This is an entry in our Keyword series, where we try to define the terms you’ll often hear when people talk about building a better Internet and put those keywords in their current context.
In the early days of the COVID-19 pandemic, people and organizations needed a new virtual space to host all kinds of interactions that previously took place in-person. Suddenly the video chat app Zoom was hosting church groups, city council meetings, lectures, seminars, community organizing, concerts, and everything else that people do offline.
Inevitably, there were some issues. Up until the pandemic, Zoom was mainly an enterprise video chat company with a significant engineering presence in China, a reality that quickly bumped up against its new role as a public infrastructure. For example, Zoom shut down a seminar at San Francisco State University with former Palestinian militant and hijacker Leila Khaled, citing anti-terrorism laws. Additionally, Zoom worked with the Chinese government to terminate Americans’ accounts and shut down events highlighting the Tiananmen Square massacre.
At iDPI when we talk about the problems facing our digital public sphere we often refer to accidental public infrastructure. Zoom is a great example of this. It is an enterprise video chat company at its core, but during the pandemic it became indispensable public and civic infrastructure. Much of our digital public sphere is made up of accidental public infrastructure: tools, platforms, and services that were designed to solve a specific problem—e.g. staying in touch with friends, putting ads on your website, preventing DDoS attacks—but have grown into indispensable infrastructure for public and civic life online and offline.
When we’re talking about infrastructure, what exactly do we mean? One simple definition of infrastructures is “the technologies and systems necessary for society to function.” Infrastructures are fundamental systems that allow us to build other systems—new houses and businesses rely on the infrastructures of electric power lines, water mains, and roads—and infrastructures are often invisible so long as they work well.
Infrastructures tend to be expensive to build and difficult to scale. When a town outgrows the capacity of a small power plant, it’s expensive to build a larger power plant, and the larger plant will be underutilized for some time to come. As a result, costly and “bulky” infrastructures are not always built by private businesses. They are often built by governments, because they are expensive, because their benefits can take a long time to be realized, and because sometimes only governments are well-positioned to capture the revenue generated by infrastructures, through taxation.
Infrastructures generate externalities, both positive and negative: the road that connects the small town to the nearby city leads to new homes and businesses (positive economic externalities) and to increased noise and air pollution from increased car traffic (negative environmental externalities).
Social scientist Eric Klinenberg argues that infrastructures also exist in the social world. Schools, universities, health care systems, news organizations, legislative and judicial systems, public safety and policing systems, public spaces and libraries all operate as social infrastructures, enabling society to function. The challenge of conducting business in a society without a functioning court system or daily life in a country without a police force serve as reminders that like other infrastructures, social infrastructures allow more complex forms of interaction to occur.
Infrastructures have a set of responsibilities that come with their critical role. Those responsibilities are codified in various ways including government regulation and professional norms. For example, water utilities’ responsibilities are codified in regulations requiring, among other things, nondiscrimination, price restrictions, and quality controls. Similarly, news organizations’ responsibilities are codified through professional norms encouraging fairness, transparency, and evidence.
Oftentimes, when public infrastructure is accidental, its responsibilities are an afterthought—either ignored entirely, or grafted on ad-hoc in response to events and public pressure. At iDPI we advocate for a different approach: building intentional public infrastructure. Intentional public infrastructure are tools, platforms, and services that recognize their critical public and civic role and center their responsibilities. Intentional public infrastructures offer an alternative to existing infrastructures, placing competitive pressure on the market. Intentional public infrastructures are designed and implemented in a way that upholds their responsibilities, leading to different experiences for the people who interact with them.
We outlined a vision for an intentional public infrastructure in our paper “Forgetful Advertising: Imagining a More Responsible Digital Ad System.” We propose a digital advertising system structured around a single design choice: avoiding the storage of behavioral data. Forgetful advertising can still target ads using information like geography, intent, context, and whatever else can be gleaned from a single interaction between a user and a website, but it cannot remember any previous interactions to inform its targeting. We believe forgetful advertising can make digital advertising compatible with the values of human agency and privacy.
Forgetful advertising sets a standard for responsible digital advertising and offers an alternative to existing surveillant digital ad systems, placing competitive pressure on the market. And forgetful advertising is designed and implemented in a way that respects human agency and privacy, affecting the experiences of people who interact with the infrastructure.
Accidental public infrastructures are an important feature of the internet’s landscape and help to explain many of the problems we’re facing as we move more and more of life online. Building intentional digital public infrastructures that recognize and center their public and civic responsibilities is an essential way to address those problems. We are partial to this approach at iDPI and it guides much of our work.