Luke Munn is a media studies scholar based in New Zealand and has just released the book Automation is a Myth. The book analyzes automation as a triple set of fictions: full autonomy, universal automation, and that of automating everyone. However, these myths ignore social, cultural, political, and geographic forces that shape technologies at the local level. Drawing on diverse stories of workers – from machine supervisors in China to human pickers in warehouses in the United States – Munn provides a nuanced, localized and racialized understanding of the so-called “future of work”.
Munn is also the author of Unmaking the Algorithm and Logic of Feeling. In this interview given to Rafael Grohmann, he talks about the new book, involving the myths of automation, left and automation, workers and automation, platform labor as a type of parasitic and alternative futures of work.
DIGILABOUR: In a nutshell: why is automation a myth?
LUKE MUNN: Automation is a myth because it is a grand narrative with a singular framing and misguided assumptions. That understanding of the “future of work” obscures actually existing labor conditions on the ground and distracts us from the issues we should be focusing on.
DIGILABOUR: You specifically address three myths: full autonomy, automation everywhere, and automating everyone. Could you explain each one of them briefly?
MUNN: Sure. There is the myth of “automating everything,” claiming that machines will take over production and supplant humans. But far from being self-acting, technical solutions are piecemeal; their support and maintenance reveals the immense human labor behind “autonomous” processes. There is the myth of “automation everywhere,” with technologies framed as a desituated force sweeping the globe. But this fiction ignores the social, cultural, and geographical forces that shape technologies at a local level. And, there is the myth of “automating everyone,” the generic figure of “the human” at the heart of automation claims. But labor is socially stratified and so automation’s fallout will be highly uneven, falling heavier on some (immigrants, people of color, women) than others.
DIGILABOUR: Do you think that even leftists have succumbed to automation myths, like accelerationists?
MUNN: The ideal of liberation from work has a long history in the leftist and Marxist traditions. So technologies which offer to automate work away, to remove that burden and free up the worker for other pursuits (political, social, leisure, etc) have long been of interest. I’m sympathetic to this desire, although also think that any vision of a post-work future needs to be interrogated carefully, not least because it’s so compelling to so many people. The leftist tradition has been very good at attending to the material conditions of labor: what’s the environment, what’s the hourly rate, how long are you working, what’s the organisational structure, what kinds of pressures are placed on the body? This book and much of my research on labor and technology in general is inspired by that approach. And if this is your analytical mode — if you’re actually getting a glimpse of conditions on the warehouse floor, or low-paid migrant labor, or the precarity and toxicity of content moderation work — then the myths of automation seem to vanish pretty quickly. So I would say it is when leftists lose this historical focus, this material focus on work “at the coal face,” and move up into the 10,000 foot view of GDP and employment statistics and Moore’s law and overblown tech rhetoric, that they begin to succumb to automation myths.
DIGILABOUR: What are the new ways of working introduced by “automated” systems?
MUNN: Yes, I explore this in detail in chapter 2: “Spotty Automation and Less-Than-Human Workers.” The key point in much of this labor, as you hinted at with the scare quotes, is that “automated” systems are often propped up by immense amounts of human labor. In some cases, this is about workers filling in the gaps, the weak points in technical systems. Human workers use their affective or cognitive labor at key moments to maintain functionality for customers: to make things work. Machine minders are one example here; “automated” checkouts are another — and these kinds of roles introduce new roles on laborers and new kinds of pressures. So it’s not that technologies don’t change labor – they do – but that the grand narrative of automation has no way to account for this kind of “piecemeal” alteration or the specific pressures placed on specific people in specific contexts. In some instances, so-called “automated” systems are used to provide a veneer of objectivity and screen actually-existing labor conditions. Content moderation is a key example of this I explore in the same chapter. Social media platforms speak about “the algorithm” when blocking and flagging content, a kind of machinic rationale which is often taken up in popular discourse. But of course, those systems are underpinned by thousands of workers, often in low-paid and precarious positions, who work for third party companies that are sometimes located in the Global South. These workers are forced to read hateful comments and watch watch toxic videos, which takes a brutal psychological toll. In these cases, we can start to see how automated systems are less about the work itself and more about the packaging of work — how it can be sliced up and offloaded to the lowest bidder, someone who is legally and operationally distanced enough from the original company to avoid any liability. And in the book, I argue that this is really the way that automation “removes” the human from labor. It removes the full worker, on full pay, with full rights (insurance, benefits, and so on). And it is these kinds of impacts, at once more subtle and pathological, which get overlooked by the automation myth but which should demand our attention.
DIGILABOUR: I found the story of the human pickers very interesting. What’s “old” and “manual”/ “analog” in on-demand labor involving automation issues?
MUNN: I like to quote Mauss about the human body being our first and oldest technical object. So in that sense, the body will always be at the center of labor and the labor struggle. And the body comes with a set of capacities but also constraints. Digital technologies, like the gamification systems used by Amazon, attempt to optimize labor performance, to ramp up the pace and intensity of work. But the body is double-edged, using its capacities also means coming up against its constraints. This might be exhaustion, or damage to the tendons and joints, or it might take the form of frustration, and in that case we see acts of sabotage: dragging feet, blocking sensors, gaming the system, or even just walking off the job and quitting. So it cuts both ways, and I think it’s key to understand the potential of the worker and her body. In the book I mention the point system and automated firing system that Amazon implements, but recently we’ve seen news stories stating the company is struggling to retain workers given its enormous turnover. How do we respond to this intersection between digital systems and the “analog” worker? In the book I highlight some work on collaborative or “augmented” automation, where researchers are attempting to understand the complex dynamics at work here. Much of this is driven by computer science or HCI researchers, and so is very technically focused, mixing engineering notions of efficiency with more recent ideas of human well-being and benefit. But ultimately, this comes back to much messier and more fundamental questions: who is “the worker”, what should work be like, and what role should technology play in elevating human agency and dignity? Answering these questions adequately would mean drawing on wider strains of scholarship: Marxist theory, feminist theory, decolonial theory, and so on. And these questions about the nature of work and worker well-being are not just “philosophical” but very urgently being asked by workers and employers in our post-COVID context.
DIGILABOUR: You speak in the crowdsourced/platform labor as a type of increasing parasitic mode. Why?
MUNN: The point of crowdsource/platform labor is not to use labor, but to extract it parasitically. Using labor is conventional: it means acknowledging that a company is the employer and they have employees. It means a commitment to paying a living wage over a longer time frame. It means supporting them with benefits and health insurance, and being responsible or liable in some way for the work they carry out under your name. Contemporary capitalism wants none of that. This is why I speak of exhaustion or extraction rather than use — and this is an argument from my first book, Unmaking the Algorithm. Exhaustion means obtaining the productivity of labor, while retaining strategically distanced from the worker and her life-world, with all of its risks and needs. The company gets the work without the worker, so to speak. From the perspective of capital, this mode of labor is a kind of “triumph” that solves for many of its core logics. In the last couple of years, we’ve really seen labor organizers recognise the dangers here and respond through unionization, protests, legislation, and so on. Uber and Amazon are two prime examples that come immediately to mind, and these even involve some of the workers like Christian Smalls I mention in the book. The victories here, and the immense amount of organisation and activism needed to achieve them, shouldn’t be dismissed. Yet the core logics at the heart of our current economic systems remain, and so the push to recreate this extractive mode — whether under a new name or in new contexts and industries — will persist. So I really see this struggle as an ongoing one that requires sustained vigilance.
DIGILABOUR: How to refuse automation – or build critical automation – towards alternative futures?
MUNN: In the final chapter, I talk about refusing the universal “future of work” offered in automation discourse and starting to develop multiple “futures of work” driven by community needs. I talk about the Lucas Plan and also mention Data for Black Lives and the Maori Data Sovereignty Network as two examples of organisations who are currently doing this kind of important work. I think part of the book’s contribution is to stress that there can be no grand, universal solutions here. We need to think concretely about who the workers are, what kind of labor is being done, what technologies are being deployed, and ultimately, what does a community want to achieve? As one of the labor activists in the book mentioned, it’s not about refusing technology, but about ensuring that workers get a seat at the table and access to its benefits. All that said, I feel like this book is really only the starting point to this question. I actually submitted a recent grant proposal that aims to build on that work. Automation is a Myth has setup a more critical foundation, and so it’s a matter now of envisioning alternatives to that future. AI, automation, and digitalization will profoundly shape how work is carried out in the next decades. And yet it’s also very clear from recent events — whether its algorithms that perpetuate historical injustice or technologies that foster racialized and gendered working conditions — that the so-called future of work does not work for many people. We need more research that is able to understand how these technical systems work but can also fold in deep insights from media, race, and cultural studies, bringing together the technical with political. My next project hopes to do that and start envisioning data-driven technologies that foster more inclusive, more communal, and more sustainable forms of labor.