What can the fate of hand spinning tell us about the AI revolution?
“The future sometimes seems to sob a low warning of the events it is bringing us.” [Charlotte Brontë, ‘Shirley’]
Hand-spinning (the production of yarn from plant/animal fibres), once provided work on this island for around 8% of the population, and nearly one in six women and children. Because it could fit neatly around the agricultural calendar, it gave many families in rural areas a crucial income, and by the mid-18th century it was a healthy, growing occupation for thousands.
And then, within about fifty years, hand-spinning had been basically eliminated as a job. It was wiped out by new automation technologies – in particular, the spinning jenny, water frame and spinning mule.
At a time when every organisation and business seems to be racing to embrace AI, it’s worth spending a moment on the fate of hand spinners. The demise of this group, who were once amongst the largest of all industrial employees in Europe, has been almost entirely forgotten. But they whisper a warning from the past.
Did the inventions that replaced them bring higher productivity? You bet. Take cotton. Your typical jenny had about 80 spindles – your classic mule had roughly 120 – and they could produce 3.25 and 4.5 pounds of cotton respectively, per day. Add in the even more productive 60-spindle water frame, and by 1787 these machines would have used up the majority of all cotton imports. As a result, the market share of cheaper, lighter cotton textiles made from factory-spun yarn, grew significantly. Mechanism in the spinning of wool and flax also led to increased productivity, and hand-spinning of fine linen yarn had been entirely replaced by the 1830s.
But if you don’t happen to be the owner of or major shareholder in an AI company, the more pertinent question might be, what was the cost of this productivity? Looking at the jenny, historian Adrian Randall estimates that just over 7 men, 6 women and 14 children using the machine, could replace the work of 153 women and their 241 child assistants. Historian Benjamin Schneider estimates that from 1770 to 1851, this all shakes out to around 600,000 women and children losing spinning work and income. 600,000.
The decline is reflected in documents like the 1834 ‘Report of the Poor Law Commissioners’, where roughly half of responding English or Welsh parishes report having little or no work for women (the north-west, south-east, East Anglia and Wales seem worst hit). And remember, this is work that had previously helped rural families make it through hard winters.
A poem from 1791 called ‘The Discarded Spinster’ describes the devastation poverty caused in rural communities (set against some benefits for industrial areas). It left many rural families depending entirely on a single male breadwinner, with no other safety net and a wider negative impact on many women’s living standards. It likely played a big role in the decrease in women’s participation in the national workforce that lasted until the 1980s.
Families with no able-bodied men, or stuck in places where there wasn’t much other work available, would likely have had a pretty terrible time. As historian Brian Merchant underlines, “spinning was an important contributor to household income, and its automation dealt a serious blow to family wealth and stability.”
If your mind immediately goes to the ‘can’t they just reskill’ argument – and why wouldn’t it, given how many times you’ve probably heard it – it’s worth noting that even by the 1830s there wasn’t enough replacement work for many of the women or children who’d made money from spinning wheels in the 18th century, or for their descendants.
There were new jobs in textiles, but for less than 1% of the population, who were mostly men. Some people were able to find work in factories, particularly former cotton hand-spinners, but many didn’t – especially women who lived further from the new mills. Other work opportunities that were considered ‘for women’, like lacemaking and knitting, didn’t suddenly grow enough to employ all the under- or unemployed women.
Was any of this resisted by the workers? Absolutely. When James Hargreaves, the inventor of the spinning jenny, unveiled his creation, a bunch of cotton spinners got into his shop and kicked the crap out of it. Richard Arkwright had one of factories burned to the ground in 1779. A booklet in 1780 refers to machine-breaking riots caused by the declining incomes and unemployment of the cotton spinners in Lancashire, Derbyshire and Cheshire.
You might condemn the violence, but for these individuals and families this wasn’t a discussion about some exciting theoretical future technology, it was a real thing being used to replace their livelihoods in the present. They were understandably furious, and desperate.
As historian Adrian Randall writes, “where machinery augmented or facilitated employment within the existing work structure, it was rarely resisted. It was when machinery threatened employment or relocated it into a new factory-based system that it generally encountered hostility.”
Lancashire cotton spinners tried appealing to Parliament in 1780. But you can guess how that went. A parliamentary committee concluded that on balance, machinery was beneficial for the industry and for the country (it certainly was for their factory-owning mates). What was that? You’ve noticed some similarities with the way modern tech titans are treated by politicians? I couldn’t possibly comment.
So, let’s say we wanted to draw some lessons from this event for the AI revolution that’s now begun. Our first could focus on the claim that enough new jobs are created by new technology to cover for any employment that’s lost. Every time this is mentioned, the experience of the hand spinners may guide us to ask where those new jobs are, how many who’ve lost previous work can easily move into them, and whether they’re jobs of the same quality or value to the individual. What might be the impact of this shift on people’s wellbeing – both if they can find the offered replacement work, and perhaps more especially if they can’t?
More broadly, media discussions around the march of work-displacing technologies seem to skip over the sheer shock of their impact on local communities. I doubt that many were having fun hand-spinning. This was very hard work. But I’m sure if you could ask them, they’d say it was a tad more fun than poverty. Modern workers being fired to create space for new automated processes might say the same.
The next lesson, or question, is who benefits? A dramatic increase in productivity and sales of cheaper clothing brought by the jenny, frame and mule, was built on the back of workers’ new lives in a factory, with its poor occupational safety, discipline and long, hard hours. AI could lead to significant benefits too, especially in areas like healthcare, but the meaningful advantages for most of us of large language models like ChatGPT, are more questionable – especially when set against their potential to replace a whole range of job areas and the huge environmental cost of data centres. Perhaps Henry David Thoreau sums that one up best: “Our inventions are wont to be pretty toys, which distract our attention from serious things… improved means to an unimproved end, an end which it was already but too easy to arrive at.”
The third lesson might be one of scale. Hand spinners were not a one off. If you’re currently looking at job losses in, say, the legal sector, and thinking this probably won’t reach your industry, think again. During the Industrial Revolution the disruption happened again and again. Soon after the spinners it was the turn of the weavers, as Edmund Cartwright’s power loom took off. Whilst productivity soured, the average weekly pay of a weaver in Lancashire fell from the 25 shillings of 1800 to the 14 shillings of 1811.
So, what could be the ahead of us? As author and historian Brian Merchant notes, it’s not the robots threatening our jobs, it’s the “managers who see a cost benefit to replacing a human role with an algorithmic one… business-to-business salesmen promising AI content and automation solutions to executives… CEOs who see an opportunity to reap greater profits in machines… they’re the ones coming for your job.”
In December 1811, The London Statesman reported… “There are 20,000 stocking-makers out of employment. Six regiments of soldiers from different parts of the country have been sent into this town… Nottingham jail is full of debtors, and the country is equally distressed. No trade; no money… God only knows what will be the end of it; nothing but ruin…”
Our remedy, as ever, could be found in stronger labour laws, union power and real community action. Or perhaps in wealth taxes focused on tech tycoons, which then contributed to boosting our social safety nets. But… that may be an article for another time.
Thanks for reading. Throughout this, I’ve borrowed a whole bunch from two brilliant sources:
‘Technological Unemployment in the British Industrial Revolution: The Destruction of Hand-Spinning’ by Benjamin Schneider – which goes into far more detail than above and is definitely worth a read.
‘Blood in the Machine: The Origins of the Rebellion Against Big Tech’ by Brian Merchant, which is 100% one of the best books I’ve read this year.
I’ve also referenced or gained further info from…
Berg, Maxine. ‘The Machinery Question and the Making of Political Economy, 1815-1848’.
Brontë, Charlotte. ‘Shirley, A Tale’.
Noble, David. ‘Present Tense Technology: Technology’s Politics’ in ‘Democracy: A Journal of Political Renewal and Radical Change’.
Randall, Adrian. ‘Before the Luddites: Custom, Community, and Machinery in the English Woollen Industry, 1776-1809’.
Thoreau, Henry David. ‘Walden’.