Robot magic has something in common with the old-fashioned kind that plays in Vegas and at kids’ birthday parties. In both cases, a ta-da! spectacle grabs people’s attention, so they don’t register the human labor that made ta-da happen. The magician draws your eyes where she wants them, away from the well-practiced move that switches the handkerchief for the dove. Similarly, the robot company talks up sensors and mobility and AI, nudging your thoughts away from the human driver who takes over when the machine gets overwhelmed.
Fauxtomation — Astra Taylor’s term for supposed achievements credited to machines that are actually performed by people — is the most egregious form of unseen work in tech, but it’s not the only kind. Also unrecognized are the myriad tiny fixes humans do when they work with robots, and the guidance they provide when machines need to learn a task. We also hear very little about the engineers who rush in to fix systems that don’t work. Even further away from the spotlight are the miners who produce the raw materials required for computers and other components, the low-wage workers who assemble them, and the people who cook their lunches and care for their kids — people who, as Noopur Raval of NYU's AINow Institute writes here, are not so much invisible as unvalued.
If you want to understand how robots are — and aren’t — transforming life, you need to pay attention to all these people. That means working against not just the intentional sleight of hand of companies but also against the classifications that divide types of work and worker. People who focus on cobalt miners, for example, aren’t often in contact with people who study content monitors on Facebook.
So it’s worth noting the new Tech Labor Forum, at Interactions Magazine, launched in this article by Sarah Fox, director of Carnegie-Mellon’s Tech Solidarity Lab, and Seyram Avle, assistant professor of Global Digital Media at the University of Massachusetts, Amherst.
The project, they write, “will consider how the human labor underpinning technological change lives and works—how workers organize, how they are viewed by peers and supervisors, employers, the law, and academics, whose work informs the actions and policies that dictate tech labor.” The aim is to support “the survival and dignity of all who labor in the very globally connected tech industry […]”
“When we speak of ‘tech labor,’ ” Fox and Avle write,
we are not simply referring to white-collar workers in the headquarters of the largest global firms. We are talking about all who work within the supply chain of technological production—from the "creusers" in the Congo who descend into the earth to find the cobalt that powers device batteries, to the Taiwanese factory workers who make chipsets, the blue-collar workers around the world who sell devices, those working in call centers and repair shops from India to Kenya, the "disinformation architects" in Manila who may not be motivated by ideology, and the content moderators in Arizona, who do not have mental health resources despite constantly fielding violent imagery. Yes, we include the designers and engineers in both Mountain View and Shenzhen, but no more than the cafeteria workers who serve them food, the janitors who clean up after them, the online order fulfillment workers, the drivers for platforms such as Uber, DoorDash, and Grab, and the families in Jidong and Mashan who breathe the graphite air and feel the smoke burn their lungs as they work recycling discarded devices.
Where We Go From Here
Avid readers may have noticed that this newsletter hasn’t exactly been coming out at predictable intervals. There are a number of not very interesting personal reasons for this, but the fundamental cause is that I’m still experimenting, trying to figure out how to do this in a sustainable way.
My natural mode of writing is a sort of essayistic weaving together of things I know and things I find out. The result tends to be (a) long, at least by the standards of blog posts read on phones; and (b) time-consuming, because in writing an essay you don’t know, when you begin, where you are going to end up. Often, what I think of as a language problem — I can’t phrase this idea correctly — reveals itself to be a thought problem (this idea is incoherent, or unclear, or untrue). As I tell people when I teach writing, it’s an error to think of language and substance as separate (as in, “I have it all worked out, I just have to write it up”). Writing is thinking is writing, which means I can’t predict very well how long it will take to say something until I have said it.
And yet, there is so much interesting stuff happening in robotics that I don’t want to confine this newsletter to the occasional intermittent essay.
So, moving forward, there will still be pieces like last week’s. There will still be Q&As and book reviews. But there will also be issues that have more brief items — news and pointers to interesting material elsewhere. As readership increases, I hope as well to have discussions in the comments section. A Discord channel I’ve just started may also become the vehicle for audio and video components (I’d like, for instance, to make recordings of interviews available to subscribers). I’ll link to it in a future issue, when it’s up and running.
In short, I’m still experimenting as I seek the best ways to write usefully and accessibly about robotics in the 2020s. Have a suggestion? Feel free to put it in the comments.