibility for its consequences.

The same framework applies to AI. Transition funds for displaced workers. Profit-sharing requirements for companies that automate. Regulatory review before deployment in high-stakes domains. Democratic input into which disruptions we pursue and which we constrain.

Disruption without redistribution is extraction. It takes value from the many and concentrates it in the few, then calls the result progress. But seeing through the narrative is the first step toward building something better — technology deployment that creates genuine, broadly shared prosperity rather than aggregate wealth statistics that mask concentrated suffering. Models for this exist: AI Cooperatives, Democratic AI Governance, and the broader Public AI movement all represent attempts to ensure that technological change serves the many, not just the few.

Related

• The Automation Debate — The broader evidence on whether AI creates or destroys jobs
• The Autonomy Illusion — How attributing agency to AI obscures human decisions
• The Shape of AI Displacement — Who actually gets displaced and where
• Power and Progress — Acemoglu and Johnson's historical case for directing technology toward shared prosperity
• White-Collar Displacement — The specific pattern of AI disruption in knowledge work
• Big Tech's AI Layoffs — The disruption narrative in practice
• Collective Bargaining and AI — One path toward worker voice in technological transitions
• GPTs Are GPTs — Research on which occupations face the most exposure

Sources

• Christensen, Clayton. The Innovator's Dilemma. Harvard Business Review Press, 1997. HBR Press
• Acemoglu, Daron. "Harms of AI." Cambridge Journal of Economics, 2021. Cambridge
• Benanav, Aaron. Automation and the Future of Work. Verso, 2020. Verso
• Winner, Langdon. "Do Artifacts Have Politics?" Daedalus 109, 1980. JSTOR
• Crawford, Kate. Atlas of AI. Yale University Press, 2021. Yale Press
• Lepore, Jill. "The Disruption Machine." The New Yorker, 2014. The New Yorker