How would you classify digital labour? Are we talking about anyone who works at a computer?
We might first try to differentiate between labour and work – to generalize, you could say labour is work where the workers are not necessarily in control of the means of production. Work might be something else, something that will always exist in some quantity, shape or form. With ‘digital labour,’ it’s often hard to say what should be included or not. The term can incorporate all of those things that are organised by algorithms and apps, famously such as Uber taxis, as well as most ‘prosumer labour’ in which you’re producing and consuming a commodity at the same time – as with Facebook and any social media platform that you add value to by participating in.
It’s the side of work that is changing that really interests me. We already have an analysis of what it’s like to receive a wage to sit in an office and do, say computer programming, because it’s very direct; you are making a product within a company, for paid time. But where the term digital labour has mostly come up is with crowd sourcing forms of work, where you have ultra precarious forms of labour, ie. you merely log into a platform, like Amazon Mechanical Turk, in order to go to work. In this sort of scenario, you don’t necessarily know who your employer is, you might not know the bigger picture of what you’re working on, and most of all, you have no contact with other workers.
How did you make contact with workers in such a fragmented field?
At first, I set myself up as an employer (it’s called being a ‘requestor’) on Amazon Mechanical Turk. This is a crowdsourcing platform which employers can use to commission work requiring human input such as labelling images, sorting through data, transcribing audio recordings, and market researchers will sometimes use it to circulate surveys. You hire someone to complete a task through this website then see the work being completed in a progress bar, so the whole interface is designed to look as though you’re interacting with a computer system, but on the other side of the system there are people doing the work. This has been called ‘artificial artificial intelligence’ because it’s not actually ‘artificial’, in a technical sense.The racialised name ‘Mechanical Turk’ stems from the story of an 18th Century automated chess machine which toured the world defeating famous people at chess. It seemed to be one of the first intelligent automatons that could defeat humans, but as it turns out, there was a small, brilliant person trapped inside the machine who was actually playing the chess. So it wasn’t automated at all. Amazon adopted the name because it made a similar thing: the interface looks like a machine and it does the work that people expect a machine to do. The fact that they get away with marketing it that way is quite scary I think.
Anyway, the fake ‘company’ was called Wages for Crowdwork and it was funded by a small grant plus some credit card debt. The paid task was to answer the question ‘would you participate in a strike against digital labour?’ With a box for comments below. There were thousands of responses from people around the world: some people just said yes or no, and others wrote an explanation or declaration. A lot of the arguments for and against related to geographic location, for example, some workers in India communicated something along the lines of, ‘I depend on this for my daily income, to feed my family, so I can’t afford to strike, and even if we do strike there are thousands who will take my place’. Other people, from various regions, said ‘absolutely, this is fundamental, we can’t let these systems continue.’ I stayed in contact with people who wanted to talk and we’ve worked on a few projects together since.
What sort of scale is digital labour operating on?
Mechanical Turk, although it has become one of the most well known platforms of its type, represents a relatively small proportion of the labour force, with around 500,000 workers. But there are plenty of projects with a huge magnitude of labour behind them. Google Books is another example. At some point Google decided to scan all the books and they want to put them into the public domain, but this also allows them to create a rentier economy: it simultaneously seems generous and scary because they are holding something they want everyone to use. The labour that is required for this is generally gruelling, repetitive and low-paid. If you look for long enough at the scanned books you will find troubling images of fingers or hands across a page because while people are working so fast all day, repetitively scanning page after page after page, every once in a while their fingers are scanned too. Andrew Norman Wilson exposed the tiered employee system at Google and was fired after asking questions about race and treatment of workers. You can see some of this in his video Workers Leaving the Googleplex, or his ScanOps photographs.
There are also things like manual stereoscopic video making – that is the mainstream way of converting 2-D films into 3-D films. It’s a painstaking frame-by-frame process in which workers edit each frame one pixel at a time. Again, it’s vast outsourced workforces around the world doing drastically repetitive tasks that most people assume is accomplished automatically by a computer or automated cameras.
These are just a couple examples of hidden labour that allow contemporary technologies to exist. To answer your question about scale more broadly, if we see the majority of internet or social media as a factory, where its contents (our productive participation) is owned, bought and sold, it is difficult to say where this paradigm of labour stops.
One of the reasons I’m interested in these fields of work is to explore whether or not any of it actually has to exist. I get a hunch that all of things masquerading as benevolent information sharing are really just creating a lot of useless work. When we counter the narrative of this beautiful new peer-to-peer world and see it instead as a massive bureaucracy that is deeply tied to everyone’s emotions, we can bypass the naive optimism that a lot of the huge and powerful tech companies seek to create.
For some people, perhaps it’s convenient that their phone and computer now connect to their bedroom dimmer lights or car security system, and their phone connects to their fridge, which automatically connects to the grocery store or their doctors’ office, to school reports and satellite positioning of their kids, and to the local police station: maybe all that is really convenient. But for me it can seem, at very least, a bureaucratic mess. I think we should take care to be optimistic and hopeful about things that actually involve cooperative sharing and cooperative control over power; to make sure they are not relying on hidden or exploitative labour and surveillance as their central premises.We should bear in mind that although digital services are presented as a ‘cloud’, as an immaterial, floating, virtual thing, they are very concrete at every level of the system. From the mining of the precious metals required to make a computer; through to assembly, to the type of work it takes to maintain data; to the energy that data is stored on; then the applications that people use it for – being attached to phones and interconnected, thus creating profits for these massive corporations that are using people’s emotions and tracking their activities. For instance, to go to a shop and have someone else know that you went to the shop, and then for that data (along with millions of other bits of personal data) to become a commodity to be sold to someone else for a profit – all of this is very concrete.
There’s always a double side to any of these technologies and networks unless they are owned cooperatively, for example when you think about the surveillance potentials that often masquerade as ‘efficiency,’ like ‘contactless’ payment devices. In terms of co-operatism, a lot of people are now talking about distributive systems – so you don’t have centralised servers humming away in big buildings owned by an elite class, but everyone is working on their own project that they own together, with their own servers, and I think that’s the only way the internet could really work well. Because as long as the internet is owned centrally then you’ll always have this problem of ‘I’m giving away my free labour.’
What do you mean by the internet being owned centrally?
There are different arguments about the origin story of the internet. The way that I think about it, which is pessimistic, is that it, as an idea, first came about during WW2 by the American military. After the invention of nuclear weapons, they needed a way to make sure that their information and communication networks could not be knocked out in a single hit – so, they drafted systems that could organise things laterally. I’d say that’s very generally where the internet (and its mathematics) began, a long time ago. One thing it still has in common with these troubled beginnings is that it’s managed by a bunch of check points that are highly securitised, because that’s what the military, and businesses, do. In the way that it’s distributed and the way it is provided, it is mostly based around decentralized nodes of control which are owned centrally, if that makes sense. This conflicts with the tech-utopianism of the 60s, 70s and 80s, which held the idea that everything will be free, everybody will be interconnected, there are too many of us to ever stop and too much information to be stopped. If that information is whirring about on Google’s servers somewhere, it has already become a commodity.
Dialectically, with this issue of ‘too much information,’ you can see almost total corporation control in the forms of ‘cloud-computing,’ or ‘big data,’ but also incredible resources such as wikipedia or possibilities for whistle blowing, etc.
Interview continued below. A short video introduction to Byron’s work:Video researched and edited by Siobhan McGuirk.
Are there any online activist projects that are inspiring you right now?
With many online campaigns I’m reminded of the psychoanalytical term ‘interpassivity’ – where rather than interacting with something you are both being actively passive together. That is how you could describe a lot of ‘clicktivism.’ It’s not enough that a virtual campaign helps people learn about something if that learning is used as a sort of decorative knowledge they can just carry with them and it makes them feel better. This can reinforce the status quo, by having a more educated populous that is still lacking any sense of action or spirit to change the world because they felt like they already did. It takes the place of solidarity and community organising and cooperative risks, so I’m interested in when things leave that passive zone.
One example could be the website called ‘co-worker.’ I’m not sure what it has become right now, but I found the initial idea of it was very interesting when I heard of it in 2013. The story goes like this: say if you‘re interested in organising a strike in your workplace, rather than the notoriously difficult ways of initial organising – sending discrete text messages, trying to meet people after work and carefully gauging opinion without letting the boss know, co-worker would simply allow you to organize a petition to the people you work with and they can sign up in private. It helps you plan an initial meeting and when you all meet in a room for the first time everyone there will have virtually said that they want to organise. Then the work really begins of course, but that difficult, mediated initial step of bringing like-minded people together in an exteriorly controlled environment, has been relatively painless. The technology did what it does best, but didn’t pretend to replace in-person, embodied relationships. It only helps create the possibility for a first meeting.
This is an example of what the internet can do best – make light work of something that is hard to do interpersonally, while not taking away from the fact that we have still got to form new relationships in person. Ideally, no-one needs co-worker after the first meeting in that room, so it just facilitates something that didn’t happen before. There are similar ideas around the world, in terms of rent strikes, or, for example, in debt strike campaigns in the USA – if enough people can manage to show each other commitment to default and coordinate this beforehand. Hey, imagine if the millions of people with medical or student debt in America who would default anyway, agreed to strike on payments at the same moment! These sorts of things can turn vulnerability and exploitation into collective power. My point is that these types of networking technology should be used intelligently to accomplish particular goals and they are best as markers for participation, rather than pretending they can hold participation itself.
Top image, Entrepreneurship is Bullshit; Black Square from Capitalism Makes us Sick, at see www.byronpeters.com