‘Your Computer Is On Fire’ draws on tech history to critique AI and the cloud

In a story last year about nine books I read about AI in 2020, I called Your Computer Is On Fire a book worth watching out for this year. It’s released this week, and I was not disappointed. The premise of the book is that techno-utopianism should die because it’s too dangerous to be allowed to continue. This argument came up recently in the context of Amazon workers in factories with robotics getting hurt more often than workers in factories without robots. But once people throw away unrealistic visions of outcomes, the history of technology looks very different.

The book attempts to interrogate how the legacy of social constructs and media narratives have shaped computing. It invites people to think critically about notions of purity surrounding data, the concealment of the carbon footprint the cloud represents, the whiteness of robots, and the wires and resources involved with making the world wireless. Your computer is on fire in part, authors argue, because of automation that perpetuates racism and sexism, and the growth of resource-intensive datacenters and the cloud at a time when climate change is an existential threat for the planet.

The title of this book is meant to prepare you for a series of 16 provocative essays that consider the history of technology, media, and policy, from Siri disciplines and the cloud as a factory to how the internet will be decolonialized and tech for the Global South. Each essay takes readers on a journey through a topic to consider the ethical and societal implications of technology over the long term, an approach former Ethical AI team lead Margaret Mitchell suggested for Google.

Contributors to the collection of essays include Safiya Noble, author of Algorithms of Oppression, who wrote an essay about race and gender stereotypes that permeate robotics and the role of robotics in policing, prisons, and warfare.

“We have to ask what is lost, who is harmed, and what should be forgotten with the embrace of artificial intelligence and robotics in decision-making. We have a significant opportunity to transform the consciousness embedded in artificial intelligence and robotics, since it is in fact a product of our own collective creation,” Noble wrote in the book.

Another essay, by Nathan Ensmenger, argues that the cloud is a factory, and it examines the extent to which datacenters demand a lot of energy, water, and the mining of rare mineral resources like cobalt, which has led to accusations that Big Tech companies aided in the death or serious injury of children. That essay also walks through a comparison between Amazon online today and Sears mail-order catalogs a century ago, and compares Amazon transportation and distribution strategy to Standard Oil.

Understanding, for example, that in the past women made up much of computation work treated as menial and feminine for most of its early history helps illuminate ongoing problems of racism and sexism in tech environments that women — especially Black women — describe as toxic.

I also found something terribly human in an essay arguing that a network is not a network, which looks at the history of large networks built in Chile, Russia, and the United States. Benjamin Peters says that history shows that just because a network works does not mean it works as its designers intended.

“[N]etwork projects are twice political for how they, first, surprise and betray their designers, and, second, require actual institution building and collaborative realities far richer than any design,” Peters wrote.

Editors of the book include Mar Hicks, a tech historian at the Illinois Institute of Technology in Chicago and an associate editor of the IEEE Annals of History of Computing. They are joined by science and technology historian and University of California, Irvine professor Kavita Philip; Peters, a media historian and University of Tulsa professor; and Stanford University history professor Thomas Mullaney.

The editors take pains to state that the book’s conclusions aren’t meant to be an overly dark view of the future or to give people the impression things are hopeless. There is hope, they argue, but recent trends should act as an alarm.

What I also took away from this book is the continuing value of critical analysis. In a recent paper, researchers recommended reporters persist in sharp questioning, declaring, “Technology journalism is a keystone of equitable automation and needs to be fostered for AI.”

In the final pages of the book, Your Computer Is On Fire also addresses the role of media and the writers of narratives in tech and AI trends.

“Tech will deliver on neither its promises nor its curses, and tech observers should avoid both utopian dreamers and dystopian catastrophists. The world truly is on fire, but that is no reason it will either be cleansed or ravaged in the precise day and hour that self-proclaimed prophets of profit and doom predict. The flow of history will continue to surprise,” Peters writes.

Even if you’re like me and follow trends in artificial intelligence through news, books, and research papers, you may still learn parts about the history of technology in this book that you didn’t know, because this book extends across an arch of history. And as editors lay out in the afterword, they hope the messages contained within will be viewed as obvious decades from now.

This lens — viewing computing and artificial intelligence across the span of decades — and consideration of social and historical context was previously espoused by Ruha Benjamin, who last year argued in the context of deep learning that “computational depth without historic or sociological depth is superficial learning.” But the collection of impactful tech issues interrogated over the span of decades in this book makes it recommended reading for anyone interested in the impact of tech policy in businesses and governments, as well as people deploying AI or interested in the way people shape technology.

This book presents compelling arguments for essential topics at the center of business and society. By using computational history as a foundation, it’s able to, as Noble put it, “underscore how much is at stake when we fail to think more humanistically about computing.”