From Luddites to Journalists
When Technology Comes for the Workers It Once Helped
1. Introduction
In the early nineteenth century, groups of British textile workers attacked and destroyed machines that threatened their livelihoods. Two centuries later, journalists, writers, editors, translators and other professional creators face a very different machine: artificial intelligence. Generative AI can already produce articles, summaries, reports, advertising copy, translations, images and other material in seconds. It does not need sleep, holidays or years of professional training, and its marginal cost is extremely low.
The comparison with the Luddites is tempting, but the real history is more interesting than the familiar story of workers foolishly resisting progress. The Luddites were not simply ignorant enemies of machinery. They were skilled workers defending occupations, incomes and ways of life that were being undermined by new methods of production. The British government responded with troops, executions and transportation. Machine-breaking was made a capital offence, and seventeen men were executed at York in 1813.
There is also a deeper irony. Some textile workers whose livelihoods were eventually threatened by mechanisation had previously benefited enormously from an earlier technological revolution. Similarly, many modern journalists and writers whose work is now exposed to AI built their careers in an industry transformed and expanded by computers, digital publishing, search engines and the internet.
Technology creates opportunities, destroys others, changes the value of skills and shifts economic power. The central question is therefore not whether technological change should occur, but who benefits from it, who pays the price and how society manages the transition.
2. Background
The term “Luddite” is now commonly used to describe someone who dislikes technology. Historically, this is misleading. The Luddites who emerged in Nottinghamshire in 1811 and subsequently in Yorkshire, Lancashire and elsewhere included framework knitters, cloth croppers or shearmen, and some handloom weavers. Their grievances differed between regions. They opposed particular machines and particular methods of production that reduced wages, weakened skilled trades or allowed employers to replace experienced workers with cheaper labour.
The comparison with modern journalism is therefore not simply between people who dislike machines. It is between skilled workers confronting technologies capable of reducing the market value of skills they had spent years developing and which had given them strong bargaining power.
Generative AI has particular significance because writing and related intellectual work were once regarded as relatively difficult to automate. Earlier machines replaced muscle, repetitive movements and routine calculations. AI increasingly operates in areas involving language, analysis and creative production. The International Labour Organization identifies journalists, writers and translators as among the media occupations with relatively high exposure to generative AI. At the same time, it stresses that exposure does not necessarily mean complete replacement: in many occupations AI is more likely to transform jobs and automate particular tasks than eliminate the occupation altogether.
This distinction matters. A journalist does much more than arrange words into sentences. Journalism can involve investigation, interviewing, cultivating sources, verifying evidence, witnessing events, exercising judgement and taking legal and ethical responsibility for publication. AI may automate parts of that process without replacing all of it. Nevertheless, if one journalist using AI can perform work previously requiring several people, employment can still decline even though journalists remain necessary.
3. History
The most revealing part of the Luddite story began before Luddism itself.
During the eighteenth century, spinning yarn was a major bottleneck in textile production. Innovations including James Hargreaves’s spinning jenny, Richard Arkwright’s water frame and Samuel Crompton’s spinning mule greatly increased the productivity of spinning. By the 1790s some forms of mechanised spinning produced many times more yarn per worker than traditional hand spinning.
This caused disruption. Traditional spinners lost work or were drawn into factories. But the enormous increase in cheap yarn created a new problem: somebody had to weave it.
Demand for handloom weavers consequently soared. Economic historian Robert Allen describes the 1790s as a “golden age” for handloom weaving. Factory spinning increased the supply of yarn, which increased demand for weavers and pushed their earnings upwards. Skilled weavers temporarily occupied an enviable economic position precisely because an earlier form of mechanisation had disrupted another group of workers.
Yet success contained the seeds of its own destruction. High weaving wages increased the financial incentive to invent machinery that could replace weavers. Edmund Cartwright had patented a power loom in the 1780s, but early versions were relatively inefficient. As the technology improved, power looms became increasingly competitive. The labour that had become valuable because of mechanised spinning had now become expensive enough to make automation increasingly profitable.
The sequence was therefore remarkable. Mechanisation created cheap yarn. Cheap yarn increased demand for skilled hand weaving. Increased demand raised the rewards from weaving. Those rewards increased the incentive to mechanise weaving. The second technological revolution then destroyed much of the advantage created by the first.
The parallel with journalism is imperfect but striking.
Computers removed enormous amounts of manual work from publishing. Digital word processing replaced typewriters and retyping. Electronic databases accelerated research. Desktop publishing reduced production costs. Email transformed communication. Search engines made immense quantities of information searchable within seconds. The internet removed geographical barriers to distribution and allowed a writer in Perth, London or New York to reach readers almost anywhere.
The web created entirely new occupations and opportunities: online journalism, blogging, digital editing, social-media production, independent newsletters, podcasts and global freelance markets. A single writer could research, produce and distribute material using tools that once required substantial institutional infrastructure.
But digital technology also damaged older jobs. Typesetters, compositors and many production workers had already been displaced by computerisation. Later, the internet diverted readers and advertising revenue from newspapers. Research published by the American Economic Association found that broadband adoption caused substantial shifts from print to online news consumption while newspaper revenue declined sharply.
Thus journalists were never simply beneficiaries of the digital revolution. Nevertheless, computers and the internet enormously increased the productive power of writers who remained in the industry. Search engines, databases, digital archives and instant communications made individual journalists capable of work that previously required far more time and support.
Get Rowland Benjamin’s stories in your inbox
Join Medium for free to get updates from this writer.
AI represents the next stage. The digital information system that helped make modern writers productive has supplied the infrastructure, data and computing environment from which generative AI emerged.
4. The Key Issues
The first issue is what economists call creative destruction: innovation creates new industries and opportunities while destroying existing ones. The benefits and costs are rarely received by the same people.
The hand spinner made redundant by spinning machinery did not necessarily become the prosperous handloom weaver who benefited from abundant yarn. Similarly, a newspaper subeditor who loses work because of AI may not become the owner of a successful AI company. Telling displaced workers that technology ultimately creates prosperity can therefore be economically correct at the level of society while being almost meaningless to the individual whose livelihood disappears.
The second issue is the changing value of skill. A scarce skill commands a premium partly because relatively few people can perform it. Technology can suddenly make that scarcity disappear. Power machinery allowed less-skilled workers to perform tasks previously requiring trained textile workers. Generative AI may allow people with limited writing, translation, research or design skills to produce work that is adequate for many commercial purposes.
This does not mean expertise becomes worthless. It may actually increase the value of exceptional judgement, investigation, originality and specialist knowledge. But the middle of the market can be squeezed. If a business previously paid a professional writer to produce routine material but can now ask a general employee to produce an AI-assisted draft, demand for professional writing declines.
The third issue concerns quantity. Industrial textile machinery did not merely replace workers. It made textiles vastly cheaper. Falling prices allowed ordinary people to buy far more clothing, bedlinen, sails, sacks and other textile products. Expanding consumption helped generate employment elsewhere in the economy.
AI may have a similar effect on information. If translation, analysis, tutoring, software development, illustration and routine writing become dramatically cheaper, society may consume far more of them. New industries and occupations may emerge that are difficult to predict in advance.
But information has an unusual problem: producing more words is not necessarily socially valuable. An economy flooded with cheap clothing gives people more clothing. An internet flooded with almost costless synthetic text may instead create misinformation, repetitive material, advertising, spam and a shortage of the one resource technology cannot manufacture — human attention.
The fourth issue is that the current effects should not be exaggerated. AI is already affecting newsroom practices, but widespread replacement has not yet occurred. In a 2026 Reuters Institute survey of media leaders, 67 per cent said their AI initiatives had not resulted in staff reductions, while 16 per cent reported small reductions and 9 per cent said additional roles had been created.
History therefore warns against two equally simplistic conclusions: that AI will destroy professional writing, or that fears about employment will inevitably prove foolish. Neither is justified.
5. Solutions
The obvious response is adaptation. Workers who learn to use a productivity-enhancing technology may gain an advantage over those who ignore it. A journalist who uses AI for transcription, document searches, preliminary research, data processing or routine summaries may have more time for investigation, interviews and analysis.
In that sense, there is some truth in the argument that workers should try to ride the wave of innovation rather than stand in front of it. Attempts to prohibit useful technology completely can protect particular jobs temporarily while encouraging investment and innovation to move elsewhere.
But adaptation is not a complete social policy. A fifty-year-old skilled worker cannot simply become an AI engineer because economists predict that technology will eventually create different jobs. Technological change produces real transition costs, and those costs deserve to be recognised.
Education and retraining are therefore important, but so are income security, unemployment protection and mechanisms allowing workers to share in productivity gains. If AI enables an organisation to produce twice as much with the same number of employees, society has choices. The benefit can become higher profits, lower prices, higher wages, shorter working hours or some combination of them. Technology itself does not decide how those gains are distributed.
Creative industries also require rules governing copyright, attribution, consent and payment for the material used to train and operate AI systems. The International Labour Organization (ILO), a specialised agency of the United Nations concerned with workers’ rights, employment and labour standards, emphasises compensation, control and consent as central questions for media workers confronting generative AI.
Finally, journalism should concentrate on the areas in which human involvement provides the greatest value: original reporting, investigation, verification, judgement, accountability and trusted relationships with sources and readers. Competing with machines purely on the production of routine words is unlikely to provide a secure long-term strategy.
6. Summary and Conclusion
The Luddites have often been reduced to a cautionary tale about foolish workers resisting inevitable progress. The real story is more complicated.
They were skilled workers confronting technological and economic changes that threatened their livelihoods. In the case of handloom weaving, there was a particularly sharp irony. Mechanised spinning had flooded the textile industry with yarn, creating extraordinary demand for weavers and temporarily raising their incomes. Their prosperity helped make mechanised weaving more attractive. Technology first increased the value of their labour and then helped destroy it.
Journalists and writers now face a comparable paradox. Computers, digital publishing, search engines and the internet made modern information workers enormously more productive and gave many of them access to audiences and information on a scale earlier generations could scarcely imagine. Those same digital foundations have now helped create AI systems capable of performing some of their work.
The lesson is not that technology is good or bad. Nor is it simply that workers should embrace every innovation. Technological progress is a continuing process of redistribution. It redistributes work, income, skills, economic power and opportunity.
The spinning machine created opportunities for the weaver before the power loom threatened the weaver. The internet empowered the modern writer before AI began competing with the writer.
The most important question is therefore not whether we can stop technological change. History suggests that we usually cannot. The important question is whether we can ensure that the enormous gains created by automation are shared widely enough that technological progress becomes social progress rather than merely a more efficient way of replacing one group of workers with another.
