Thomas S. Mullaney
Humanists and social scientists who work on computing and new media are subject to daily reminders about how little most technologists reflect upon our work or take our scholarship under advisement. We are gadflies, it would seem, always dwelling upon the negative sides of technology rather than its unquestionable positives. We are technophobes who, either because of our failure to understand the technologies we critique, or perhaps out of envy for the power of those who build them, are akin to millenarian street preachers waving signs that read “The End Is Nigh.” Our critiques are made in bad faith, moreover, as we are so often spotted using in our classrooms, our essays, and our books the very systems, platforms, and devices that we lament.
The profound self-confidence and self-assurance of these same technologists are on daily display as well. Their “bias toward action” is considered refreshing; their predilection to “move fast and break things” is celebrated. Even when systems fail in spectacular fashion, a seemingly endless supply of assuring words are at the ready to defuse or deflect.
Technology firms selling bleeding-edge surveillance systems to authoritarian regimes at home and abroad?
Bad actors.
Latest-generation web cameras incapable of recognizing the faces of African-American users (yet functioning flawlessly with Caucasian users)?1
Bad training data.
Rampant and historically unprecedented growth of corporate surveillance?
Terms of usage.
No matter the problem, it seems, a chorus of techno-utopian voices is always at the ready to offer up “solutions” that, remarkably enough, typically involve the same strategies (and personnel) as those that helped give rise to the crisis in the first place. We can always code our way out, we are assured. We can make, bootstrap, and science the shit out of this.
A largely unreflective rush to automating immense sectors of the labor economy out of existence?
Creative destruction.
Widely documented gender, class, and ethnoracial inequalities across the IT labor force?
A pipeline problem.
Camera “blink detection” technology repeatedly flagging users of Asian descent with the prompt “Did Someone Blink?”2
A bug, not a feature.
Google’s image-recognition algorithm tagging the faces of African-American individuals with the word “gorillas”?3
They fixed it in the next release.
From the standpoint of technologists, the appeal of this argument is not hard to understand. If one believes in it, after all, then one also believes that technologists should continue to enjoy a monopoly, not only over the first-order creation of an ever-increasing number of computationally grounded political, social, cultural, and economic frameworks, but also over the second-order repair and replacement of said systems when they (repeatedly) fall short or break down. Who would want to relinquish such a plum position?
The contributors to this volume—scholars who approach computing and new media from a variety of critical perspectives derived from humanistic, social scientific, and STEM disciplines—have come together to issue a manifesto that can be summarized as follows:
Your computer is on fire.
Humankind can no longer afford to be lulled into complacency by narratives of techno-utopianism or technoneutrality, or by self-assured and oversimplified evasion. Every time we hear the call of a lullaby—soothing words such as “human error,” “virtual reality,” “the Cloud,” or others meant to coax us back to sleep, leaving the “adults” to continue driving—our response should be a warning siren, alarming us and those around us into a state of alertness and vigilance. Every established or emerging norm needs to be interrogated—whether the taken-for-granted whiteness of humanoid robots, the ostensibly “accentless” normative speech of virtual assistants, the near invisibility of the human labor that makes so many of the ostensibly “automated” systems possible, the hegemonic position enjoyed by the English language and the Latin alphabet within modern information-processing systems, the widespread deployment of algorithmic policing, the erosion of publicly governed infrastructures at the hands of private (and ultimately ephemeral) mobile platforms, the increasing flirtation with (if not implementation of) autonomous weapons systems capable of selecting and engaging targets independently, and the list goes on. The long-standing dismissal or evasion of humanistic and social scientific critiques of computing and new media is over. It has to be over, because to allow it to continue is simply too dangerous.
When we speak of “fire” in this volume, we do so in three interconnected ways. Our first usage is literal. Despite widespread tropes that portray computing and new media as immaterial and disembodied—whether through an emphasis on “virtual” reality, “telepresence,” “the Cloud,” “streaming,” the “postindustrial” economy, or otherwise—computing and new media are nothing if not entirely physical, material, and organic. They are physical machines, propelled by fire both material and metabolic. When they run, they run hot; and when they work hard, they run hotter. Data centers alone account for more than 2 percent of global energy use, energy consumption predicted to grow with the expansion of the Internet of Things.4 (Google emitted over 50 kilograms of CO2 in the time it took for you to read this sentence.) Through the studies of platforms and infrastructure, bitcoin mining, programming languages, underground cable networks, and much more, this volume drives home what is often termed the “materiality of the digital”—that is, the physicality of computational and new media technologies that are too often described in ethereal terms.
Computing and new media depend upon flesh-and-bone metabolism. Our “virtual worlds” are made possible by battalions of human beings (as well as nonhuman organisms): cable layers, miners, e-waste recyclers, content moderators, call-center operators, data-entry technicians, and repair technicians, many of whom come from marginalized class, racial, and gendered positions. Computing and new media run on a vast metabolic conflagration. In certain cases, the work of fabricating so-called virtual experiences exposes laborers to a daily regimen of toxic by-products of electronics manufacturing and disposal and, in other cases, to forms of post-traumatic stress disorder (PTSD) that grow out of long working days spent concentrating on realms of the digital world that the rest of us rely upon them to keep out of view: still-frame and live-action portrayals of extreme violence, child pornography, and vicious or hateful speech.
With all of this in mind, we wrote this book to remind ourselves and our readers: Every single thing that “happens online,” “virtually,” and “autonomously” happens offline first—and often involves human beings whose labor is kept deliberately invisible. Everything is IRL. Nothing is virtual.
Fire also signals a state of crisis. Whether in the context of criminal justice practice, accelerating ecological crisis, access to credit and capital, governance, or elsewhere, computation and new media increasingly have life-or-death consequences. They are becoming ever more woven into the fabric of our social, political, and economic lives, and as this happens the conscious and unconscious values of system builders become encoded into the very algorithms that undergird said systems. Inequality, marginalization, and biases are transposed from the social and political world, where they can (at least in theory) be struggled over in the light of day, and rendered more durable, invisible, unavailable to human audit, and thus largely unassailable within the realm of computation. In this process, inequalities of gender, race, class, religion, and body type find their way into robotics, automated decision-making systems, virtual assistants, code academy curricula, search algorithms, and much more. Expanding beyond the first theme above, then, not only is it essential to remind ourselves that “nothing is virtual” but more broadly that there are dangerous consequences whenever one describes computational and new media forms using sanitized, deodorized, and neutralized vocabularies that exempt them from critical analysis.
Fire also propagates—in which direction and with what velocity is nearly impossible to forecast with certainty. It can spread steadily and predictably, moving out along fronts across contiguous territory. But it can also leap across space, riding atop a single airborne ember, and rapidly surround us.
We invoke this dimension of fire to emphasize the profound difficulties we face when deciding where we as scholars and activists should concentrate our attention and deploy our resources most effectively. Where are the frontiers of computing and new media, whether in terms of emerging and future forms of hegemony or, by contrast, novel forms of subversion and liberatory possibility? Many have already begun to realize that such frontiers may not be located where we might assume. If the vanguard of artificial intelligence research once resided in the defeat of Russian Garry Kasparov by chess-playing Deep Blue, more recently it lies in the 2016 defeat of Korean Lee Sedol at the “hands” of Google’s Go/Weiqi/Baduk-playing system AlphaGo.5 In the same year, China took top prize as the global leader in supercomputing for the seventh time in a row, with its Sunway TaihuLight clocking a theoretical peak performance of 125.4 petaflops, or 1,254 trillion floating point calculations per second.6 Meanwhile, banking systems long reliant upon state-governed infrastructures are rapidly being displaced on the African continent by the cellphone-based money transfer system M-Pesa—the largest mobile-money business in the world.7 And bringing us full circle, many of the cellphones upon which this multibillion-dollar economy runs still use T9 text-input technology—a technology that, although invented in North America, found its first active user base in Korea via the Korean Hangul alphabet. Where should the fires be fought? Where should fires be kindled (perhaps even lit)?
Each essay in this book is framed as a direct and uncompromising declaration of fact. When you peruse the table of contents, you will see immediately that the essays do not argue that the Cloud “can be thought of” as a factory or that sexism “can be considered” a feature rather than a bug in the history of computing. Instead, the essayists draw upon their formidable research experience and deep knowledge of the technical subject matter at hand to cut directly to the heart of the matter. The Cloud is a factory. Your AI is a human. Sexism is a feature, not a bug.8
Our choice of phrasing is deliberate. Collectively, these essays are meant to serve as a wake-up call to both our colleagues and students in STEM fields, articulating these points in ways that, by being unequivocal, do not permit equivocation from our readers. Debate, certainly. Disagreement, quite likely. But evasion or facile techno-optimism? No.
These declarations of fact are also meant as a wake-up call to humanists and social scientists. By insisting on empirically rich, nuanced, yet unapologetically direct and bold arguments from all of our contributors, this volume represents a recognition, as it were, of a kernel of truth in the widespread dismissal of our work by those who are closest to the centers of production in the computing and new media economy. Critics like us, we must admit, have pulled our punches far too often. In spite of years of research and intellectual labor, a deep and hard-earned command of the technical systems under examination, as well as rich evidentiary and theoretical bases, we too often stop short of where we could take our arguments and conclusions. Too often, we hedge our bets and use formulations like “can be thought of as,” or “is analogous to,” or “is comparable to,” and so forth—formulations that leave readers wondering: Does this author really mean what they say, or are they merely proposing a “way of seeing”? We are in a position to make urgent and reasonable demands of ourselves, our elected officials, and our most decidedly unelected industry leaders whose actions (and inactions) in large part define the lives we lead. The time for equivocation is over.
It is also important to stress that the goal of this volume is not to leave readers with the impression of a world, or a future, that is foreclosed and dark. Each contributor strikes a balance between certain sober realities about computing and new media, while always acknowledging their ubiquity, often their indispensability, and the sometimes awe-inspiring nature of technologies that have or are quickly becoming part of our everyday lives. Likewise, while the contributors grapple with questions of inequality—inequalities of gender, race, language, and more—we also remain vigilant about not inadvertently reproducing such inequalities by means of stereotype threat. While we feel it is essential to take an uncompromising look at ongoing dilemmas of gender and racial inequality, for example, the goal is not to dissuade readers—whether women, persons of color, first-generation students, or others—from pursuing their passions within the context of one or another STEM field. This volume is not, in other words, crafted as a call of despair but as a call to arms. We very much hope that the students who read this book will go on to take up positions in STEM fields, and then to agitate therein on behalf of the issues we raise.
I wish to thank my coeditors, all participants in the two Stanford conferences, and one conference at the Computer History Museum, that helped give rise to this volume; our editor Katie Helke, and her colleagues at the MIT Press; and Joseph Rulon Stuart, for his capable work on the index.
1. “HP Computers Are Racist,” YouTube (December 10, 2009), https://www.youtube.com/watch?v=t4DT3tQqgRM.
3. Kate Crawford, “Artificial Intelligence’s White Guy,” New York Times (June 25, 2016), https://www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.htm.
4. Fred Pearce, “Energy Hogs: Can World’s Huge Data Centers Be Made More Efficient?” Yale Environment 360 (April 3, 2018), https://e360.yale.edu/features/energy-hogs-can-huge-data-centers-be-made-more-efficient.
5. Cade Metz, “Google’s AI Wins Fifth and Final Game against Go Genius Lee Sedol,” Wired (March 15, 2016), https://www.wired.com/2016/03/googles-ai-wins-fifth-final-game-go-genius-lee-sedol/.
6. Thomas S. Mullaney, “The Origins of Chinese Supercomputing, and an American Delegation’s Mao-Era Visit,” Foreign Affairs (August 4, 2016), https://www.foreignaffairs.com/articles/china/2016-08-04/origins-chinese-supercomputing; Thomas S. Mullaney, “‘Security Is Only as Good as Your Fastest Computer’: China Now Dominates Supercomputing. That Matters for U.S. National Security,” Foreign Policy (July 21, 2016), http://foreignpolicy.com/2016/07/21/china-taihulight-sunway-encryption-security-is-only-as-good-as-your-fastest-computer/.
7. See Paul N. Edwards, “Platforms Are Infrastructures on Fire,” ch. 15 in this volume.
8. The essays in this volume address a wide variety of subject matter and case studies, ranging from human-in-the-loop content moderation to Bitcoin mining, from mobile banking platforms to game design, and from undersea cable networks to keyboard interfaces, among many other examples. At the same time, the volume does not aspire to be a “conspectus” on computing and new media. There are noteworthy omissions, such as high-frequency trading, 5G and the question of cellphone radiation, disability studies, biohacking, election hacking, and a great many other subjects. Our hope is that this volume, rather than trying to exhaust all possible avenues of discussion, provides a clear enough road map to help inspire the student and reader to grapple with subject matter we have not addressed. In this sense, the volume is well positioned for use in a classroom and for providing a well-structured yet open space for students to take some of the issues explained here into new areas of consideration. Moreover, the essays are written in a crisp, propulsive tone, and the word count is kept intentionally low, to enable instructors to assign them in clusters to their students in both STEM fields and the humanities. In this way, our goal is to help instructors help their students think comparatively and expansively.