The Industrial Revolution Had Winners and Losers. AI Will Too. The Question Is Who Decides.

History doesn’t repeat itself. But it does identify who pays the price.

In 1780, British textile workers began smashing spinning machines.

They weren’t wrong to be afraid. The spinning jenny and water frame were eliminating the hand-spinning work that had supported rural families for generations. A technology that would eventually raise living standards for millions was, in the short term, destroying the livelihoods of the people who built it.

The Luddites lost. The machines kept running. And two centuries later, we cite the Industrial Revolution as evidence that technological progress ultimately creates more jobs than it destroys.

That’s true. It’s also incomplete. And the incomplete part is relevant right now.

What the Data Actually Shows

The broad narrative — technology creates disruption, then abundance — is supported by the evidence. Overall employment rose through industrialization. Real wages eventually increased. Life expectancy improved.

But aggregates obscure distribution. And the distribution is where things get complicated.

Researchers analyzing British census data have documented approximately 140,000 bootmakers who experienced technological unemployment as mechanization swept through the industry between 1851 and 1911. Between 1910 and 1950, the automobile displaced over 500,000 jobs in the United States — carriage makers, harness manufacturers, stable hands, and the broader ecosystem of services built around horse-drawn transport.

The winners of the Industrial Revolution and the people who paid its transitional costs were rarely the same people. Factory owners profited. Skilled craftsmen either retrained or watched their earnings collapse. Children went to work in factories because their parents could no longer earn a living through craft.

Progress happened. But it happened unevenly, across decades, and the people living through the transition didn’t get to skip to the good part.

The Patterns That Are Repeating

According to the IMF’s January 2026 assessment, nearly 40% of global jobs are exposed to AI-driven change. The research identifies a consistent pattern: 2023–2025 brought task automation, hiring freezes, and role compression. 2026–2028 is projected to see career transitions spike and displacement peak.

What makes the current moment feel specifically familiar is where the displacement is concentrated.

Goldman Sachs analysis shows that workers aged 22–30 are experiencing AI-driven displacement at nearly three times the rate of workers aged 40–55. Employment for software developers aged 22–25 has fallen nearly 20% since 2024 — not overall developer employment, which is holding steady. The collapse is specifically at the first rung of entry into the profession.

This is a precise historical echo. During the Industrial Revolution, it wasn’t the master craftsmen who lost their footing first. It was the apprentices — the people who were supposed to learn the craft by working alongside someone who already knew it. When mechanization eliminated the entry-level work, the pathway to mastery disappeared with it.

The aggregate numbers improved. The individual transitions were not always successful.

The Fundamental Difference This Time

The Industrial Revolution unfolded over roughly 80 to 100 years. Child labor restrictions in Britain didn’t arrive until the 1830s and 1840s. The first trade unions were legalized in 1871. The 8-hour workday became standard in most of the industrialized world in the early 20th century.

Society had time — painful, contested, often brutal time — to adapt. Laws got written. Norms got established.

AI is compressing this timeline dramatically. The technology that is restructuring labor markets emerged at scale around 2022. The regulatory responses are already years behind the deployment. Only 19% of AI users at work are classified as “Frontier” users — the most capable and integrated. The gap between those who use AI effectively and those who don’t is already generating measurable wage and career divergence.

Workers, educators, and policymakers have less time to respond than their counterparts in any previous technological transition. The patterns are the same. The pace is not.

What History Suggests About Distribution

The most important lesson from the Industrial Revolution isn’t that progress ultimately prevailed. It’s that the distribution of that progress was not determined by the technology itself.

The 8-hour workday didn’t emerge from efficiency calculations. It emerged from labor organizing and political pressure. Child labor laws didn’t appear because factory owners decided on their own that children shouldn’t be working 12-hour shifts. They appeared because society decided — slowly, through conflict — that this was not an acceptable price for industrial output.

Minimum wage legislation, antitrust enforcement, workers’ compensation, social insurance — none of it was automatic. All of it was the result of deliberate choices about who should bear the costs of technological transition and who should capture the benefits.

The AI revolution will ask the same questions. Who captures the productivity gains? Who bears the transitional costs? What obligations do companies that deploy AI have to workers displaced by it? What does society owe to the generation that happens to be entering the labor market during the transition?

These are not technical questions. They will be resolved — as they were during industrialization — through the accumulated decisions of companies, governments, and workers over years and decades.

Where This Leaves Companies

We build software. We use AI agents in our own development process. We have a direct stake in this technology continuing to advance.

We also think the Industrial Revolution parallel is worth taking seriously, specifically because it doesn’t lead to a comfortable conclusion.

The technology will advance regardless of what any individual company decides. The distribution of its benefits will not be determined automatically. Companies that deploy AI have choices: whether they invest in workforce transition, whether they maintain pathways for junior talent to develop into senior talent, whether they treat transparency and governance as core engineering requirements rather than compliance afterthoughts.

The Industrial Revolution ended well for humanity, by aggregate measures. It was considerably harder for specific people who lived through the transition without adequate support.

We don’t think “it worked out in the long run” is a sufficient framework for thinking about what we build and how we build it.

Wamisoftware has been building software products since 2014. We work with senior engineers and AI agents on projects where the quality of what gets built matters as much as the speed and where thinking about the broader implications of what we’re building is part of how we work.