History's clock is running. Does your workforce strategy know what time it is?

Sometime in 2019, a researcher at OpenAI prompted an early language model and saw something that had not been seen before. The prompt was straightforward. The response was not.

Within three years, the technology was in the hands of 100 million users. Within five, it was cited by US employers as the primary reason for 40% of all announced job cuts in a single month — up from 7% just four months earlier, according to HRD's reporting on May 2026 Challenger, Gray & Christmas data.

That speed feels unprecedented. Historically, it’s not. What is unusual is how many senior HR leaders are making strategy as though it is.

The cycle that always runs

Every general-purpose technology — the kind that doesn't just improve one industry but restructures the entire economy — follows a recognizable sequence. The stages vary in duration but not in order.

Invention. A working technology exists, usually in crude form, often solving a narrow problem. Thomas Newcomen's steam pump, patented in 1712, was designed to remove water from mines. James Watt's improved engine arrived in 1765. Neither man imagined a railway network or a factory system. The technology existed decades before its full implications were visible (World History Encyclopedia, January 2026).

Demonstration and early adoption. Pioneers deploy the technology. Results are uneven. Costs are high. The majority of the economy proceeds largely unchanged. Watt's commercial engines were manufacturing workhorses by the 1780s, but for the first half of the 19th century most British industries still relied primarily on water, wind, horse, and human power.

Public reaction — fear, resistance, and political response. As the technology scales, the people it threatens can see it clearly; the jobs it will create cannot yet be seen at all. This is the stage of maximum fear and minimum accuracy. The Luddite movement peaked between 1811 and 1816 — burning mills and smashing looms across the English Midlands. The British government deployed more troops to contain them than Wellington had taken to the Iberian Peninsula. Parliament made loom-breaking a capital offence. The weavers were not wrong that their specific livelihoods were ending. They were wrong that the economy was ending with them.

Slow, uneven adoption. This stage is longer than any executive presentation acknowledges. Electricity was demonstrated by the 1880s, but U.S. factories did not meaningfully run on it until around 1914–1917 and did not fully utilise it until the 1920s — a lag of three to four decades from working invention to mainstream industrial deployment. Economists now call this the J-curve: productivity dips first, as organisations absorb learning costs, rewire infrastructure, and restructure around the new tool, before eventually rising substantially. Robert Solow's famous 1987 observation — "you can see the computer age everywhere but in the productivity statistics" — captured the same paradox for computing (Brynjolfsson, "The Productivity Paradox of IT," 1993). Computers had been in offices for a decade; the productivity gains did not materialise clearly until the late 1990s.

Displacement — visible, concentrated, painful. When adoption reaches critical mass, the jobs it automates disappear with striking speed. Between 1910 and 1950, the automobile displaced over half a million jobs in horse-drawn transport, carriage manufacture, harness-making, and related trades (Manyika et al., cited in arXiv:2401.09718, 2024). These losses were geographically concentrated and socially devastating for the communities that depended on them. They were also well-documented, loudly protested, and politically visible in ways that the subsequent creation phase was not.

Creation — diffuse, delayed, harder to see. The automobile ultimately generated an estimated 7.5 million jobs across manufacturing, supply chains, roadside commerce, auto dealerships, repair, fuel, long-haul trucking, and the interstate highway system — roughly ten times the jobs it displaced, amounting to approximately 10% of US employment at the time (Manyika et al., cited in arXiv:2401.09718, 2024). But those jobs arrived slowly, across new sectors, in new geographies, requiring skills that the displaced workers often did not have and could not easily acquire.

The net outcome was positive. The transitional experience was not.

Where AI sits in the cycle, and why timing matters for HR

Understanding the cycle matters because the correct organizational response differs sharply depending on which stage you are in. Treating a displacement phase as though it is still an adoption phase leads to workforce plans that are structurally wrong — not directionally wrong but wrongly calibrated in time.

The honest assessment of AI's current position: we are somewhere between the end of slow adoption and the beginning of real displacement. The Solow paradox may already be resolving.

The World Economic Forum's Future of Jobs Report 2025, drawing on surveys of more than 1,000 employers representing 14 million workers, found that 86% of employers expect AI to transform their business within five years, and that 40% of core job skills will change in that period (World Economic Forum, Future of Jobs Report 2025, January 2025). Sixteen leading economists surveyed by the Wall Street Journal on June 9, 2026 — including Daron Acemoglu of MIT, winner of the 2024 Nobel in Economic Sciences, and former senior White House advisers — unanimously agreed that AI will meaningfully boost labour productivity. The J-curve, in other words, is approaching its inflection (Te-Ping Chen and Justin Lahart, The Wall Street Journal, June 9, 2026).

The displacement data supports this. As HRD has reported, entry-level developers in California saw employment fall nearly 20% from 2024 levels even as headcount for more experienced developers grew. AI is compressing the junior end of the workforce first — consistently with historical patterns where displacement starts at the most automatable tasks and works upward. Andy Challenger of Challenger, Gray & Christmas offered a measured read: "AI isn't yet the jobpocalypse some predicted. But the labor market is being reshaped by technology in real time."

What the historical cycle tells us is that this is early in the displacement phase, not late. The creation phase that follows it is still largely invisible. The jobs that AI will ultimately generate — in oversight, in novel risk, in human-centred roles that only become commercially valuable once everything else is automated — do not yet have job titles, salary bands, or career pathways. They will. The question for HR leaders is whether their organisations will be positioned to fill them or scrambling to find people who can.

The pipeline problem is structural, not cyclical

The most consequential implication of the cycle for HR is not headline redundancy numbers. It is what happens to the talent pipeline when automation removes its foundations.

HRD's coverage of research by Deel identified the risk plainly: organizations may gain short-term efficiency while eroding the very pipeline that produces tomorrow's managers, specialists and executives. David Autor of MIT, in the WSJ survey, named the roles most immediately at risk: workers in "routine information-processing roles — adjusting insurance claims, translating documents, writing standard ad copy — face genuine displacement risk." Every industry has its own version of these roles. They are also, in most organizations, the roles through which people learn the business, develop domain judgment, and become the senior professionals the organisation will depend on in ten years.

The historical cycle makes this risk more acute, not less. During the electricity transition, factories that moved fastest to electrify often did so by replacing the workers who operated steam-powered equipment — the same workers who carried the institutional knowledge of the production process. A generation later, those same factories faced skills shortages in precisely the areas they had hollowed out. As HRD has reported, the CHROs driving genuine transformation in 2026 share a conviction that the old definition of human resources no longer fits the work ahead — and that the precondition for anything strategic is clean processes, strong workforce planning, and deliberate skills-based thinking.

Research on workforce agility published by HRD puts numbers on what good practice looks like: firms operating a connected system of workforce intelligence and talent activation are eleven times more likely to describe their workforce as highly adaptable to change, and achieve up to eight times stronger financial performance than lower-performing peers. The differentiator is not the AI tools deployed — it is whether the organisation treats workforce intelligence, development, and deployment as a connected system rather than isolated functions.

The skills that survive every cycle

Across every technology wave, the same pattern holds in the skills that remain valuable long after the displacement phase. They are not technical skills. Technical skills specific to any given technology become commodities as that technology matures — the railway engineer, the telephone operator, the COBOL programmer. What endures is the capacity to operate at the boundary between the technology and the human system it serves: to exercise judgment the machine cannot replicate, to manage relationships the machine cannot hold, and to navigate the political realities the machine cannot see.

The WSJ economists were consistent on this point. Acemoglu argued that interpersonal and social skills will become more important as AI takes on codifiable knowledge work. Harvard Business School's Rafaella Sadun identified the ability to build coalitions and manage resistance as a premium skill precisely because AI creates internal winners and losers as reliably as it creates external ones. Ajay Agrawal of the University of Toronto described what the most productive workers will look like: not those who code or write best, but those with "the judgment to know which AI output to trust and how to integrate it into a broader system."

HRD's reporting on CHROs preparing for agentic AI finds three in four HR leaders expect AI agents to increase demand for soft skills — collaboration, adaptability, the capacity to work alongside digital labour rather than compete with it. Most plan to move employees into relationship-building roles as automation absorbs the transactional layer of their work.

Phillips Pham, chief people officer at Mainetti Vietnam, and Kate Major at ACC New Zealand told HRD what the evidence from every previous technology wave suggests: keeping the human outcome at the centre of every technology decision is not a soft preference — it is what determines whether implementation succeeds or fails. ATB Financial deployed AI tools across 5,500 staff without setting clear usage expectations. The tools were available and heavily promoted. Adoption was patchy; accountability was absent. Availability, as every technology transition proves eventually, is not strategy.

The honest time horizon

Michael Strain of the American Enterprise Institute, one of the WSJ's sixteen economists, offered the calibration that HR leaders most need to hold: "It is true that, over the medium run, new technologies make (basically) everyone in society better off. But the Industrial Revolution left average real wages stagnating and the quality of non-wage amenities declining for four decades."

Four decades. That figure is not a counsel of despair — it is a planning parameter. The organisations that navigated the electricity transition well were not those that deployed the technology fastest. They were those that understood the full cycle, invested in building the workforce capable of operating in the world the technology would create, and resisted the temptation to mistake early displacement for the whole story.

As LinkedIn's CEO has argued in HRD, five-year career plans are becoming unreliable as AI tools redeploy tasks faster than planning cycles can accommodate. But that is an argument for shortening the planning horizon and building scenarios, not for abandoning the long view entirely. The long view is exactly what history offers — and what it consistently shows is that the organisations and individuals who understand where they are in the cycle, rather than reacting to each stage as though it were unprecedented, make consistently better decisions.

Rebecca Henderson of Harvard Business School put the uncertainty honestly in the WSJ survey: she expects this disruption to be significant enough that affected workers will "get very, very angry and change the politics," and she does not pretend the historical analogies are a perfect fit. "I don't think we've ever really seen anything moving with this scale and speed before," she said. "It's going to be a wild ride."

She is probably right about the speed. The cycle, however, is the same one. History's clock is running. The question for every senior HR leader is a simple one: do you know what stage you are in, and is your workforce strategy built for the one that comes next?