The proliferation of artificial intelligence (AI) and other automated technologies throughout every facet of modern life is fuelling widespread speculation about the future of work. But Gavin Mueller, a media studies professor and author of Breaking things at work: Why the Luddites were right about why you hate your job, says the popular notion that this technology will inevitably free us from the drudgery of work is a “dramatic oversimplification” of how it interacts with labour processes.
“It’s more accurate to say that machines restructure the work process, and oftentimes it’s restructured in ways that increase levels of exploitation and/or deteriorate the working conditions for the people in those jobs,” he says.
Citing the example of large language models (LLMs), Mueller notes how they rely extensively on human labour to check and moderate the masses of data they hoover up, which means “someone across the globe has to spend eight hours a day looking at the most horrible content put on the internet until they literally can’t do it anymore, until they have so much trauma that they have to stop”.
Mueller adds that a similar dynamic is at play in the “efficiency” of Amazon deliveries that arrive the day after ordering. “That’s coming at the expense of the people in the warehouse whose work has to accelerate constantly,” he says. “It comes at the expense of the delivery drivers who don’t even have a break to urinate.”
This is why Mueller is advocating for the adoption of an adversarial, “Luddite” attitude towards technology, so the current drivers of its development can be effectively interrogated, while also making the case that various forms of “sabotage” can help articulate new critiques of technology while forging the social bonds necessary to challenging it.
From the Luddites campaign of sabotage against the mechanisation of the textile industry and the Wobblies rank-and-file rebellion against factory automation, to students protesting military computing projects at their universities and the free software movement, Mueller argues it’s important to show that technology has always been politically contested, because people still have a tendency to view technological development as somehow detached from the wider society in which it occurs.
“The idea that technology is independent of human decision-making and just progresses along its own trajectory is in fact an ideology,” he says, adding that this promotes a sense of “inevitability” around technology that is both depoliticising and disempowering. “This belief tends to benefit the people who are making the decisions over technology, who own these companies.”
Drawing on the many historical examples of resistance to new technologies that litter the history of automation, Mueller argues that adopting similarly “antagonistic practices” is key to restoring people’s sense of agency and autonomy.
This, in turn, can help articulate alternative political approaches “from below” (i.e. from the perspective of workers) that undermine the increasingly centralised control states and corporations have over the development and deployment of new technologies.
Adopting a similar stance to Wobbly activist Elizabeth Gurley Flynn – who defined “sabotage” in 1916 as anything that leads to the deliberate “withdrawal of efficiency” by workers, from absenteeism and work slowdowns to strikes and the physical destruction of machine components – Mueller also says such practices are key to forming new bonds of solidarity between workers, which can in turn inform new political strategies.
He adds that, for those with qualms about the more destructive practices of sabotage, it should be viewed as a strategic consideration and one possible tactic among many.
Taking Luddism seriously
Although today the term “Luddite” is dismissively used as shorthand for someone wary or critical of new technologies for no good reason, Mueller notes that Luddism was specifically about protecting workers’ jobs, pay and conditions from the negative impacts of mechanisation.
While workplace sabotage occurred sporadically throughout history during various disputes between workers and owners, the Luddites (consisting of weavers and textile workers) represented a systemic and organised approach to machine-breaking, which they started doing in 1811 in response to the unilateral imposition of new technologies (mechanised looms and knitting frames) by a new and growing class of industrialists.
“The stereotype of the Luddites that we have with us today is that they were kind of throwing a temper tantrum by breaking these machines, and how pointless it was because breaking a machine isn’t going to stop the industrial revolution,” says Mueller.
“But that’s not really the best way to understand it: breaking machines was part of a more organised labour struggle to maintain control over their working conditions, over what they were getting paid, who was allowed to do the work, [and] the quality of the goods being produced.”
For Mueller, the Luddite’s physical opposition to the introduction of new mechanised technologies in their workplaces was therefore a considered political strategy designed to help them maintain control over their working lives, rather than a spontaneous expression of dissatisfaction.
He adds that while the practice of workplace sabotage is often dismissed as an “immature tantrum” by critics of all political stripes, it’s important to pay attention to instances where it has happened because it reveals what people do not like and why.
“Breaking it is in some ways a practical critique of these new technologies,” he says. “It’s also a way to slow things down and to advocate for workers positions in other ways. Typically, in really contentious labour struggles where there’s really intense conflict between management and labour, sabotage happens whether it’s authorised [by unions] or not – I think what we should do is see it for what it is, rather than write it off.”
While Luddism was eventually crushed by the British state via a series of brutal hangings, Mueller says it’s important to bring their critical perspectives on and militant actions towards new technologies into our current historical moment.
“I agree that most people work too much, [and] that we should be working less and have more time to do the things that are really important for us,” he says. “We have societies productive enough to do that – to provide a decent quality of living for everyone while working a lot less.
“But the way we get to that – and this is my difference with the post-work utopians – is not through new technologies, it’s by articulating it through labour politics.”
Applying this perspective to the past century or so of automation, Mueller notes that although people have long been predicting and expecting far shorter working weeks, increases in productivity via automation over recent decades have instead translated into people generally working longer and harder.
“A lot of technology we’re told is efficient, and we read that as ‘it makes things easier to do’, but it’s not always so straightforward,” he says. “Efficiency means there is more output for the input, which actually translates into people working harder somewhere down the line.
“We’re not looking at a cybernetic future of either dystopia or utopia – we’re potentially looking at something that’s basically a worse version of what we have now for the people that have to work for a living, which is most of us.”
Part of the issue, adds Mueller, is the way in which work has come to be completely orientated around productivity and growth, which in turn has shaped the direction of technological development.
The Taylorist turn
In his book, Mueller explores how this technologically determinist and productivity-focused mindset has historically been fostered by a confluence of disparate actors with overlapping interests in maximising productivity and growth.
For example, while Fredrick Taylor began developing his theory of “scientific management” in the late 1800s for US industrialists – which essentially revolved around breaking down work into a set of discrete tasks to make the production process as efficient as possible – the Soviet Union ended up adopting its own “Taylorised” production processes.
In both cases, Mueller notes the Taylorist obsession with atomising and measuring an increasingly mechanised labour process was not simply about achieving efficiency and productivity gains, but as a means of exercising power and control over workers by fracturing their source of power at work: knowledge of the process.
“Factory owners and managers might have only a dim idea at how products were actually put together, and no ability to do it themselves,” he writes. “This control of knowledge meant that workers could control the pace of work. Out of desire or necessity, they could slow it down, even stop it all together.
“Scientific management, for all its pretensions, was less about determining ideal working methods and more about shattering this tremendous source of worker power. By breaking apart each work process into carefully scrutinised component tasks, Taylor has cracked the secret of labour’s advantage, thereby giving management complete mastery over the productive process.”
In the case of the Soviet Union’s brand of Taylorism, Mueller notes that “massive state-driven industrialisation and bureaucratic management became the means to create a new socialist subject” and “as a means to enforce labour discipline at the expense of the autonomy of workers”.
Aside from the shared interests of capitalist enterprises and nation states, Mueller says the bargaining strategy of organised labour in the 20th century shifted away from increasing the free time of union members and towards increasing pay in line with productivity, which made them much more open to the introduction of new technologies by bosses.
“There’s a difference of perspective between the upper echelons of the large trade unions, making a lot of those strategic decisions, and the more everyday experience of workers who actually have to deal with these new technologies and how they change their jobs,” he says, adding that while the bosses agreed with linking increased productivity to increased pay, the benefits of this for workers themselves were short-lived.
“That lasted for a relatively brief amount of time, and it was fairly uneven in its implementation,” says Mueller. “In the meantime, the ways that technology restructured the labour process ended up undercutting the very organisations, the unions, that had agreed to adopt them.”
Both the turn to scientific management and the acceptance of new technologies at work by unions happened much to the chagrin of workers, who he says began engaging in various “sabotage” practices, including demonstrations, riots, strikes, slowdowns, industrial sabotage and the circulation of “subversive” materials.
Through these practical experiences of struggle, Mueller says workers were able to compose and organise themselves, as well as forge new practices of solidarity and militancy that did not previously exist.
Linking this to the idea of degrowth (which at the most basic level critiques the capitalist notion of economic growth at all costs) he adds that any general reorientation of economic activity away from the logic of growth and productivity would “be connected to a very different perspective of thinking about how we do work”.
“This is where it connects with Luddism, because Luddism is also a way of thinking about work that is not about growing the profits of the firm, it’s not about doing everything in the most efficient way,” he says. “It’s saying work is a big part of our lives, we deserve to do it in ways that are dignified, in conditions that are decent, and in ways that we agree to, that we have some say over.”
Bonds of solidarity
Relating this to the shift to remote working many experienced after the onset of the pandemic, Mueller notes how enterprises have turned to automated “productivity monitoring” software to track their employees’ behaviour, including everything from keystrokes and mouse movements to time spent on websites or drafting emails.
“[Working from home,] you have more control over your time, and most people found they could do what they needed to do for work in less time than if they had to sit in the office for eight hours,” he says. “But of course this is not something that the bosses like – your manager wants you in the office so they can watch what you’re doing, and if you’re not busy enough, find something for you to do to keep you busy.”
Mueller adds that while this deployment of analytics software is often justified by managers and the suppliers on the basis of reducing worker burnout, the technology can easily be adapted to monitor workers in other, more intrusive ways. However, he says that workers were quick to resist the imposition of productivity tracking technologies, and have already developed a range of practices to subvert employers’ surveillance attempts.
Examples of this include using an analogue watch under a mouse so that it jiggles enough to trick any software tracking mouse movements into registering the workers as active, buying cheap devices online that achieve similar results, or even downloading programs from GitHub to simulate it instead.
“I’m always really interested in these forms of resistance because I think a lot of people that do these things don’t see that what they’re doing is political,” says Mueller. “The conflict over work is always political to me, and these forms of resistance are potentially a place where you start to recognise that. So you’ve got your mouse jiggler, you find out this other person in the meeting is doing it, too, and that’s where things start to happen. I’ve had a lot of jobs, and the jobs where I felt the most solidarity with my co-workers were when I realised we were all breaking the rules together. That creates a different kind of bond.”
He says these initial bonds of solidarity can then help people realise they’re part of something much bigger. “That to me is the beginning of coming to a realisation that work is not just something you have to do, but it’s also a whole political situation that you’re in,” says Mueller.
High-tech Luddism: The alternative politics of technology today
Mueller says that while technology’s development and deployment is undoubtedly directed by capitalist enterprises for their own purposes, Luddism is not about a flat rejection of technology but instead “recognising the politics of technology”.
Noting that both businesses and the technology they deploy are often structured to maximise workers’ atomisation from one another and reduce opportunities for collective organising, he adds that technology also offers opportunities for the exact opposite.
Citing the example of Turkopticon – an “advocacy organisation” for the collective interests of Amazon Mechanical Turk workers – Mueller says “you can also use forms of technology to try to maintain worker autonomy, dignity and enable organisation”, adding that while we should generally be skeptical of new technologies developed for capitalistic ends, there are interesting counter-examples of how it can be used for different purposes.
“Turkopticon is one,” he says. “I think open source software in general can be considered a Luddite technology – it was definitely a deliberate attempt to say, ‘Look, we don’t want corporations to take over programming. We want to do it our way. And we’ve identified that copyright is the line of attack they’re taking’,” says Mueller, adding that this led to the creation of alternative software licenses that allowed people to continue copying and tinkering with code despite the concerted efforts of giant corporations like Microsoft.
Writing in his book, he noted that free software is “an innovation in the interest of the preservation of practitioners’ autonomy against the imposition and control over the labour process by capitalists”.
“By ‘breaking’ software copyright and challenging closed and proprietary business models connected to it, free and open-source software has helped preserve independent and craft-like working conditions for programmers for decades.”
On the example of programmers and similarly technical professions, Mueller adds that while they do tend to be paid much high wages relative to many others throughout the IT sector, it’s important to understand that as wage earners they also have no control over the flow of capital or the direction of the production process.
He further adds that while there’s a lot of debate around how social classes should be defined, there are real limits to an understanding of class that’s tied to how much someone earns or the particular jobs they do. “I think you can derive some insights from that, but I think … it becomes a very complicated procedure to understand what people’s interests are, as people’s interests and situations are mediated by a lot of things,” says Mueller.
“I think what is maybe more important – especially because we live in times that in some way seem political but in many ways are quite depoliticised compared with prior moments in time – is to see who’s kind of already on the same page and who wants to take up those struggles.”
Based on an understanding of class as something born out of practical struggles, he adds that there has been “encouraging solidarity” between, for example, some of the highest paid programmers at Google and some of the lowest paid cafeteria staff.
“These are Google employees, too, and some of the highest paid workers recognise that, so start thinking about, ‘Can we make a change here? What can we do or what does it say about where we work?’” says Mueller. “To me, that’s where you see class in a political sense starting to emerge.”
He further adds that, given the extent to which modern work is broken down into tasks, it’s important to engage with every type of worker, as everyone has skills and knowledge that are helpful in envisioning alternative ways of doing things.
“Whether it’s an IT professional, whether it’s the person emptying the dumpsters, everyone has a load of knowledge about their job that, if deployed correctly, we can use to do these tasks in ways that are fairer, not full of drudgery and less exploitative, and can give us more of a world we’d like to live in.”