In 2013, Tyler and Cameron Winkelvoss announced their decision to invest around $11 million in Bitcoin—around 1% of the cryptocurrency’s total market cap at the time. They declared: “We have elected to put our money and faith in a mathematical framework that is free of politics and human error.” 

That decision has paid off handsomely for the twins. But it’s much less certain whether their reasoning has borne out. Nearly a decade on, the blockchain systems they praised as faultless have birthed a chaotic new world filled with excessive hype and elaborate cons — some leading to financial implosions that have devastated ordinary participants. 

These dynamics may be particularly extreme in crypto, but they’re not unique to it. What these stories underscore is that no technology or digital space—however elegantly conceived—fosters trust solely through its design. 

Trust is an elusive variable that lies within most of the key questions about tech’s role in our lives. How do we know a machine, or program, will work as planned? How do we determine if someone online is who they claim to be? How can we make sure a digital space is truly safe?

But trust is also about power. When a dominant platform asks for our trust, do we have a choice, or just the illusion of one? If a technical system predetermines a marginalized person to be less trustworthy, do they have any recourse?

We take these provocations as a starting point for this issue of New_ Public Magazine, which explores the idea of trust in digital systems. As always, we look for the human concerns that lurk just behind the technological ones. Our goal is not simply to pose whether trust is necessary in digital infrastructure—it undoubtedly is—but to approach the topic from unexpected directions that inspire new thinking. 

There are 11 pieces in this issue, under three broad themes. We hope you’ll discover insights within these stories that can accompany you as you imagine, build, and inhabit the digital spaces of the future. We are grateful for your attention. 

Wilfred Chan
Editor, New_ Public Magazine

The Blockchain Con Men

On a night in 1990, Tommy Carmichael woke up with a brilliant idea. He had just invented the “Monkey Paw”: a flexible metal wire connected to a rod that could be used to cheat slot machines. Twenty-six years later, a hacker (whom we will refer to as “0x15def”) sat in front of his computer screen, inspired by another brilliant idea. He came up with a way to exploit an “immutable” software protocol which enabled him to amass an amount of cryptocurrency worth $55 million at the time. He became known as the attacker of TheDAO: one of the world’s first decentralized autonomous organizations.

This story is about the debate over whether a technology can truly be “trustless.” As shown by recent events like the dramatic implosion of the Terra cryptocurrency, a so-called “algorithmic stablecoin” which relied on code to help it maintain a peg to the US dollar, supposedly trustless systems can have ruinous consequences for communities organized around them when they fail.

We ask: is there a better way to think about trust in blockchain technology, that does not reject it, but brings it thoughtfully into the picture?

Cheating TheDAO

TheDAO, launched on the Ethereum blockchain in 2016, was envisioned to operate like a decentralized version of Kickstarter. Participants could pool their cryptocurrency together, and then vote on how to invest it—without a centralized operator taking a cut. It was supposed to counteract the rent-seeking inherent in “platform capitalism” and enable participants to have greater control in the decision-making of an online platform. The basic building block of TheDAO was its “smart contract,” which—like the slot machine—was designed to operate mechanically, without any possibility for human interference.   

As originally conceived by blockchain pioneer Nick Szabo, smart contracts are embedded in a deterministic world. If the world of the slot machine is one made of metal, levers, and gears, the world of the smart contract is one made of code—which, unlike natural language, is unambiguous and executes exactly as it is written. 

The uniqueness of TheDAO was that it created an organizational form in which every interaction is technologically determined. No human would ever need to be trusted, as nobody would ever be able to cheat the system—in theory, at least. TheDAO would thus be a towering testimony to the widespread idea that blockchain is a “trustless” technology; that we don’t need to trust any banker, accountant, or stateswoman anymore. That notion would be put to the test on April 30, 2016, less than three months after TheDAO went live, when the attacker used an unintended feature of TheDAO code to drain millions of dollars worth of Ether—the native Ethereum cryptocurrency—into a single wallet. 

This was a devastating blow to the users of the Ethereum protocol, which sent the price of Ether tumbling as panic and anger raged on online forums, tearing a rift in the Ethereum community. 

Many considered the attack to be theft, calling for an immediate intervention. They advocated for a “hard fork”: a new version of the Ethereum blockchain, identical to the original but with the stolen funds transferred to a new smart contract, from which the original token holders could redeem their funds. Such a solution could not, of course, be achieved unilaterally; it would require an intervention by all network nodes to collectively “switch” to the new protocol.

Others, however, believed that regardless of the theft, the principle of immutability should at no cost be violated, as it was a basic prerequisite for the technology’s “trustless” nature.  Informal public polls leaned against forking, with one user asking: “What’s the point of using blockchain tech if a 3rd party acts as judge and police deciding what’s right and wrong?” Forks should, according to this view, only relate to technical fixes intended to improve the security of protocols. 

Eventually, the lack of consensus between the two groups led to a split of the Ethereum network into two separate networks: a dominant version, where the theft has been undone, and a less popular “classic” version where the theft has been left untouched.

TheDAO attack and its aftermath revealed something important about the claim that blockchain is a trustless technology. Like Tommy Carmichael and his Monkey Paw, the anonymous hacker 0x15def taught us that even the most mechanical and deterministic machine can be cheated. As advanced and sophisticated as the code of a smart contract might be, we can never assume that it is immune to abuse. 

When these systems fail, it is humans who always remain in the loop. Even in the most deterministic and tamper-resistant blockchain networks, participants still need to trust the decisions of core developers, validators, miners and crypto-currency exchanges, and other relevant stakeholders. 

When the confidence in Ethereum was compromised in the course of TheDAO attack, core developers had to be trusted to develop a practical solution to address the hack, while ensuring the proper functioning of the blockchain network, in line with the majority of token holders’ sense of fairness and justice. Miners and validator nodes had to trust each other to collectively upgrade their software to implement the hard fork when the update was released by the core development team. The fact that a minority of miners decided to remain on the original protocol is a sign that such distributed trust could not be taken for granted.

Designing trust for trustless technology

TheDAO attack showed that blockchain technology isn’t as trustless as it is claimed to be. And yet, the myth of trustlessness and immutability continues to capture everyone’s imagination. So let’s reverse the question: if blockchain is intended to operate without trust, what is it offering instead?

In 2000, the German lawyer and sociologist Niklas Luhmann made an argument that trust should be distinguished from confidence. He claimed that trust is a feature of social interaction that allows us to take risks by putting ourselves in a vulnerable position. Consider trusting a friend to deposit a thousand dollars in cash at the bank: this trust might be betrayed, for she might run away with the money. 

Confidence, argued Luhmann, cannot be betrayed. It does not rely on human agency but instead on general knowledge, prior experience, or statistical evidence about the way a system operates. It derives from the predictability of future events that comes from the deterministic qualities of the system. Now, consider the slot machine. When we operate a slot machine, we are confident that putting in a coin could, according to the logic of the machine, result in a payout. 

Blockchain technology, like the slot machine, has been designed to build confidence. By combining technical features of deterministic computation, cryptographic primitives, and distributed consensus mechanisms, blockchains manage to generate confidence in a system that stretches across borders and cultures. You don’t have to know your peers in a blockchain network to be confident that a transaction will be executed. In other words, it is a “confidence machine.” 

Yet, even though trust and confidence are distinct, they are still related. Indeed, building confidence in a system—as we saw with TheDAO—requires trusting a network of actors operating the system. The actions of con men like the TheDAO attacker revealed that confidence in a system can never be absolute, and that when it is violated, trust needs to be rebuilt.

The difficulty of bringing trust back in

One popular approach to create more trust in the blockchain space has been the establishment of “off-chain” governance mechanisms, such as non-profit foundations that raise money to support the development and maintenance of protocols, and interact with real-world policymakers. These foundations, while sometimes described as unaccountable to the networks they serve, generally do not exercise direct control over these protocols. 

Some cryptocurrency networks, like Dash, have taken a different approach by vesting their “masternode operators” (participants who are paid in the Dash token to maintain a copy of its blockchain and validate blocks) with the power to vote on governance and funding proposals. Additionally, these masternodes can participate in the election and removal of the Dash Core Group’s board of directors through the democratically-controlled Dash Trust. This techno-legal arrangement enables the Dash network to own property and interact with the physical world, while still being controlled and administered “on-chain.”  

Yet, other examples reveal the limitations of an on-chain plutocracy. 

Consider the case of Steemit, a “decentralized social network” with over 1.5 million users . To avoid centralized control, the network was designed to be governed by a limited number of “delegated witnesses” elected by the users of the Steem blockchain, whose votes are weighted by the amount of cryptocurrency they lock up in a smart contract. 

Yet, there was a problem: a commanding percentage of Steem’s cryptocurrency tokens had been pre-allocated to the network’s founders. In 2020, Steemit fell into crisis when one of the founders sold his tokens to Justin Sun, a wealthy crypto-entrepreneur, who declared he would transfer Steemit to his own TRON network. Despite the protests of its community members, Sun was able to force through a takeover of the Steem network, violating users’ implicit trust that major stakeholders would not use their influence to sway voting behavior. 

Cases like this are why many blockchain communities prefer the chimera of “confidence without trust” over “trust without confidence.” Indeed, as the trustless character of a blockchain is often the raison d’être for its development and adoption, many blockchain communities prefer to hide behind the façade of “trustlessness” rather than admit that the systems require trust to operate. 

But this means many users go on falsely assuming that blockchain eliminates the need for trust—and the con men exploit this ignorance to their own advantage. Each time they do so, users end up learning hard lessons.

TheDAO attack was both a curse and a gift to the Ethereum community. By showing the limitation of the ‘confidence’ machine, it opened the eyes of many blockchain users to the existence of trust and importance of governance, and the need to design structures that could account for both. But there are many others who still haven’t gotten the message. 

Recognizing that slot machines and blockchains are not trustless doesn’t make them useless. Indeed, people use slot machines not because they are trustless, but because they come with a promise of potential gains, at low entry costs. Various applications of blockchain technology also hold a similar promise. But for a machine to truly generate confidence, we must focus not just on the design of the machines themselves, but on building a better trust infrastructure around them. That means people—even con men—will always have a role to play. 🌳

Primavera De Filippi is a Research Director at the National Center of Scientific Research in Paris, and Faculty Associate at the Berkman-Klein Center for Internet & Society at Harvard. She is the author of “Blockchain and the Law” published by Harvard University Press.

Wessel Reijers is a philosopher of technology, working as a postdoctoral researcher at the University of Vienna. He investigates the ethics of emerging technologies, with a focus on blockchain and social credit systems. Wessel is the author of Narrative and Technology Ethics.  

Morshed Mannan is an academic and lawyer. He is currently researching blockchain governance and platform cooperatives as a Max Weber Fellow at the European University Institute in Florence. He is part of the ERC project on BlockchainGov and a Research Affiliate at The New School.

Illustration by Josh Kramer.

Editor’s note: Names have been changed throughout this piece to protect the privacy of individuals.

George needs work, but technology keeps getting in the way. Because he’s currently unhoused—sometimes at an Austin-area shelter, but other times in a hand-me-down tent—his phone goes for long periods without a charge or cellular data. A talkative, sixty-ish man with a ready smile, George has known people in his neighborhood for years, including the staff at the public library, where he can use the computers and try to borrow a phone charger. When he can’t, his texts pile up unread, and that means he misses opportunities for small jobs. 

Once, George had trouble accessing his email account at a library branch because the email service provider had flagged the login attempt as suspicious, and he was unable to receive text messages at that time to confirm his identity. “Damn thing,” he laughed, recollecting. “Too smart for its own good, huh? Couldn’t prove it was me for the life of me.”

I met George at a local nonprofit that offers free technology classes, like the one on computer basics that George was taking. In the course, students go from learning how to use a mouse and keyboard  to browsing the web, creating email accounts and using other common applications like spreadsheets. But as a cybersecurity expert, I was there because I was curious about how a recent  evolution in online authentication impacts some of the web’s forgotten users.

In the cybersecurity industry, long plagued by overpromises and hype cycles, the current buzziest buzzword is “zero trust,” a network security paradigm which most users encounter in the form of multi-factor identification. Zero trust is a departure from the traditional network security paradigm, which focuses on defending against cyberattacks at the perimeter of a network and implicitly trusts requests made within the network. When the perimeter model was first implemented, with the invention of the firewall in the late 1980s, computers were stationary machines connected to a network shared by an organization in a single physical location; it seemed only natural that any potential threats would be external.

In subsequent decades, however, as the capabilities and demand for remote access has grown, it has become too easy for attackers to pose as legitimate users. In response to this reality, the zero trust framework treats each request as equally untrustworthy, eroding the notion of a perimeter, but increasing the level of scrutiny on each authentication attempt and each device that attempts to connect to the network. 

For example, if an attacker wants access to your online banking account, it’s possible that they could gain direct physical access to your computer, but it’s overwhelmingly more likely for them to gain access through credential theft from thousands of miles away. Therefore, a zero-trust implementation requires not just a username and password, but proof of possession of your personal device, through a code in a text message or recited over a phone call. And since both phone calls and SMS messages have documented vulnerabilities, the more secure methods require either a smartphone or an additional accessory, like a security key.

But this does not work for everyone. For one thing, it assumes each user has a personal device. This alone would already exclude many of the people I met at the computer class—who often simply can’t afford one. 

According to Pew Research Center, about three-quarters of U.S. adults own a laptop or desktop computer, and 97 percent of U.S. adults own a cell phone of some kind, with 85 percent owning a smartphone. Smartphone dependency, or reliance on smartphones for most or all online access, has declined from 20 percent in 2018 to 15 percent today, but remains at 27 percent among people with a household income less than $30,000 and 32 percent among those with less than a high school degree. One-quarter of Hispanic Americans are smartphone-dependent, as compared to 17 percent of Black Americans and 12 percent of White Americans. 

Most of the students at the class fell into one or more of these categories, including Diane, a Hispanic woman in her mid-forties.

When Diane emigrated to Austin, Texas three years ago, she had a hard enough time navigating English, to say nothing of new technology. She did not have a cell phone and had never used a computer. One of Diane’s sharpest memories was her initial puzzlement at QR codes, which are used by some multi-factor applications to register phones. “I saw it first at the restaurant where my nephew works,” she said. “They got rid of all of the menus because of COVID, but there were no instructions, no anything.”

Diane stayed in a spare room at her sister’s and hoped to find work housekeeping, like she had done previously, or better yet, a custodial job that would provide more regular income. However, when she went to the office of a custodial staffing firm, they told her that they were only accepting online applications, which required her to list both an email address and a phone number. She ended up writing down the contact information of a family member. “That was when I felt kind of stuck, because on the one hand I needed work, but on the other hand, how was I gonna work if I didn’t have money for a cell phone?” Diane told me.

Some families share devices, which creates complications. I met Rosana, a woman in her early fifties who shared a flip phone with her husband for many years. She told me about one particular headache when she hoped to set up an online account with her bank, only to find that doing so required her to register a phone number and verify it through an SMS passcode—but her husband had taken the phone to work that day.

But having a phone of one’s own doesn’t solve everything. Recently, Rosana got a secondhand smartphone at the insistence of her daughter, Rocío. But Rosana felt overwhelmed. “(Rocío) knows how to do everything, but I just see all this stuff come in and it’s like, whoa,” said Rosana, referring to messages she receives over text and on apps like Facebook and Whatsapp. After someone she knew got scammed out of some money, she worried she’d fall victim as well.

Cybersecurity products and platforms have always been aimed primarily at companies and government bodies trying to secure their workforce, for the simple reason that that’s where the money is. According to cybersecurity firm Duo’s most recent State of the Auth report, people who are currently employed are 20 percent more likely to have used two-factor authentication (2FA). In the enterprise market, the typical user profile is a knowledge worker who logs in at regular times, from the same network and the same device. There may be variations, such as during business travel, but most importantly, the user is associated with at least one known device that they consistently use and that can be manually updated to the latest software. 

But as more web services begin requiring device-based second factors, it is the most technologically vulnerable who get locked out. In the zero trust framework, the behavior of people like George, Diane, and Rosana might be considered anomalous. The fact that they use shared computers means their IP address constantly changes, which subjects them to more security challenges—putting an extra burden of proof on the individuals least equipped to provide it. 

On one hand, such measures improve the security of the account enormously. On the other hand, erecting these barriers present real hurdles for low-digital literacy individuals who are trying to open an email account, a bank account, or otherwise participate in a world that others take for granted. And while this group’s online activity is low by definition, and their accounts usually aren’t high-value targets for hackers, it is these same individuals who stand to lose the most from having their account locked out or taken over by an attacker, since they tend to occupy more precarious positions in society.

Some veterans of the cybersecurity world have already raised concerns. Wendy Nather, Head of Advisory Systems at Cisco Systems, has long decried the state of usability in security, writing in a 2018 blog post, “Security is important, but it shouldn’t be this hard. It shouldn’t require a user to have to learn multiple user interfaces and acquire several apps and accessories just to perform one simple task.” At her 2020 keynote address at RSA Conference, one of the biggest cybersecurity conferences in the world, Nather talked about what she calls the “democratization of security”—the need to shift the focus of industry away from control and toward collaboration, with products and policies that are “designed to be adopted rather than just engineered to be enforced.” It’s a message more of us should heed.

To be clear, zero trust is a big step forward in securing online services. But we can’t give up on protecting those who don’t have a cell phone or a computer. The Covid-19 pandemic has exposed a stark digital divide in our communities, as more people have had their livelihoods determined by whether they can reliably do work over the internet. To democratize this security framework, we must directly address the needs of the personal-device have-nots.

In one promising initiative here in Austin, the city’s Office of Telecommunications and Regulatory Affairs has invested $3.5 million into digital equity-focused nonprofits over the last two decades through the Grant for Technology Opportunities Program, and since 2016 has helped to provide people with refurbished smartphones, tablets and computers through the Community PC Program. Other major cities have started programs to subsidize broadband service and offer digital literacy training. Then-mayor of Chattanooga, Tennessee, Andy Berke, argued last year that to supply such services was akin to paving roads, something that could not be limited to only the wealthy. “We have to start talking about this as infrastructure, not as luxury.”

When I returned to the computer class a few weeks later, students in the computer basics class had already made lots of progress, typing slowly but surely to fill out web forms. Although security was not formally part of the curriculum, their instructor encouraged them to choose long, random passphrases to protect their accounts, and the students navigated through the various applications they had practiced. George was proud to have joined a meeting on Zoom earlier that week, to participate in a conversation hosted by another community organization around providing services to veterans. “I didn’t know how to do that before,” he said, referring to the video call, but “I spoke my truth, and they all heard.” 

Maggie Engler is a technologist and researcher focused on security, trust, and safety. She currently builds machine learning for platform health at Twitter and previously did data science at Global Disinformation Index and Duo Security. She lives in Austin.

Design by Josh Kramer.

Photo via Manchester City Library on Flickr, with an Attribution ShareAlike Creative Commons license.

The last time I used the BabyConnect app, it was to record nursing my son at bedtime. It was April, the day after my birthday, when X was 14 months old. I had gotten my second Covid vaccine shot that day, and I was ready to emerge (I thought) from the claustrophobic huddle of life with a pandemic baby. The feed lasted 23 minutes and at the end of it I hit pause on the timer, rather than stop. The clock kept running in the background, dutifully keeping track, until the next time I opened the app, which, as it turned out, was 7,148 hours and 35 minutes later. Perhaps it seems odd that I didn’t close it after that last feed with a sense of finality. But that’s how change happens with children: you don’t notice you’re done, or they’re done, and you certainly don’t get to say goodbye. Not to the binky with the fluffy white bunny rabbit attached, gripped, dropped, and washed so often it turned gray like roadkill, and not to the months when it dangled out of his mouth morning and night until one day, he just stopped needing it. The binky floated off on its canoe to who knows where, and we never noticed it go. 

BabyConnect is a timer and tracker app that exists to mark the details of a baby’s development, to reassure new parents that time is passing and progress is being made. It was recommended to us, in no uncertain terms, by other parents we trusted, so we duly downloaded it. With its grid of gentle, pastel-colored circles and white icons, it’s simple to set up and use through the fog of anxiety exhaustion. The most important trackers are at the top, for sleep, diapers, and feeding—breast, bottle, or solid—but you can also record your baby’s mood, activity, medicines, and milestones. If you’d like, you can go even deeper: the app allows you to share information with your pediatrician, upload photos, keep a diary, make shopping lists, and call your partner. We stuck mainly to the first row of the data grid, and the only photo we uploaded was the header image, taken when X was a few hours old, tiny and sleeping and hooked up to monitors. Every time I opened the app, multiple times each day and night, there he was in his own little circle: a peaceful, fragile little stranger. 

It turns out that five dollars is a very small price to pay for the illusion of control: a vital, sustaining illusion in the first weeks of parenthood. Tracking his feeds, sleeps, and diaper changes was a way of marking our progress, trusting that the amassing of experience was progress, that we knew more after ten, 45, 600 naps than we did before. More about ourselves, more about this strange new reality, more about this new person we’d brought into the world without consulting anyone. In the kinder, gentler schools of parenting advice, there is a tendency to reassure you that “you know your baby best.” But did we? In those weeks before smiles, before even eye contact, he felt as mysterious in my arms as he had in my belly. If he was known anywhere, it seemed, it was in the charts and graphs of BabyConnect. The data, at least, had our backs. For the rest of it, we were on our own.

X arrived three and half weeks early, at the end of February 2020, when visitors tramped into my shared postpartum room in regular clothes—no masks, no gloves. My mother and stepfather flew over from London for a three-day visit to meet the newborn, and we planned for when they’d come back later in the spring. The day after they left, the flights were grounded. We live in Astoria, Queens, around the corner from a major hospital, and by the time we reached my baby’s original due date, New York was a pandemic epicenter. Through the rest of March, April, and May I sat up at night in an armchair by the window feeding the baby while red lights slid across the glass and sirens rose, wailed, died, and rose again, relentlessly. 

After those first social days we were alone: even our closest friends, neighbors who live upstairs and whom we’d planned to count on for babysitting, were shut away, sharing cocktails over Zoom instead of in our living room. We knew it was evening by the seven o’clock claps and hollers for the essential workers. Our friends who lived alone went months without a human touch, while I had never felt so touched—clung to, bitten, cuddled, pummeled, puked on, and worse—in my life. The worst of it was that we couldn’t share, couldn’t say Here, hold this baby. There was nobody but us to take his weight. 

I’ve never been more reliant on my phone than during those months, watching social media feeds like a world through glass, then toggling between BabyConnect and a library e-book app when Twitter got too much. I manipulated the phone one-handed while I settled in to feed, trusting the minutes ticking by to translate into calories and ounces that would flesh out X’s skinny chicken legs enough to eventually fill out a newborn onesie. At first, I also hooked myself up to a breast pump, which at least proved that I was producing milk, but with nowhere to go and no prospect of separation, I gave up. I transferred my faith to the timer in the app. 

In ordinary times, BabyConnect might be revealing about the division of domestic labor, since it labels who entered the information. But in our abnormal state of isolated togetherness, both working and parenting from home, my husband and I shared everything. Did you start the timer? I’d whisper-call from the other room, when we were sure the baby had gone down for a nap. My husband would change a diaper and I’d leap into action as some kind of bizarre scatological stenographer, to record the result. The app gave us a wealth of options: buttons for poopy, wet, or dry, and yes/no options for leaks, diaper cream, and “open-air accidents”—which, with baby boys, are more like gleeful target practice. Poop can be further classified by size, consistency, color. Was it more “soft” or “claylike”? “Runny” or “watery”? “Hard,” or “little balls”? A shade of brown or mustard, or alarming red, white, black? We drew the line at uploading photos, but used the comment field to vent our feelings: “Good god so much poop.” “Five-wipe tsunami.” “La Brea tar pits.” Once, simply, “DUDE.” 

This is the daily reality of early parenthood, universal and utterly particular: brain fog and poop jokes. In this way, BabyConnect resembled a much older form of recording, the notebooks mothers have kept for generations, like the ones my mother found and read to me over the phone, a patchy record of my older brother’s first months (these systems usually drop off with subsequent children.) I didn’t keep a notebook consistently, feeling both the practical limitations of finding a pen and paper and a light, but more heavily the pressure to make the writing good. I couldn’t even fill out the pre-printed pages of a perky “new mum’s diary” I received as a gift; the experience was too much, overwhelming the checklists and pages set aside for those elusive milestones. It should be poetry, this experience, this overwhelming, endless, fleeting life, and I didn’t have it in me. And it was also too sad, a reminder of the strangeness of our particular experience: our park outings masked and distanced, no meeting of grandparents, babysitters, other babies. But the app didn’t ask for poetry, nor impose a picture of normal babyhood we couldn’t match. It was experience stripped of feeling, stripped of the specificity of 2020. Just an endless scroll of lovely, timeless, ultimately meaningless data.

The app would turn the data into bar charts and timelines, which we’d study to see if there was a pattern emerging. (There wasn’t.) Even at the time, I suspected that the data was a hedge against the powerlessness of simply waiting for him to grow older, to become physically able to sleep through the night. Recording it, parsing it for patterns and messages helped us feel that we were doing something in that waiting period, while we lay awake listening to coughs and coos and whimpers from the crib across the room, tensing for the escalation into screams. 

When I look back on the data now, I can see growth happening inexorably, sleep stretching out, all the changes that are part of the incredible pace of growth and development in a baby: the doubling in weight, the learning, the kicking-in of new bodily functions and the abandonment of old needs. But I can’t see the story. The app offers a few pre-set menu options for obvious milestones, like “Born,” and let us add our own: “first bath without screaming!” But most of the time you don’t know what’s a milestone, what matters—like the first time he slept nine hours, and it felt like a window opening. The app gives no interpretation of the data, no analysis that will spit out a grade, much to the chagrin of mothers like me, who were dutiful straight-A students. I could tell the app that my baby’s nap lasted 13 minutes, or three hours, and record two hours of wakefulness at 4 a.m., but it wouldn’t tell me what to do about it. 

In a recent WIRED article, the disinformation researcher Nina Jancowicz offered a damning assessment of popular pregnancy apps as “a fantasy-land-cum-horror-show, providing little realistic information about the journey to parenthood.”  Although leading apps like What to Expect and BabyCenter have discernibly worked to counter the woo-woo—being forthright that childhood vaccines are essential and touting the medical credentials of the authors and reviewers of their articles—they are still, Jancowicz argues, in the business of building their subscriber base and mining user data for the benefit of advertisers and corporate partners. But BabyConnect didn’t look at our nights of broken sleep and suggest we spring for a Snoo; nor did it bombard me with the hotly contested pros and cons of breastfeeding. That reticence makes it rare, perhaps uniquely trustworthy, in the suite of pregnancy and new-baby apps. In our pandemic isolation, it was reassuring simply to feel we had another set of eyes on us, and on the baby. Not a judge, but a witness. 

As much as we trusted BabyConnect, I see that we cheated on it, leaving no visible record of the worst disaster of our first year. At the beginning of December 2020, my husband tripped over the outstretched leg of the high chair, holding the baby. A day or so later, we discovered that X had broken his leg in the fall. The unfurling nightmare that followed—of hospitals mid-Covid surge, full-body X-rays, talks with surgeons and social workers, and getting the leg of a baby just starting to crawl immobilized in plaster—is nowhere to be seen in the app. I fed X in curtained cubicles and recorded it, but I didn’t say where I was. There are no gaps in the scroll of data: as mundane and unrevealing as ever. It was a way of reassuring myself that he was still a healthy baby, sleeping, pooping, feeding on schedule, despite the fracture. He would never remember it, the doctors reassured us, and we would never forget it. We made it to London for his first Christmas, and within weeks, he recovered completely.

In that experience, we turned to the community we had—our friends and family—and realized how much they were with us, even if we couldn’t see them in person. My friends told me stories of accidents with their kids that they would never otherwise have shared. In the new year, we put X in daycare for the first time, and he spent time with other babies. Immediately, he went down with a cold that he’s pretty much had ever since. But we found a real community, and we relied less and less on the app. We could see how much he was eating, at the table with us, and it didn’t matter so much how long his naps were once he reliably slept at night. We trusted the data for a little over a year. Seven thousand or so hours later, we’ve started to trust ourselves. 

Joanna Scutts is a historian and literary critic based in New York. Her writing has appeared in venues including The New York Times, The Guardian, Slate, and the Paris Review. She is the author of The Extra Woman and the forthcoming Hotbed: Bohemian New York and the Secret Club that Sparked Modern Feminism (Seal Press/Hachette).

Illustration by Josh Kramer.

Photo by Christian Bowen, via Unsplash.

Editor’s note: This piece was updated on June 16, 2022 with additional information from Google.

In 2018, I received an angry email from a Google engineer over an article I wrote that described a project he had been working on as “not really transparent.” The project I was critiquing in that article was Google’s Perspective API, a machine learning project that analyzes toxicity in language, and was intended to be used in commenting sections online. I’m a researcher of machine learning and online harassment, and at this point, I had been analyzing and following the Perspective project for over two years.

The engineer wrote, “You claim that our code and data is not public, but it is, and has been since the start,” and he was partially correct: some of their code for the project I was critiquing was on GitHub, and so was one-fourth of their dataset. But not everything was open sourced for this project, and there wasn’t any meaningful data on their model and its real-world performance. 

I’ve been thinking about transparency in technology for a few years now. Too often, software teams create the appearance of transparency, without really achieving it. Open sourcing a part of a data set is a good initial first step, and sharing code is good too, but none of that contextualizes or explains what the code does, or how it was made. Or importantly, what the pitfalls or harms that could exist within that code. 

What I’m interested in is meaningful transparency: a kind of transparency that is equitable and understandable, particularly for complex technology like machine learning. That’s why I felt it was important to look closely at Google’s Perspective API web documentation. If the goal of Perspective is to make healthier public spaces using machine learning, then the public must know how it works, and that includes a general audience, not just engineers. 

First, a little more background on the product. Perspective API tries to mitigate and detect online harassment by rating toxicity and harassment in language. Its machine learning model is trained on commenting data from English-language Wikipedia, the Economist, The New York Times, and The Guardian. To my knowledge, Wikipedia is Perspective’s open source data set, while the data from the news outlets has not been shared publicly but did help influence the model. But there’s no clarity on how exactly the model was trained, using what metrics, and who trained it. 

It’s important to note that not all harassment data is the same. The linguistic construct of a harassing news article comment can be different from the kinds of arguments someone has on Twitter, Instagram, or Facebook, which in turn differ from harassing posts on posts on Reddit, and Metafilter, or sites like Wikipedia. In machine learning, the quality of the data matters, along with the context of where a model will be placed in a product, and how that model is used or interacted with by users. 

Some questions I had when learning about Perspective included: How does this model understand things like toxic language? Who trained it and where did the data come from? How well does it work in an everyday context, and how do people use it? Measuring toxicity with machine learning isn’t perfect, and Perspective recommends on its website that it be used alongside humans. But is that enforced, and how? What are the feedback mechanisms or harm reduction mechanisms put into place when Perspective is used? Do clients even like it? 

When I first visited their website in 2019, I found not a lot of answers at all: just a hip, purple layout, an oblique definition of toxicity, some samples of already scored language, and notably, an interactive demo at the bottom that often produced poor results. Some phrases I entered into the demo would be rated as “toxic” when they were clearly not: for example, just the word “Muslims” was rated as over 80% likely to be toxic.

I could not find who made the project, maintained the project, or anything about their case studies. To learn more about the API itself, I had to click on a link that said “For developers.” It turns out Perspective was developed by a Google research initiative called “Conversation AI,” which wasn’t clear from any documentation on Perspective’s website, other than one sentence mentioned on the “Developer’s page.” 

Buried toward the middle of the page was an explanation that the model is trained on “wiki data,” which is open source data from the Wikimedia projects. The page broke down a very in-depth process that a seasoned researcher would understand, but would be incomprehensible to anyone without engineering knowledge.

To its credit, Perspective’s current website has much more information, including links to updated blog posts and model cards. Clicking on “How it Works” now offers a more robust and in depth explanation of the model and how toxicity is defined, along with case studies. The developer site leads to Kaggle and GitHub pages that appear well-maintained. 

It also made an important improvement to its interactive demo: it would no longer show a rating for a phrase below a certain threshold of toxicity. This time, writing “I am a Muslim” brought up no toxicity rating. 

But even with this amount of information, Perspective still isn’t sharing anything concrete or legible for non-technical readers. If you can’t code, it’s nearly impossible to understand what Perspective does and doesn’t do well since none of the models, data sets, or experiments are concretely explained in any way. If you can code, it still requires time to analyze Perspective’s findings. 

It feels like opacity through engineering speak. There’s no way to tell what works, what doesn’t work, and how the product truly functions from reading their materials alone. Even clicking on the case studies provides little analysis as to how Perspective was actually used by The New York Times, what their product flow or integration looked like, what went well, what didn’t go well, and what had to change for one case study versus another. To truly, meaningfully understand how machine learning works, there needs to be analysis not just on the model, the data set, and the code, but also how it performs when launched and interacting with people. 

After a few emails back and forth with the Google engineer, I stopped hearing from him a few months later. He moved to another country, and like a lot of corporate tech employees, he moved to new products. Last month, I emailed Perspective for their thoughts. Given that the original engineer said this was a transparent project, does Google still see Perspective that way? I received no response until after this article was published, when a communications rep from Google sent me a list of links to “additional information” about Perspective—most of which I had already read, and none of which addressed my central concerns. The rep added that “we are limited by what data our partners are willing to share, but always happy to share where we are able to.”

Perspective continues to market itself as a large, self-contained, and ready-to-use product. But machine learning isn’t a discreet, self contained entity at all. Machine learning is like salt: it’s not that interesting on its own, but like how salt transforms a dish, machine learning attains meaning as it interacts with data and users, and changes whatever product, software, or problem it’s working within. 

That’s why it’s not enough just to share raw data. Even if data is shared, we still need legible descriptions of that data. What’s in it, who made it (when?), has it been updated (when?), and how is it used? This can be articulated in graphs, percentages, and paragraphs. But for there to be true transparency, to engender trust from users, what must also be shown is how the product, and the algorithm are reacting to that data, along with how the model was formed. This can be shown through demos, gifs, plain language explanations, and yes—open source code. 

I think meaningful transparency consists of three integral pieces: legibility, auditability, and impact-ability. Legibility is whether most audiences can understand it. Auditability builds on legibility, and is whether an outside party can understand a process, data point, or intention well enough to request changes or give feedback. Impact-ability builds upon legibility and auditability, and refers to the ability for users or individuals to affect change and/or decision making from a project and then see the ramifications of that change or interaction.

That last part is important. Mols Sauter, a professor, researcher and author, said in 2019, “What we’ve lost in this rush towards transparency is now we have the ability to just see things. but just seeing something isn’t really have an impact on something.” Questions we should ask: What happens if and when someone leaves feedback, either via a contact form or Github? How does the dev team respond to them? Can users see that their feedback is effecting change?

Perhaps, the future of transparency in machine learning means shipping products with well-designed archives of explanations, gifs, dials, and contextual examples. For every model card and data sheet: a series of legible graphs to help guide what that analysis means and why it matters. Such an archive would record the changes a product has undergone, and take time to explain the product’s processes. These kinds of explanations should be a duty of care.

Today, I have the same questions about Perspective as I did in 2018, and I still want them answered. If we are to truly understand Perspective, to be able to talk about its strengths and weaknesses, there has to be meaningful transparency. Otherwise, we are just guessing alone together in the dark—forced to take Google, and other companies, at their word.  

Caroline Sinders is an award-winning design researcher, an artist analyzing technology’s impacts in society, and a Contributing Editor at New_ Public. She’s worked with the Tate Modern, the United Nations, Ars Electronica’s AI Lab, the Harvard Kennedy School, Mozilla and others. She lives between New Orleans and London.

Illustration by Josh Kramer, using MidJourney’s AI-powered text-to-image generator with the prompt “toxicity and harassment in language.”

For a period from the 1950s to the mid-90s, New York’s Times Square was home to scores of pornographic movie theaters, populated by men of various sexualities masturbating and cruising among each other. Detractors complained the venues drove crime by encouraging drug use and unsafe paid sexual encounters; these fears reached a fever tilt in the 80s, with the twin perils of the crack and HIV/AIDS epidemics. When, by the mid-90s, Rudy Giuliani’s pro-business administration began the full-scale shuttering of these theaters and other sites of public sex in Manhattan under the rhetoric of public safety, it was heralded by many as the long-desired cleansing of the city. Out were gay men, drug users, and the homeless, all seeking refuge in the glowing skin-flick celluloid darkness; in was Disney’s Lion King, the four-story Toys “R” Us, and a spate of new luxury development that heralded the city’s final emergence from decades of decline.

Still, if you asked science fiction writer Samuel Delaney, who frequented Times Square’s theaters for decades before their final closure in 1996, safety was almost never a concern inside the theaters. Recurrent visitors developed an understanding of the various uses the theaters offered, from a quiet, unjudgemental place to sleep, or a potential site of sexual gratification. Varying levels of desire coexisted, and regular visitors developed relationships both ephemeral and consequential, creating a field of intimacy built on subtle communication. While newcomers were more vulnerable to potential harm, even a handful of return visits would endear you to “the particularly social queens clustering in their corner of the theater,” those watchful eyes who were quick to dispense warnings like “Honey, watch out for that guy over there. He’s up to no good!” Delaney suggests that, across thousands of visits, he witnessed just one violent robbery. 

In Times Square Red, Times Square Blue, first published in 1999, Delaney makes the case for pornographic theaters as a critical social infrastructure for intimate cross-class contact, the kind of social space that both increases pleasure for its users while serving as a vital, accomodating environment for marginalized people of all stripes. Today, in our capital-flush metropolises, such places for public sexual culture have been largely obliterated. While some of their functions have shifted online, like the use of hookup apps, such digital environments offer only a pale imitation of what the theaters represented to their patrons. While the connection is not necessarily intuitive, the legacy of these theaters prompts us to ask: how we might redesign safe digital public spaces that allow us to experience life with more beauty, vulnerability, and pleasure? 

At first glance, it may seem unusual to draw comparisons between these oft-forgotten porn theaters with the social media platforms that mediate our lives today. While sexual uses of social media do exist, their relative marginalization — especially in the wake of SESTA/FOSTA, a 2018 law which criminalized various online aspects of sex work — means that such uses are far from the norm for many online. But if we can observe the marginalization of a public sexual culture in both digital and urban space as intertwined forces, in ways that punish the most marginalized and otherwise deny unexpected pleasure as a vital part of being alive, we can ask how the internet might be remade to meet the messy totality of our lives. 

Gentrification of the city, and the internet

As New York’s porn theaters were under attack, other aspects of its complex urban fabric were also coming undone. During the 1980s and 1990s, neighborhoods like Chelsea and the Lower East Side became rapidly gentrified as HIV/AIDS killed significant numbers of gay residents, opening their rent-controlled apartment units to market-rate transformation. 

Gentrification is many things, not just the economic displacement of existing tenants and businesses with higher-income residents, but also a mental and emotional shift, according to Sarah Schulman’s Gentrification of the Mind. For Schulman, a New York-based author, activist, and dedicated member of AIDS activist group ACT UP, one consequence of New York’s gentrification was the loss of avant-garde lesbian theaters and other radical artistic communities, lovingly chronicled in her early novel Girls, Visions and Everything. With high rents pricing out these kinds of heterodox uses, which created spaces to encounter people from different backgrounds, gentrification erased the possibility of living in mixed communities. 

Meaningful urban diversity exposes us to different ways of being, and as Schulman argues, “the daily affirmation that people from different experiences are real makes innovative solutions and experiments possible.” This analysis also helps us see why so many digital spaces lack the kind of texture that makes authentic living difficult. Too many digital spaces, however much they present themselves as revealing the reality of other people, create a false sense of sameness and homogeneity, replacing messy, uncertain environments that acknowledge the needs of others with something smoother and less challenging. 

Gentrification includes both racial and aesthetic homogenization, two forces that became apparent in danah boyd’s research on the shift from Myspace to Facebook in 2006 and 2007. boyd found that white high schoolers perceived Myspace as “ghetto” due to its varied visual interface, enacting a kind of digital white flight to Facebook’s cleaner user interface and elite college-adjacent network as “safer” environments. Of course, the idea that a cleaner UI and an air of exclusivity would confer users a greater degree of safety has proven false in the years since boyd’s research, and in time we’ve watched the internet become less variegated, more surveilled, and less likely to facilitate the meaningful cross-identity encounters that non-gentrified urban space has provided its residents.

Many online spaces today are widely (mis)understood as digital public environments, despite being significantly more surveilled and controlled by their own rules than real-life public spaces. (Physical “public space” is also under attack today: consider the hyper-surveillance of New York City’s Hudson Yards and its attempt to assert ownership of its visitors’ digital photographs.) In practice, social media platforms erase the variations that distinguish human beings, reinforcing certain aspects of class stratification (whether through one’s follower count or verified status, for example) while flattening other critical distinctions, as they seek to enact a streamlined model for targeted advertising. Shoshana Zuboff argues in The Age of Surveillance Capitalism that social media platforms use “dehumanized methods of evaluation that produce equivalence without equality,” an approach that “reduce[s] individuals to the lowest common denominator of sameness—an organism among organisms—despite all the vital ways in which we are not the same.” 

The social world of the porn theaters stands in stark contrast to these narrowed visions of user categorization, as theater patrons formed idiosyncratic bonds that could only emerge in a complex physical environment. Delaney describes having dozens of sexual encounters with men who struggled with homelessness and addiction, disappeared for long stretches at a time, and reappeared unexpectedly, each a meaningful part of Delaney’s life. These affinities transgressed the bounded expectations of a suburbanized heteronormative sexual life, in which physical encounters are curtailed and carried out in private. 

Likewise, digital environments could indeed create rich possibilities for novel sexual and emotional encounters, in ways that push beyond the limitations of physical geography and embodiment. But the majority of digital spaces today seek the opposite: bounding our imaginations for what we might desire from these virtual platforms.

Lesbian digital activism

If Times Square Red, Times Square Blue models the kinds of intimacy that remain possible in public spaces, then Cait McKinney’s Information Activism: A Queer History of Lesbian Media Technologies shows how marginalized communities have appropriated resources and used tools not always of their own design without losing community values. By creating lesbian indexes, newsletter networks, phone hotlines, and archives, lesbians have sustained communal life by making do with imperfect foundations, creating ad-hoc systems that allow queer experiences to proliferate. 

Information Activism shows that community norms are essential both on- and offline. While the rules governing a lesbian bar and a lesbian newsletter may differ, their creation by those who identify with the identity category “lesbian” means that negotiating expectations for both individual and collective behavior is critical. Lesbian information activists worked out these agreements in real time, and McKinney’s research finds shifting norms and expectations emerge in the margins of phone hotline call logs, with differing opinions on the inclusion of trans women—a mix of hostility, uncertainty, and a gradual recognition growing from the quotidian work of sustaining community resources. These social agreements, developed through conversation, organizing, and productive conflict, helped many women define a distinct lesbian identity, even as McKinney finds the term lesbian shifting in meaning over the course of decades. 

Despite McKinney’s playful question, “Did lesbians invent the internet?”, and ample evidence that core aspects of Black and queer digital practices have permeated web culture, marginalized people have rarely held full control over the web platforms they inhabit, leaving them vulnerable to harassment while producing billions in revenue for web companies. In this climate, lesbian information activists offer further lessons for today’s users, adopting a “good enough” approach that emphasizes tactical concessions to systems beyond individual control, without relinquishing core community values.

This is apparent in the ways that trans people have operated online, where we’ve had to make do with platforms ambivalent (at best) at our presence. In earlier formations, trans people worked around AOL’s attempts to ban trans digital spaces by using the names of community elders to find one another, a canniness that proved necessary until the ban on trans chatrooms was lifted in 1994, which resulted in the creation of the Gazebo, the first-ever dedicated and open trans chatroom. Then, in the late 2000s and early 2010s, much of this crafty use of digital spaces shifted over to Tumblr, which also created space for alternative pornography and other niche communities to thrive. 

As one research paper on the platform’s impact on the community argues, “Tumblr allowed trans users the changeability, network separation, and identity realness, along with the queer aspects of multiplicity, fluidity, and ambiguity, needed for gender transition.” Tumblr users described it as an “open space,” an environment where one could strategically express one’s emerging identities, and where “Anonymity was less about being technically anonymous, and more about being separate from the rest of one’s everyday network.” For my friend and fellow writer Sasha Geffen, the site was a space that gave them access to trans communities years before they were fully out to the world: when their photograph was shared by a popular genderqueer blog, it affirmed a core part of their identity that was still in-the-making in their IRL relationships. Trans Tumblr also connected them with others who have stayed friends and whose relationships have migrated to other digital spaces, primarily Twitter, where shifting expectations again delimit what is and is not possible.

Still, Tumblr’s emergence as a key site for many trans users was unintentional: the platform did not seek them out, and the community still faced challenges with content moderation and censorship. Despite that, Tumblr proved vital for numerous bloggers to share details about their transition processes, until a 2018 ban on adult content (enacted by new owner Yahoo!) and “female-presenting nipples” rendered it nearly unusable for these purposes. The loss has left us to once again mourn complex digital spaces, just as Schulman and Delaney lamented the parts of New York that made queer life decades ago so meaningful.

Today’s dominant tech companies are interested in minimizing user fluidity and self-determination while maximizing profit. These simplified platforms have crowded out earlier web tools that were weirder, more niche, and most importantly did not seek to encompass the totality of the web or delimit how we can encounter one another. Ironically, Yahoo’s purchase of Tumblr proved the challenges of a profit-only approach as it alienated the platform’s existing user base, severely curtailing how it could be used without generating any significant financial return. The backlash should reveal obvious lessons: we share an abundant desire to inhabit digital environments that enable productive instabilities, our identities and affinities never fixed in place, always open to negotiation, unhappy with the status quo. 

Gentrification has been a decades-long process, still ongoing in cities around the world today. By contrast, the destruction of niche online space has been carried out almost as soon as the internet came online, a speeding-up process that we can speculatively see ricocheting back into meatspace, amplifying the physical destruction of mixed communities. Reversing these trends in both contexts will require an appreciation of the virtues of living amongst those who are not like us. If we are to remake both our cities and the digital environments that have become a second home today, we must ask for more: spaces that are not perfect, or even necessarily of our own design, but still capacious enough to live as if our belonging does not require the erasure the messiness of others, and, just as importantly, ourselves. 🌳

Annie Howard is a housing organizer and freelance journalist based in Chicago. She writes about queer and trans life in cities, radical urban histories, and plenty more. Her website is annie-howard.com, and she’s on Twitter @t_annie_howard.

Illustration by Josh Kramer.

On November 23, 2021, South African scientists, while doing routine genetic sequencing of coronavirus samples, identified a new, infectious, and rapidly mutating strain of Covid-19 that PCR tests were struggling to detect. On November 26, they alerted the world to the Omicron variant.

Instead of acknowledging South Africa’s world-class disease surveillance, Western countries turned to a typical response: exclusion. Against the recommendations of the WHO, the United States, Canada, and many European countries imposed bans on travelers from southern African countries. Canada went further and stopped accepting QR-bar coded molecular Covid-19 test certificates issued in South Africa and nine other African countries, forcing travelers to get a test from a third country.

The move was part of a long-running architecture of suspicion against Africa: from medical tests to travel documents, to social media, and even web3, African people and technology are excluded from global technical infrastructure. And this has consequences that designers, engineers, and policymakers in the West should think more about. 

We can find the origins of this digital mistrust in Western colonialism. In the thinking of colonial corporations like The Dutch East India Company or The British South Africa Company, Africans had to be dispossessed of their resources so that “responsible” colonial administrators could exploit them for the benefit of “modern civilization.” Even then, Western colonial armies did not trust that Black inhabitants of Africa had the skill, intelligence, and know-how to mine the continent’s enormously rich coal, diamond, gold, or silver deposits. Colonial scientists carried out medical experiments on the bodies of Black Africans, not believing that the bodies of Africans could feel pain.

Centuries later, that mistrust remains. If someone from my country, Zimbabwe, wants to get a visa to stay in a Western country for more than a few months, we are usually required to get tested for HIV and tuberculosis and submit digital results. But Western embassies will only allow these results from a tiny handful of labs, ruling out most of the local clinics that people rely on. Then, if our visas aren’t denied and we make it to our destination, our passports are aggressively scrutinized—even though they contain chips and a scan-friendly, numerical barcode that any airport machine-reader should be able to easily authenticate.

It reminds me of my daily life. As a writer, I receive my income from publications around the world. But to get paid digitally, especially from the US, is an uphill endeavor. When billing American publications, I’m required to insert numerous digital signatures certifying that my personal wages don’t violate US Department of Treasury sanctions. But numerous Western editors whom I work with only pay via PayPal, which is almost impossible to sign up for in my country of Zimbabwe due to those sanctions. That forces me to use expensive online correspondent banks in places like Germany, which deplete my earnings due to rerouting fees.

It’s true that African nations like Nigeria have earned global notoriety for hosting internet scams that swindle Westerners. However, digital crimes don’t only originate in Africa. The UK has been named as a “global epicenter” for online financial cyber crimes. The US hosts the largest number of botnet servers in the world, which are commonly used to launch DDoS attacks or steal sensitive data. The majority of dark websites—where drugs, phishing, bots and malware, and illicit money are said to be traded—are either heavily used in the US and originating in Eastern Europe, particularly Russia. A 2019 Price Security study revealed that North America leads the world when it comes to the use of the dark web. So is Africa really the problem? 

Yet Africa’s exclusion persists. Yasin Kakande, a pan-Africa technologist, TEDx speaker, told me he felt this when sampling American dating apps as part of informal personal research on the under-explored aspects of online racism. “I was disturbed to see straight-up on some Westerners’ profiles: “I don’t accept love proposals from users in Africa…sorry,’” he said. 

It’s not just Western users who ignore Africans. Western tech platforms have excluded Africa’s internet spaces from some of their most important initiatives. For instance, Big Tech AI algorithms, which are reasonably busy flagging hate speech in the West, have a poor or non-existent capacity when it comes to incorporating African languages, and often do very little to address hate speech-language formations in Africa. Crucially, Facebook’s algorithms appear to have failed to flag online hate speech posts in Ethiopia’s two dominant languages in the lead-up to Ethiopia’s bloody civil war, according to disclosures from whistleblower Frances Haugen. 

Unfortunately, the “web3” movement hasn’t brought more inclusion, either. In the booming world of tokens and NFTs, many platforms ban blockchain miners from my country—including Zimbabwe citizens living abroad. As Michael Musekiwa, one of Zimbabwe’s most prolific NFT artists told me: “To mine NFTs on the blockchain from Zimbabwe is such agony. We usually use Virtual Private Networks to connect to American blockchain platforms like Open Sea and thus mine NFTs, but as soon as they notice that the connection is coming from Zimbabwe, they usually slow the network or throttle it.” And then, of course, there’s the emerging, troubling data that NFT buyers prefer white male avatars, and that light-skinned avatars on average fetch higher money than you know…darker ones.

There are some initiatives that give me inspiration. Home-grown movements like WikiAfrica are trying to improve African representation on the internet, by organizing efforts to add African knowledge to platforms like Wikipedia, and digitally rescue the languages that Western tech ignores. Nigeria-based publication TechCabal—Africa’s answer to WIRED—documents the grit, innovations, disappointments, and triumphs of Africa’s tech scene. And then there’s Rest of World, a global nonprofit media project that documents the internet’s effect on society outside of the Western bubble.

These efforts give Westerners—including its coders, programmers, UI designers, illustrators, and network engineers—an important opportunity to get past their suspicions of Africa and its technology, and hopefully develop a deeper understanding. Then, we might be able to talk about building a better internet—where trust is more fairly distributed, and a world where African technology is integrated at last. 🌳

Nyasha Bhobo is a freelance journalist whose focus is the interplay between technology and society beyond the Western bubbles of California or London. Her work appears in Rest of World, Newsweek, The New Arab, The Africa Report, and The Globe and Mail newspaper.

Illustration by Josh Kramer.

In May 2020, a dog owner made a post in the Facebook group for alumni of Fenzi Academy, a popular online dog training school. “I understand that this will  be an uncomfortable topic,” she wrote, but she thought the group was the best place to have an “honest and civil” discussion. The issue in question? Whiteness in the dog training world. It was days after the nationwide protests for George Floyd’s murder, and the question had been on my mind too: Dog ownership was diverse, so why was the dog training world—especially its force-free, positive reinforcement-based sub-community—so overwhelmingly white?

The racial makeup of our 12,000-plus member group was no exception. Its active users were mostly white, and so was the school’s leader, a charismatic woman named Denise Fenzi. The trainer’s post hit a nerve, and as a member of the group, I watched as the thread spiraled out of control. Some members were supportive and eager to discuss it, while others were upset and dismissive—the question seemed confrontational to them, and out of place in a group for dog training. Suddenly, the comments were locked, and soon after, the whole thread was erased. 

For many of us, that moment felt like betrayal. I joined the group after I took a few classes for my first reactive rescue dog, and I’d been fawning over Denise Fenzi and the school’s teaching philosophy. To me, the alumni Facebook was a blessing: I’d post about problems I was having with my dog, and get dozens of thoughtful responses, reassurances and solutions. It was a rare, generous community—one that had, until that conversation about racism, felt inclusive. 

The Fenzi Incident, as I’d think of it later, changed everything. It split a beloved Facebook community into two, raising complicated questions about what it meant to be BIPOC community members of a mostly white world, the cost of speaking up, and the challenges of creating online spaces that genuinely support its marginalized members. 

One of the members who had tried to respond to the thread before it was shut down was Jennifer, a Black woman, who had first joined Fenzi because she admired the trainers who were teaching in her sport, rally obedience. She started taking classes in 2014, and by 2020, she’d become a teacher’s assistant for some Fenzi classes. Until the thread on the lack of diversity got deleted, she’d felt that the group had been welcoming. With the thread gone, she sent a message to Fenzi directly. She told Fenzi that shutting off the comments was hurtful, that it took away her chance to speak up.

But Fenzi argued the thread had to close once people started getting angry. It was becoming an echo chamber, she said, and it would make the divide in the dog world worse, not better. 

For Jennifer, Fenzi’s response was crushing. It reminded her of a past experience, when a man from another majority-white dog organization that she had been part of for decades slapped her and called her the “N-word” in public, and then the group dismissed her concerns. Recounting the incident, even years later, still brings her to tears. “It’s heartbreaking because you have a group that you feel almost a familial closeness to. And in one opportunity, one instance, someone can change that very drastically, just by saying, ‘no, we can’t talk about issues that pertain to you,’’ Jennifer said. 

In the next few days, Denise Fenzi would post multiple apologies to the Fenzi group about her decisions: apologies about how she hadn’t meant to hurt people, apologies that to some BIPOC members felt more like an attempt to quickly repair things than a genuine attempt at reflection or change. Jennifer began to distance herself from the Fenzi group, and many other BIPOC and allies in the group felt the same. 

In one of the first deleted threads, Denise Fenzi had suggested that members create their own space to talk about diversity in the dog world, if they needed to. In the following days, a few Fenzi members did just that, and Fenzi eventually shared the link to the new Facebook group. It was called Inclusivity in Dog Training: a group for dog people to discuss the racism entrenched in the dog world. 

Jennifer joined the new IDT group, as did many other BIPOC, and many non-BIPOC allies (Denise Fenzi herself never joined the IDT group, but a few Fenzi instructors did). The group was much smaller than the main Fenzi group, but it grew quickly—from 200 members to over 2,000 in a few months, including some big-name trainers I recognized. The group was very active in its early days. There was a sense of collective anger, frustration, and hurt, and many BIPOC who spoke up about the problems they’d experienced, and others who shared the long, racist history inside the dog world. It was more startling and intense than I could have imagined.

I learned, for instance, that the cute and cuddly American Kennel Club, which made dog breed standards and judged the ever popular AKC show each year, had never once had a non-white Best in Show judge in its 145 year-old history. That out of the hundreds of dog-specific breed clubs in the U.S., only two had non-exclusionary clauses in their membership requirements. That the history of dog breeding is linked to eugenics (for wasn’t dog breeding about preserving lineages, selecting arbitrary ideal traits, over others?). And that modern-day dog breeders often held deeply racist beliefs—some refused to sell puppies to specific minorities, some posted racist memes. At dog sports shows, some dog owners openly disparaged BIPOC, LGBTQ and other marginalized members. 

The new volunteer moderators of the group were also figuring out how to set up a trusted, productive space where BIPOC could feel safe and heard, and where everyone else could ask difficult questions or learn. Mods like Oluademi James-Daniel, a dog trainer, had never moderated a big group before. But she found herself spending four to eight hours a day combing through posts in the group, curating, managing, trying to find the right balance of being honest and holding people accountable (if they said something racist, or didn’t know when to step back from a conversation), while still caring about the people behind the screens. She’d sit in front of her laptop with a can of cider, sorting through her alphabetized files of resources: one to explain what privilege looks like, one to explain how the innocuous things people were actually racist, how to be a genuine ally, how to step up. 

“Trying to help 2000-plus people to make that journey was a lot of work, but it’s one of the most fulfilling things that I’ve done,” Oludademi told me. Holding true to a lesson she learned from her mom: she decided to trust that there were no stupid people, just blissfully ignorant ones, so part of her work as a moderator became kindly trying to teach them how to understand.

IDT wasn’t only a place to learn and teach about racism. Outside of the IDT group on Facebook, there were no other spaces, in the real world or online, for BIPOC in the dog world—and the group became a vital source of support for them. The group also had ambitions of creating more resources for BIPOC—perhaps someday, even opening a BIPOC-led dog training school, a school by BIPOC, for BIPOC. 

The IDT group came with its own challenges, though: its non-BIPOC members still vastly outnumbered BIPOC, which meant white members still easily overtook the discussions. Jennifer continued to feel frustrated when she saw white members who posted repetitive questions, or expected to be rewarded for taking basic steps toward anti-racism. 

For the moderators of IDT, it also meant constant unpaid emotional labor, Which they had to juggle with their own dog businesses offline.The work took a toll. They soon realized they had to learn how to step away for their mental health. 

They started setting “BIPOC-only posting days”. If a non-BIPOC member posted on those days, moderators would send them a gentle but firm reminder to save their post for later (repeat offenders would get temporary or permanent bans). The days were really designed to be breaks for the mods, but also felt true to the spirit of the group: to promote and prioritize BIPOC voices above all. 

From February 2021, more days became BIPOC-only for the IDT group. The group was archived for a weekend, and then a week. The moderators made a new, small private space exclusively for BIPOC (with fewer than 100 members), where they could vent and be heard, even if it would not connect them to the broader dog world. Jennifer felt safe in the BIPOC-only group, but also felt its limits. Many of the BIPOC members in the dog world were struggling—financially, emotionally—the posts were often about their pain, a place to listen. But without non-BIPOC allies and ties to a bigger world, the small group couldn’t push for change, the way that the bigger IDT group did. 

Months passed. Some of the BIPOC group members thought about leaving the dog community, like German Shepherd breeder Jackie, a Black woman who especially enjoyed working with young Black dog handlers. But she knew that the cost of leaving the community would mean having even less representation of BIPOC, less guidance for them to enter the dog world. To stay in it, though, meant facing dog friends who wore MAGA hats, and maintaining pleasant interactions with people face-to-face while reading their racist posts online. 

The Fenzi Academy Alumni Facebook group continued on, too. Denise Fenzi declined to comment for this story, but at the end of 2020 Fenzi Academy did send an email to its community about actions it had taken to address problems of diversity, which included steps like working on a diverse moderation team for its alumni group, updating its website to reflect inclusive wording, and providing its staff with trainings for inclusive practices. 

But many BIPOC members never received the email—they had left for good. That included Jennifer, who stopped her TA work with Fenzi, and completely removed herself from the group six months after the initial incident. And although I stayed in the Facebook group, I never interacted with the group or took another Fenzi class again. 

When Denise Fenzi reached out to Jennifer a year after their first conversations to try to talk again, Jennifer did not write back. “I didn’t have anything new to say,” Jennifer said. 

Sometimes, Jennifer thinks about how different everything might have been if Denise Fenzi reacted to the very first conversation differently. What if she had acknowledged the problems in the dog training world, and opened up space for a hard conversation? What if, instead of only apologizing for her own behavior, Denise worked on rebuilding trust, and fostering a community that would genuinely feel inclusive? But that wasn’t how things went.

I, too, had to figure out a new relationship to the dog world. I stopped taking classes at Fenzi. Sometimes, I thought about signing up for agility or obedience classes in person, but then I imagined the crowd I might encounter, and what it would be like being the only BIPOC in the group. 

Like Jennifer, I also wondered if things could have been different. Was there a way to repair trust, once broken? Was there a way to build better online communities? 

Lately, in the BIPOC only IDT group, there’s been conversations about how to restart the main IDT group. “I would like to see it back up because I would like to see diversity and dog training celebrated in a central place,” Jennifer said, “And I’d like to see people that aren’t BIPOC to listen, and think.” 

The moderators are trying to figure it out—how to give it the time and resources that it needs to be safe, how to revive a group that once sprang out of betrayal into something new. 🌳

Laura Yan is a writer and teacher in Brooklyn. Her stories have appeared in over 25 print + online publications including Wired, Vogue, GQ, The Cut, Longreads, The Verge and elsewhere. She dreams of moving to a farm.

Illustration by Jasmin Bina.

Much ink has been spilled this year about digital infrastructures that might let people make decisions, transact, and play without needing to know or trust each other. Over in the febrile waters of web3, small groups plot new organizations where decisions will be automated through smart contracts. In the fustier offices of government bureaucrats around the world, people consider whether laws and regulations could be made machine readable, enabling more decisions to be automated as code becomes law. Trust in institutions and government may be low, the logic goes, but perhaps we can trust computers instead.

These experiments are part of a broader trend of people exploring how technological systems can support new models of organizing people, money, and activity. But in the process of automating complex processes, automated decision-making systems and “trustless” infrastructures alike risk making complexity invisible and recourse impossible, further entrenching existing inequities and alienating already-underrepresented people. 

In this essay we focus on automated decision-making deployed by governments, whose decisions around automation and AI have far-reaching consequences. And we know that people are concerned—because we asked them.

In 2020, we were part of a research project looking at trusted and trustworthy and automated decision-making in Aotearoa New Zealand. In the course of the project, we heard from 187 people from around the country about how automated decision-making affects their lives, how they feel about it, and what could be done to make them feel more comfortable. The research project was led by the Digital Council for Aotearoa New Zealand, an independent advisory group to the Minister for the Digital Economy and Communications, with the participatory research conducted by Toi Āria’s design research team. We, the authors of this piece, drafted the final research output, bringing together findings from the participatory research, a literature review, and a report from Māori experts into a report to the Minister with recommendations to the government.

In the past few years, there has been significant discussion in New Zealand about the role of algorithms in decision-making—particularly those deployed by government agencies—with a focus on ensuring fairness and transparency. In 2019, the government released a stocktake of operational algorithms used across agencies, and in 2021 the official statistics agency Stats NZ released the Algorithm Charter for Aotearoa New Zealand, outlining guidelines for participating government agencies in their implementation of medium and high-risk operational algorithms. 

Unlike previous work on algorithms in New Zealand, and much of the broader research on trust and automated decision-making, our project with the Digital Council prioritized hearing from folks who are often impacted by complex digital systems but rarely have agency or input into their design. We heard loud and clear from participants that they are doubtful the current approach to automating important processes will lead to more equitable outcomes, even if the automation is intended to reduce human bias or make the decisions more trustworthy. But we also came away with important insights about how institutions that use automated decision-making can improve—and it starts with including people in the process.

Relationships come first

Our research team knew we wanted to hear about trust and automated decision-making from people whose voices aren’t usually heard on the topic. Research participants included young people with experience in the care system, Māori and Pacific youth, blind and vision impaired people, and migrant and refugee women. But to begin talking about trust, we had to build trust—especially as we wanted to hear from folks who’ve had their goodwill and capacity drawn on so many times before by researchers, government and nonprofits, with little reciprocity or progress to show for it. That meant forming relationships with people from the communities we hoped to hear from, communicating transparently and consistently, and making people feel that their words would be heard, understood, and respected. 

Eventually we all made it into the room together. Workshop sessions were held in community centers, offices, and Zoom rooms around New Zealand. Each workshop focused on participants from a different group or community, and each started with a welcome, introduction, and shared kai (food). People split into small groups, and with the guidance of facilitators, discussed a set of scenarios in which algorithms played a key role in a decision-making process. The scenarios ranged from low-stakes situations like having a film suggested by a recommendation algorithm, to high-stakes situations like automated risk-assessment instruments that inform parole decisions. They largely focused on the government’s use of automated decision-making where the decisions had significant impacts on people’s lives.

You can’t automate your way to trust

For years, academics and activists including Safiya Noble, Cathy O’Neil, and Joy Buolamwini have written about the potential for algorithms and automated decision-making systems to be developed and used in ways that embed and perpetuate systemic biases and racism, or the individual biases of the people building the technology. Many of the people we heard from pinpointed these same issues, and were wary that automated decision-making systems would do nothing to help remedy the systemic bias or demeaning bureaucracy many of them often faced in interactions with the government.

Algorithms are only as good as the people who designed them. Machine learning might help with that, but right now most of the algorithms are people-designed, so people’s individual biases … can come into play.

—Blind and vision impaired workshop participant

People from poorer and minority communities are also well aware that data collected about them does not reflect the full picture of their lives and aspirations. People experiencing poverty are required to give up copious information about their living situations, spending, and family lives in order to access services like welfare and housing assistance. But when these data sets are used to inform algorithmic decision-making processes, it can feel like you’re forever defined by past challenges.

Who wants their life to be based on stink stuff from their past, that came from things from their parent’s past that they had no control over? Stop focusing algorithms on what you think is the matter with us. Instead focus them on what matters to us, the changes we want to make. Ask us, and start collecting that data.

—Whānau Ora navigators workshop participant

It didn’t come as a surprise that folks we heard from had strong and perceptive opinions on data and bias—people with lived experience of discrimination often have to also become experts in the government systems they’re required to navigate. 

As well as noting that algorithms might encode the biases of the engineers and organizations that design them, people were quick to identify that the datasets that train algorithms and inform decisions are likely to reinforce historic patterns of discrimination and selective measurement. 

While discussing a scenario where automated decision-making was used to inform parole decisions, one workshop participant said, “This is the justice system and I can’t imagine a training set that didn’t come from past decisions. … [Assessments about a person’s] risk of offending would be based on data on reoffending which is based on getting caught, getting convicted—which we already know has got a huge amount of bias in this country—so it would just self-perpetuate.” 

While algorithms and data sets are central to the function of an automated decision-making system, solving for bias in algorithms and building better data sets will not be sufficient to solve the trust problem. “It’s not whether the algorithm is testing what it’s supposed to test, it’s what they’re doing afterwards,” one young person with care experience told the team. People clearly saw algorithms and data as just one small component of a wider system of system design, governance, and organizational culture.

We shouldn't separate the system and the algorithm because, for something to work, we have to consider both. It has only ever been designed to be part of the system.

—General public workshop participant

A trusted and trustworthy digital system requires users to have trust in the organization building and maintaining that system. And when that organization is a government agency, for many people the trust just isn’t there. A significant number of the people we heard from—and especially Māori, Pacific and blind or low vision people—had very low trust in scenarios where government departments used algorithms for high-stakes decisions. 

This distrust has complex historic reasons. For Māori in particular, distrust in government decisionmaking is informed by the history and ongoing experience of colonization. In their contributing report “Māori perspectives on Trust and Automated Decision-Making,” the Te Kotahi Research Institute authors noted that “the whakapapa of distrust is rooted in a broader distrust of the systems in which ADM’s [automated decision-making systems] are embedded.” Work to build trust in government decision-making will necessarily start with work to address the root causes of this distrust. As the authors state, there is “no current incentive for Māori to trust the systems in which ADM may be employed.” 

In the course of our research, we learned from Māori experts and workshop participants that the project’s framing and key research question—which centered questions and concepts of trust—wasn’t well suited to enable the issues to be discussed from Māori perspectives. The word “trust” does not have a commensurate word in te reo Māori, nor is the Western concept of trust the key issue for Māori when it comes to navigating relationships and power. Researchers at Te Kotahi emphasized “the importance of being able to frame questions in ways that align with Māori concepts and values, allowing for discussion and debate within a te Ao Māori view and from the point of view of Māori interests”. 

This disjunct prompted us to reflect on the need to design research engagements around participants’ cultural values and ways of working, and emphasized the limitation of “trust” as a framework for understanding relationships, technology, and power.

Toward something better

So how can the state—or other organizations that provide services or spaces for a broad public—build digital systems and employ automation in a way that doesn’t further disenfranchise people? From what we heard in our research, a key to building people’s comfort with automated decision-making is summed up in the disability rights and participatory democracy rallying cry: Nothing about us without us. “We want to see Iwi, hapū, whānau involvement in creating them [systems]. Co-develop the solution,” said a Whānau Ora navigator participant.

The people directly affected need to be consulted about the criteria being written for the algorithm and definite checks and balances are needed, reviewing and monitoring them, and also that things are being created to take into account social disadvantage.

—Blind and vision impaired workshop participant

Despite strong feelings of discomfort towards many automated decision-making scenarios, most people weren’t opposed to all uses of automated decision-making, even by a government they had little reason to trust. Some decisions were considered simple or low-stakes enough that automation was appropriate, and might help speed up previously-slow processes. People could also clearly imagine how automation could be mobilized in equitable, even liberatory ways, if conditions were different and if systems were designed in a way that included them. “If you want to know an area that an algorithm could help with, find an area that actually matters to famil[ies],” one Whānau Ora navigator participant said. “Create the algorithms around that—what makes up a happy person, a happy whānau,” said another.

People also wanted to see more transparency about when automated decision-making was being used, how it worked, what data and criteria were used to inform decisions, and what part it played in the wider system. They wanted to see this transparency accompanied by clear and open communication that made space for asking questions and opportunities for recourse if something did go wrong. The ability to talk directly with a person, not just a chat bot or via one-way missives from an organization, was also seen as an essential part of building trust through community.

Algorithms work best in conjunction with relationships.

—General public workshop participant

Ultimately, we learned that digital projects and systems are likely to succeed or fail, build trust or diminish it, because of the relationships involved. 

Postscript 

The Digital Council presented their report to Minister David Clark in late 2020, and launched publicly in 2021. Although our work on the project finished over a year ago, the insights from the research have significantly shaped our thinking about trust and digital technologies and our approach to research and design more broadly. We have more trouble than ever imagining a scenario where trust in institutions could be replaced with trust in a technical system. And we have more appreciation than ever for the hard work of communities and activists who fight to be heard and included by designers, organizations, and governments who so often (despite frequent good intentions) build digital infrastructures that disregard people’s best interests or even cause harm.

Automation allows for increased speed, scale, and sometimes immutability, all of which can have huge benefits—especially for people profiting from the efficiency gains as a result. But without taking steps to build trust over time and involve users and diverse teams in design, trust in digital systems will remain unevenly distributed. Privileged people will likely trust that systems will serve them while less-privileged communities have little choice but to engage with digital infrastructures they don’t trust but which define the course of their lives. 

If governments and organizations want to have trustworthy systems that truly serve a broader public, sometimes decision-making needs to be slowed down to a human scale, with the space to make adjustments to account for the needs of people affected. Because you don’t build trust with technical systems. You build trust through relationships. 🌳

Kelly Pendergrast is writer, researcher, and sometimes art worker based in San Francisco. She is the co-founder of Antistatic, a consultancy that focuses on complex issues around technology and the environment. Her writing has appeared in Real Life magazine, e-flux, and Business Insider.

Anna Pendergrast is a writer, strategist and policy analyst based in Wellington, New Zealand. As co-founder of Antistatic, Anna helps people communicate about the complex issues they are working on in order to drive positive social and environmental change. Much of her recent work has focused on digital technology and data systems and their impacts on individuals, communities and the environment.

Illustration by Huriana Kopeke-Te Aho.

In 2020, Xinyuan Wang, a PhD student at University College London, conducted a 16-month ethnography of Chinese residents to gather their perceptions of the country’s social credit system. A typical response she captured: “I can’t wait for the implementation of the social credit system, there will be less fraud for sure.” 

Westerners might dismiss this sentiment as false enthusiasm to avoid danger from state censors. But in her analysis, Wang found that many Chinese citizens saw their nation’s proposed social credit system—currently a fragmented set of policies that aim to enforce certain regulations and incentivize good moral behavior—as a way to alleviate a lack of social trust. 

The connection of social credit to trust raises interesting questions about our own systems of algorithmic surveillance in the United States. The truth is that we have our own trust problem. According to the World Values Survey, in the early eighties, 44 percent of Americans thought their fellow citizens could be trusted most of the time, but that percentage fell to the mid-thirties by 2014. This lack of trust is particularly pronounced among partisans: according to Pew, in 1994, 21 percent of Republicans and 17 percent of Democrats had “very unfavorable” views of the other party. By 2016, those numbers had nearly tripled.

Mistrust has made us vulnerable to our own version of algorithmic control. Algorithmic surveillance aims to ensure that citizens behave in a manner consistent with a society’s core values—as interpreted by the state, in China’s case. But we can think of ours as a private credit system that regulates behavior algorithmically for the purposes of private gain rather than state control. Our version is pernicious in different ways because it folds into our individualistic notions of individual agency: we choose which Netflix shows to watch, which political Facebook post to like, which Spotify songs to play, which Amazon item to purchase. To the extent that companies use their algorithms to recommend items to us, they are presented as extensions of items we already consume. It is as if algorithms are “giving us back to us.” But this comes at a cost. 

It’s not just that we are being mined for information. I believe this algorithmic catering to our private desires only serves to make us trust each other less. We live in a self-perpetuating social cycle where the more time we spend as isolated individuals consuming algorithmically-curated culture, the more we trust algorithms over each other. If what we watch, who we listen to, and how we speak become structured by processes of data extraction, we may become a people unable to come together across differences to solve collective problems. We may begin to see politics as an exercise in personal wish fulfillment rather than as a site for collective action. 

In my class on Algorithms, Data and Society, I ask my students to reflect on why they object to China’s social credit system. “Is your objection to the fact of surveillance or is it that a state is doing the surveilling?”

I remind my students that even they may be in their seats due to algorithms. Increasingly universities are hiring vendors like Salesforce or Kira Talent to aggressively identify and market to prospective students. They run a complex web of formulas on a wide range of personal data that determine when and how students should be contacted, whether they should be admitted, how much aid should be offered, and what other strategies would get them to enroll. They even monitor student behavior to ensure they stay on course for graduation. Some universities have gone as far as tracking athletes using geolocation data to make sure they are attending class. 

These algorithmic relationships are not two-way exchanges. Increasingly, algorithms have become a tool of the powerful to reduce the costs associated with contract enforcement. In a 2015 paper, Shoshana Zuboff cites Google Chief Economist Hal Varian describing how driverless car neural networks, geolocation data, and consumer databases can work together to empty contracts of uncertainty:

If someone stops making monthly car payments, the lender can ‘instruct the vehicular monitoring system not to allow the car to be started and to signal the location where it can be picked up.’

It’s not unlike how local governments increasingly turn to policing and algorithmic surveillance, rather than community engagement, to address crime. And while most of the algorithmic encounters we have in this society do not leave us locked out of our cars, Zuboff’s example points to a possible dystopian future. If massive amounts of data and computational power are being employed to guide or constrain our decisions, then how much can we trust the authenticity of our choices, and how can we trust each other? 

Sociologist Robert Putnam’s work in the early 2000’s highlighted the damaging social and political effects of mistrust. Communities with low levels of trust engage in what he calls “hunkering” or withdrawing from social life, which, he argued, reduces their ability to engage in collective action to address problems in their communities. But it can have even more dire consequences. In The Origins of Totalitarianism, philosopher Hannah Arendt saw social isolation from one’s community as a precondition for accepting totalitarian views. 

To regain trust, we must regain control of the algorithms that govern our lives. So how do we use algorithms in ways that preserve individual autonomy and freedom? To address this dilemma, we need to understand algorithms as sites of power and find ways to restore the balance. We should assert our digital rights to not have data be used unfairly to manipulate or control us—and this must be done collectively. 

One such model might be found in the United Kingdom, where a group of Uber drivers have fought to close the balance of power with the ridehail platform by creating an organization called the App Drivers and Couriers Union and demanding access to their driver data under the EU’s new data privacy laws.

As the union’s cofounder James Farrar explained in an interview with (New_ Public Magazine Editor) Wilfred Chan for Dissent Magazine, this data would empower workers to know:

How much did I earn? How was my time utilized on the platform? What was the quality and quantity of work I was offered? And if I was fired, why was I fired?

Crucially, Farrar’s group has also sought access to information about the algorithms that control drivers. For example, how they nudge drivers to make decisions that are in Uber’s interests. The union members hope to use this data to inform legal action and labor negotiations.

How can we scale this kind of collective action? Getting citizens to play an active role in the institutions that govern them has never been easy, but we have decades of research into best practices for encouraging participatory governance. Citizen oversight boards for police departments, Chicago’s local school councils, and platform cooperatives offer models for how we could empower communities to ensure the responsible use of algorithms. And to avoid “digital nimbyism,” where a minority of loud and well-resourced people serve on these boards to protect narrow interests, an algorithm could randomly select community members to participate.

Getting private organizations to reveal their algorithms—which many firms regard as proprietary information—might be harder without legislation. As we consider crafting comprehensive data rights policies in the United States, it’s worth noting that a number of scholars have proposed ways for companies to retain their intellectual property while providing users with knowledge of algorithmic processes. Ethan Zuckerman’s idea of algorithmic audits is a promising approach to balancing the needs of companies and workers. 

Our goal should be to make algorithms into tools to help us expand—not restrict—the ways we can live our lives. In her book, The Politics of Possibility, sociologist Louise Amoore argues for designing algorithms as tools that can present decision-makers and community stakeholders with multiple possibilities for identifying and addressing social problems, rather than seeing algorithms as the solver of problems. This means agency stays with people, not machines.  

I imagine few people in the United States want a Chinese-style social credit system that enforces the state’s ideas of morality. But neither should we accept a private version that steers us away from exercising our power. If corporations use proprietary data and algorithms to optimize for problems of their choosing, workers and communities can use the same data and algorithms to optimize for our own goals. But this can’t be done individually. Whether through labor unions, data trusts, citizen councils, or platform cooperatives, only collective action can counter algorithmic control—and teach us to trust one another again. 🌳

José Marichal is a professor of political science at California Lutheran University. He studies the role that social media plays in restructuring political behavior and institutions. Currently, he is working on a book that looks at how algorithms impact political identity.

Illustration by Josh Kramer.

As Elon Musk made his bid for Twitter, I reached out to Silicon Valley’s Representative Ro Khanna, one of Congress’ most visionary technologists. I wanted to talk about his forward-looking new book, “Dignity in a Digital Age”—which lays out an argument for how to create an economy and internet built on human dignity and community rather than warped commercial imperatives. 

In it, Khanna describes his aspirations to build a “digital public sphere, with a plurality of speech communities.” He calls for the development of platforms, “public and private, that do not promote lies or divisiveness but facts and constructive discussion.” At stake, he argues, is trust in our democracy itself. We, of course, agree!

We spoke about the line between public and private platforms and the responsibilities of both, the importance of place and community, and what a future internet might offer people that our current one does not. 

The following interview has been lightly edited for clarity. 

—Eli Pariser, Co-director, New_ Public

Eli Pariser: I have to ask about Elon Musk and what it means that the world’s wealthiest person is buying Twitter. What does that say about the prospects for digital dignity?

Rep. Ro Khanna: The very fact that we are so concerned over a change in ownership of Twitter, and the impact that that may have on the digital public sphere, highlights the complete lack of regulatory oversight and the lack of any set of ethical norms that have been established for social media.

When Jeff Bezos bought the Washington Post, there were obviously rumblings—is The Washington Post going to be biased? And on the deepest levels, if you read Chomsky’s Manufactured Consent I’m sure there’s some systemic bias that media may have towards corporate interests. But no one thought on a practical basis that Jeff Bezos was really going to change the character of The Washington Post and be making day-to-day decisions. 

And that’s because there’s a regulatory framework governing newspapers and there are strict journalistic norms, and it was made very clear that there was going to be separate corporate governance from editorial decision-making at the Post.

So one question is, is Musk really going to run Twitter day-to-day, or is he going to do a more prudent thing and have Twitter be separate? The fact that we’re even having to ask that question shows the problem: unlike in the newspaper business and in the broadcast business, social media has no separate sphere that’s divorced from profit maximization.

I’m always interested in this question. What we’re trying to do at New_ Public is make the case for and build nonprofit digital spaces for exactly this reason. But I’m curious about where you would draw the line between broadening the focus on public interest in private corporations and actually building public digital infrastructure. 

We need both. And the reason we need both, I think, stems from a philosophical basis of the public sphere. If you look at (German philosopher Jürgen) Habermas’ ideal speech community, everyone has to be equals, and everyone has to have good intentions. This is the sense of an ideal public speech that then at its best leads to moral truths and legal legitimacy.

But he realizes in his later works, Between Facts and Norms, that democracy is messy and you’re not going to have ideal speech conditions. So what he and many others argue is: what you really need is a lot of different discursive spaces.

I would argue that what you need is a mix of the public and the private. You need more digital public spheres—both local governments could create it; nonprofits could create it.

The federal government could create a space. It could be like a Nextdoor, but a little bit more robust in terms of engagement in a local community. The best model may be what Audrey Tang is doing in Taiwan, where you get citizen engagement locally; that’s a case of individuals actively participating in solving legislative dilemmas and finding common ground.

The thing people say to me sometimes is, “You’re starting in one of the hardest countries to make the case for publicness. This is an easier project in Europe than it is in America.” How can we make the case for publicness in a culture that’s very used to private companies?

Well, we need to make two cases. 

We need to make a case where even private companies have a sphere of public obligation and public interest. Right now social media doesn’t have that, and that is a glaring weakness because it has such an effect on democracy.

Beyond that, we have to call for true public forums like PBS, but I think a better case can be made for those if they’re seen as active engagement in citizenship. 

Why is it that we don’t even know what Congress people are voting on? If, let’s say, we put congressional bills online, and allowed people to comment on that, or participate in that—I think that would catch on, and people would say, yeah, that’s a public function.

In your book, you started by talking about the idea of geographic place and why it matters. Why do you think it’s hard for Silicon Valley platform people to remember that different places have different characters and qualities?

Well, I don’t think it’s just hard for Silicon Valley. I think it’s been hard for our country over the last 30, 40 years. 

We said, globalization is happening. Don’t move to the jobs; new industry is coming. Let the markets figure this out. And we underappreciated people’s attachment to place. We underappreciated the necessity of community. I think now we’re suffering the consequence of that with a lot of communities destroyed—many of them now feeling they don’t have a place in a modern economy, and don’t have a sense of identity.

So what we need to do is rethink this sense of unfettered capitalism in an age of globalization and say, no, we still have to value community, we still have to value place, and we need to have a capitalist system that allows for enough state intervention, that helps places survive and thrive. 

I hear a parallel between that way of thinking and your vision for a “plurality of speech communities” online. It feels like a similar argument where instead of one monolithic totalizing system, we’ve got lots of different sorts of special purpose communities. Am I reading that right? 

That’s a great analogy. One of the advantages of a plurality of places is the sense that if America is going to be a composite nation—what Frederick Douglass said is sort of a composite of all different cultures at any given time, with all of us working in a democratic way to create it—that it works better if there are also spaces in such a democracy to allow for thicker versions of tradition and culture to exist that may not be predominant in the national culture.

For a multiracial multi-ethnic democracy to thrive, my intuition is that having different places with different customs and traditions makes it easier for people to also embrace a national culture, even if the national culture is not quite exactly what they want. Because they can have their locality to live their daily life and hold on to some of their traditions. 

Let’s say the internet that you want—and I think that I want—comes to pass. What do you think that offers people that the current internet doesn’t? 

It would be more empowering.

People right now are disconnected from the decision making of government. And the idea of a public sphere is not just conversation. Habermas said this is critical for democracy: to have informal channels that can actually shape government action. 

Currently, I would argue that’s not really the case with digital forums. Yeah, you can spout off, but there’s not this sense of an empowered link where people feel like they’re really shaping congressional legislation, they’re really shaping the president’s action.

One of the points I make is that the amount of time people spend on Twitter and Facebook and like and share and comment is partly a function of not having any other way of having a voice in government, a sense of disaffection.

So what we have to do, in my view, is to make the participation of the digital sphere more active, more empowering. There are many ways you could do that, but I’m thinking of a site that’s a bit more formal and inclusive of everyone, discussing issues affecting the community. 

I imagine there may be models like that around the country that I’m not aware of, where local towns or communities are doing that. And I think that could be very interesting. 🌳

Eli Pariser is an author, activist, and entrepreneur focused on how to make technology and media serve democracy. He helped lead MoveOn.org, co-founded Avaaz.org (now the world’s largest citizen’s organization), wrote the 2011 bestseller The Filter Bubble, and co-founded Upworthy. He is currently Co-director of New_ Public at the National Conference on Citizenship.

Ro Khanna represents Silicon Valley in Congress. He has taught economics at Stanford, served as Deputy Assistant Secretary of Commerce in the Obama Administration, and represented tech companies and startups in private practice. He is the author of Dignity in a Digital Age: Making Tech Work for All of Us. He enjoys spending time with his wife and two children in Washington, DC, and Fremont, California.

Design by Josh Kramer.

Photo via the Office of Congressman Ro Khanna.

Despite the techno-utopian myth that most citizens trust the digital infrastructure underpinning our modern world, the recent data abuses of Big Tech have fueled newfound skepticism toward these digital innovations. At Siegel Family Endowment, we recognize that technology has reshaped foundational elements that underpin civil society. As such, we focus on understanding and shaping the impact of technology on society, and support work that advances its use for outsized public good. As Executive Director of Siegel, I’m a firm believer that technology has the potential to better our world, but think it’s naive to expect that we can use these systems without thoughtfully and carefully working to build trust in them first. In order to create digital infrastructure that works and makes a difference, we cannot take trust for granted. 

Trust in digital infrastructure is not a given… and it’s certainly not free.

Technology doesn’t exist in a vacuum where everyone blindly accepts it. Instead, it operates in a complicated and nuanced world in which people’s lived experiences give them reason to be wary of new inventions. If a person is hesitant about a digital tool, it’s likely they simply won’t use it—making community trust just about as important to a system’s success as the code it runs on or the funds it needs to grow. We can’t come close to harnessing the full potential of these systems to better our world if they aren’t accessible—through actual or perceived inequities—to all.

This became clear when it came to deploying contact tracing apps during the pandemic. While a crucial tool for containing the coronavirus early on, the apps were met with significant hesitation across the United States. While the focus was on the apps, the underlying concern was the digital infrastructure on which they were built. Could downloading them open up personal data to further surveillance from the government? Could the information be used to disqualify someone from a job, a place to live or something else? While contact tracing apps are tremendously innovative solution, they only work if people trust them. 

This hesitancy didn’t come out of left field. People have legitimate reasons to be wary of handing over their personal information to what may feel like a black hole—especially those who have been previously betrayed by authorities. For example, Black communities were disproportionately impacted by COVID-19, and yet have expressed skepticism over contact tracing and in some places, vaccine hesitancy. Considering this in the context of history, it should not necessarily come as a surprise. Black individuals have long been marginalized by data, victimized by medical institutions and profiled by the government. From slavery, (one of our nation’s earliest form of infrastructure in many ways), to the Tuskegee Syphilis Study, which both took advantage of Black bodies to profit the medical profession, communities of color know all too well about the potential for exploitation. We have to recognize that for some, they are viewing these interventions through the lens of generational trauma and wrongdoing. 

Likewise, trust in digital infrastructure shouldn’t be presumed.

When we each approach these systems, we don’t do so with a blank slate. Instead, we bring our unique identities with us and our values, beliefs and backstories shape how we understand and respond to the technology, even when it’s built to help us. Even with my experience working in and around the tech industry for more than a decade, I found myself taking a pause when I was signing up for my first vaccine dose through a city-run website, double checking the security of the web address and the fine print about how my information, including demographic data, would be shared and used. 

Going forward, we must prioritize centering trust as a crucial building block when it comes to digital infrastructure.

Today’s emerging technologies have the capacity to launch our societies leap years ahead, but it’s evident—even in these systems’ earliest stages—that trust will be difficult to earn. Advances in biotech and precision medicine, including the COVID-19 vaccine, have the potential to prolong and save millions of lives, but have still been met with intense skepticism. The metaverse and related technologies could serve to change the way we collaborate, democratize experiences, and create powerful new ways to connect, but suspicion about the extractive behavior of the companies leading the charge, a lack of regulation around data and privacy, and failures to safeguard against the worst behavior could potentially derail these benefits. We can’t achieve an equitable future, powered by these cutting-edge tools, networks, and systems, until we invest in building trust from the ground up. 

How do we do this? Those at the helm of infrastructure projects can reach out to communities who may be hesitant to new systems, listen to their fears, and bring them in on the design and implementation. And when it comes to marginalized populations, we’ve admittedly got a long way to go. It won’t be easy, but we must continue and double down on our efforts to bring overlooked and underrepresented voices to the room where infrastructure is being built to ensure it accounts for diverse perspectives. We also must listen and understand where the hesitations lie in order to address them in a substantial manner that lays the groundwork on which to establish trust across communities. 

As we stand on the dawn of a historic opportunity to modernize America’s infrastructure systems following a $1 trillion investment from the Biden administration, let’s not take trust for granted. While reimagining the networks that connect us all, it’s crucial we also take this moment to double down on building trust in these innovations to ensure their impact is felt across all communities, for generations to come. 🌳

Katy Knight is the Executive Director and President of Siegel Family Endowment, a foundation focused on understanding and shaping the impact of technology on society. She draws on a diversity of professional experience from previous roles in education, technology, and community-based organizations.

Design by Josh Kramer.

Photo via Maryland GovPics on Flickr, with an Attribution Creative Commons license.

In a short commentary posted to her website in 2004, science fiction author Ursula K. Le Guin wrote that the concept of technology is “consistently misused to mean only the enormously complex and specialized technologies of the past few decades, supported by massive exploitation both of natural and human resources.” Rather, she argued, technology should be understood more expansively, as the way “society copes with physical reality… the active human interface with the material world.” Technology, as the sum of “what we can learn to do.”

This magazine is about technology. It builds on Le Guin’s formulation by asking how society copes with digital reality, what happens at the interface between humans and the digital world. It does not separate the internet from IRL. It understands that there are serious problems with the complex, specialized digital technology that we use, that cannot be solved simply by more complex, specialized technology. It urges us to re-examine our own agency in our relationship to the digital world, and think about what we can still learn to do. 

The theme of our first issue is “decentralization.” I am drawn to this word because it contains social and cultural concerns as much as engineering ones. Decentralization in contemporary “tech” discourse tends to reference protological innovations (such as blockchain, peer-to-peer software) that offer alternatives to “centralized authorities” (the banking system, Silicon Valley platforms). But it also refers to an ongoing push and pull throughout human history between rebels and rulers; the way power is distributed, accumulated, transformed, and dispersed at every layer of society. I think these continuities offer us a useful starting point to speculate about what comes next. They are a reminder that we already understand what’s at stake. 

We hope you will read the following stories as meditations and quietly radical interventions, suggesting the ways in which technological revolutions are hidden in plain sight, and often already within our grasp. We are grateful for your attention. 

Wilfred Chan
Editor, New_ Public Magazine

Burbank, at golden hour. I’m driving the 134 freeway, alongside a sliver of Griffith Park tempering city and wilderness. In the distance—through not a particularly bad haze today, after a spring rain—sits the ridgeline of the San Gabriel Mountains. Over the concrete channel of the LA River, the freeway climbs and curls away on itself, giving glimpse briefly to a median planted with ragged eucalyptus trees, all leaning hard, surrounded by snack bags, soda bottles, and those plastic snap-top containers they use to sell weed. It must be the saddest forest in the world.

In a healthy forest, the roots of these trees would be mingling with helpful fungi to transmit water and nutrients to one another, in a mutualistic network scientists call the “wood wide web.” At 70 MPH, I wonder if these Los Angeles eucalyptuses are still able to connect to one another under the roar of the freeway. Surrounded by asphalt, do they bind together, pooling meager resources, surviving one day at a time? I wonder how it feels at the edge, where the forest ends, and their messages have nowhere left to go.

The wood wide web is ancient, but it’s new to us. We owe much of our knowledge to Suzanne Simard, a Canadian ecologist who has spent her career revealing the cooperative nature of forests. In field experiments beginning in the 1980s, Simard traced the ways roots and symbiotic mycorrhizal fungi commingled beneath the surface of the soil in the old-growth forests of British Columbia. Enmeshed in a fungal embrace, she discovered, trees communicate, sending chemical warning signals to one another and passing sugar, water, carbon, nitrogen, and phosphorus between species. Wild forests operate, Simard wrote, as “an intelligent system, perceptive and responsive.”

Nature gave Simard’s research front cover-treatment in August 1997; it’s the journal’s editors who, inspired by the development of the World Wide Web, dubbed her discovery the “wood wide web.” The term was apt, and as the web itself evolved, the metaphor of forest-as-network took root. More recent research into the wood wide web’s particularities—along with a broader popular interest in the planetary and mind-altering implications of fungi—has brought the term into wider use than ever. 

The wood wide web has been a powerhouse metaphor for popularizing the mutualistic relationships of healthy forests. But like a struggling forest, the web is no longer healthy. It has been wounded and depleted in the pursuit of profit. Going online today is not an invigorating walk through a green woodland—it’s rush-hour traffic alongside a freeway median of diseased trees, littered with the detritus of late capitalism. If we want to repair this damage, we must look to the wisdom of the forest and listen to ecologists like Simard when they tell us just how sustainable, interdependent, life-giving systems work.

Ecosystems are like human societies: they’re built on relationships. “Our success in coevolution—our success as a productive society—is only as good as the strength of these bonds with other individuals and species,” Simard wrote. “Out of the resulting adaptation and evolution emerge behaviors that help us survive, grow, and thrive.”

The life of a forest is many lives, entwined. It’s also intergenerational. The eldest trees—having survived storms, droughts, ravenous insects, and the damage wrought by capitalism and colonialism—serve as the central hubs of the wood wide web. They are the strongest, the most resource-rich, with taproots stretching far beneath the earth. Suzanne Simard calls these elders “Mother Trees,” as do many indigenous people.

“Mother,” here, is also a verb, because that’s what these trees do. In drought, they share water. They pump carbon to the seedlings in their shade, compensating for the light their mighty canopies prevent from trickling to the forest floor. When they die, Simard has shown, they give all of themselves, pulsing every last bit of carbon to their kin. 

Mother Trees serve as hubs in a decentralized network. A single Mother can be connected to hundreds of other trees, but in a healthy forest, multiple Mother Trees with overlapping connections ensure that a single elder isn’t responsible for the continuity of the forest as a collective organism. Simultaneously, multiple species of mycorrhizal fungi—sometimes dozens at once—plug into the network with their own specialized offerings, swapping carbon-rich sugars for nutrients foraged from the soil. Thanks to these redundancies, the wood wide web is remarkably resilient; only a clear-cut can destroy it. 

If a single Mother Tree dies, the network adapts. If they are all felled, the forest loses precious genetic information—the instructions for resilience and survival contained within each Mother Tree seed—and replacement plantings are forced to run without a backup. New seedlings will not get extra nutrients when they need them most, nor receive sips of water in a drought. As the climate changes in ways that stress and endanger young trees, these things could mean the difference between life and death.

The web isn’t what it used to be. When the editors of Nature compared mycorrhizal fungi to a computer network, the web was still predominantly peer-to-peer, its users sharing their thoughts on personal home pages and homespun message-boards. Online advertising was in its infancy. But as the web has centralized, it has strayed further and further from the ideal presented by the wood wide web. 

In her recently-published memoir, Finding the Mother Tree, Suzanne Simard described “free-to-grow” plantations in Canada—forests that loggers strip of native plants like birch, alder and willow in order to make space for high-value timber crops. Such clear-cutting emerges from the misguided belief that growth is a zero-sum game. As a young forester, Simard saw firsthand how acutely seedlings suffered in a vacuum. Her experiments revealed why: birch trees that appeared to hog ground water in the summer dropped their nitrogen-rich leaves in the fall, replenishing the soil and feeding young conifers. The birch weren’t competing with the valuable conifers; they were cooperating. The forest was strongest as a community. 

By selecting for the most inflammatory and emotional content, Big Tech has algorithmically weeded the forest into a field of commercial timber.

Suzanne Simard’s early field experiments with birch and fir showed, too, that although firs grew initially well in a bare-earth clear cut, they suffered from lack of water and nitrogen when their collaborative relationship with neighboring birches was severed beneath the soil. After a few years, any gains they might have made from increased exposure to direct sunlight were obliterated. In the forest, lowered biodiversity doesn’t give high-value crops an advantage: it ultimately reduces productivity, invites rot, pests, and disease, and amplifies the risk and spread of wildfire.

If we take the metaphor of the wood wide web seriously, it’s hard not to see an analogy here to the context collapse endemic to social media. A few corporations control the lion’s share of public cloud infrastructure, and monopolistic ISPs exploit everyday users. Tech and social media giants have clear-cut the web, privileging high-value crops—viral content, controversy, and clickbait—over a healthier ecosystem of people, opinions, and perspectives. By selecting for the most inflammatory and emotional content, Big Tech has algorithmically weeded the forest into a field of commercial timber. As users, we’re incentivized to chase the quixotic, highly ephemeral celebrity algorithmically meted out by the platforms. Having attained it—through a viral tweet or TikTok dance—individuals, like trees, might initially flourish. But when we sever people from their context and thrust them into the glaring exposure of the sun, bad things happen. The trolls, like mountain pine beetles, proliferate, digging under the bark. Controversy sparks like wildfire, scorching the earth. Most of all, it’s lonely in the clear cut, where there are no teachers, no friends, only consumers.

“We emphasize domination and competition in the management of trees in forests… We emphasize factions instead of coalitions,” wrote Simard. It applies to the web, too. Social media companies segment and isolate us in order to package us up in tidy tranches for advertisers, just as big timber clear-cuts healthy, complex ecosystems and replaces them with sterile plantations designed to yield the maximum profit.  These practices privilege factions over coalitions too, over the cooperative, interdependent relationships that bind healthy systems and societies. They ignore the forest for the trees.

The project of decentralizing the web is vast, and only just beginning. It means finding a way to uproot our expression and communication from the walled gardens of tech platforms, and finding novel ways to distribute the responsibilities of infrastructure across a collective network. But we needn’t start from nothing.

To build resilient decentralized networks, let us create “Mother nodes”—sites in the network bearing a responsibility of care. We’ve built institutions like these before: consider public libraries, which serve both as bearers of cultural memory and as generous sources of nutrients for our minds and communities. As Joanne McNeil wrote in Lurking, her excellent people’s history of the internet, “librarians are what the internet is aching for—people on task to care about the past, with respect to the past and also to what it shall bequeath to the future.” Can we reimagine libraries for the digital age?

The forests that Suzanne Simard has spent her life studying show us how the internet could be: a mutualistic entanglement of platforms and users, in which resources are distributed according to need. A network where nobody is left to fend for themselves in the clear-cut, or expected to feed a never-ending desire for content. A place where elders with thick barks and deep taproots weather the wildfires, connecting us all to history, bringing continuity to our communities and preventing us from repeating old mistakes. 🌳

Claire L. Evans is a writer and musician. She is the singer and coauthor of the Grammy-nominated pop group YACHT, the founding editor of Terraform, VICE’s science-fiction vertical, and the author of Broad Band: The Untold Story of the Women who Made the Internet (Penguin Random House). She lives in Los Angeles.

Illustrations by Josh Kramer.

Editor’s note: This story contains descriptions of sexual coercion.

When Katherine and her partner Eric decided they wanted to have a baby, they knew that their route to conception would be complicated. Katherine already had three kids from a previous relationship, and so knew she was able to conceive through the traditional means, but Eric was transitioning trans masculine. 

Their income from manual jobs in a rural US state wasn’t enough to foot the bill for an expensive sperm bank. Semen costs a lot—over $1000 per vial for top-shelf—and rarely does one round do the job. Plus, there’s the cost of artificial insemination techniques such as IVF, the cost to keep the specimens on ice, and payments to secure any backup vials to conceive future siblings. Without a friend willing to give up the goods, Katherine and Eric had only one place to go: the internet.

They were far from alone. For years, increasing numbers of people have sourced privately donated sperm from a decentralized network of websites and social media groups around the world. Many of these people are excluded from traditional fertility pathways because they are gay, single, POC, old, or poor. Others choose this path because they want to see the person contributing half of their child’s DNA. Some reports suggest that the number of people going online for sperm has risen dramatically during the pandemic, as donations to cryobanks have fallen because of lockdown-related restrictions.

The US fertility industry is regulated by three governmental agencies: the Centers for Disease Control and Prevention (CDC), which collects and publishes data produced by researchers and clinicians on fertility procedures, the Food and Drug Administration (FDA), which controls the approval and use of drugs and testing of biological and technical products, and the Centers for Medicare and Medicaid Services (CMS), which oversees lab testing. This regulated, hyper-medicalized process ultimately presents recipients a choice of donors with squeaky clean, anonymized sperm whose fact sheets are written up like dating profiles. 

Outside of the state-vetted system, so many things can go wrong: STDs, genetic disorders, harassment, sexual abuse. Without official oversight, donors and recipients have had to devise their own systems to stay safe. There is no single solution—but what has become clear in these decentralized digital marketplaces is that accountability and trust between the people using, operating, and moderating them makes a huge difference.

Online sperm exchanges are not new. Some long-in-the-tooth donors recall dipping their toes into the marketplace in 2008, when the only game around was Craigslist. From what they have told me, it was a distinctly unsavory option. There were more people looking for no-frills sex for themselves than hoping to create a human being. Other prominent networks began to emerge as early as 2009, finding a welcome audience in existing informal networks of lesbian couples and single mothers-by-choice. In the US, people seeking donors began to coalesce around a site called the Known Donor Registry, or KDR, launched in 2010. 

KDR was founded by donor recipient Beth Gardiner. It was part marketplace, part gathering place: there were ads by people seeking and providing sperm, and they exchanged samples on mutually agreed upon terms. The private exchanges were not governed by health-related regulations from any jurisdiction, nor were there specially-trained mediators to help with the process of conception. As Gardiner once said, “If it’s legal to go to a bar, get drunk, and sleep with a random stranger, then it can’t possibly be illegal to provide clean, healthy sperm in a cup.”

Any woman who’s been on the internet for minutes knows that it’s full of creeps. The online sperm donation networks are, unsurprisingly, a magnet to them. According to one analysis from 2015, one in two women who seek sperm online through sperm donation websites or social networks experience some form of abuse, from harassment to sexual assault. This makes the design of these digital spaces—and the people they admit—incredibly important. 

Drew is a long-time donor. About a decade ago, he realized he was aromantic and wasn’t going to have a traditional family, which bothered him greatly. So giving away sperm, to people who’d let him be Uncle Drew, seemed like a logical thing to do. He began casting around for recipients, and the first place he landed was Craigslist. “I very quickly realized it was a terrible idea,” he told me. “I found a lot of people looking for sex. I was not interested in that.” He started Googling and found three or four fledgling communities for sperm donors. “None of them were very sophisticated. Things changed most significantly when KDR first came into existence.”

KDR was not only free of charge—attracting roughly equal numbers of donors and recipients, according to Drew—it also had one of the first known reviewing systems of online sperm donation websites. But beyond its matchmaking services, its most significant feature was a magnifier of social capital: the Off-Topic forum. 

Message boards and chatrooms like these are what make the internet feel like a place rather than space; in an ocean of information exchanges, it is here that trust—the only true currency of online interaction—develops and grows through regular interaction, shared vulnerabilities and reciprocated exchange. 

Solving the problem of matching sperm with egg isn’t just about health and fertility science. It’s also about protecting people by building a system that generates trust and accountability.

Without trust, you’re just a faceless person offering something another faceless person wants. With trust, you’re Bob, who’s vouched for by Katie and Trish, who’ve given Mary and Heloise great advice on fertility cycles. Bob didn’t donate to Katie, but he did tell her about a great movie he watched the other night and thought she’d like. He didn’t donate to Trish either, but shared some questions she might want to ask potential donors that would weed out the creeps. Bob donated to Mary, who told Heloise that he didn’t hassle her for sex. 

Over time, Drew watched the interactions of other donors and began building a list of the trustworthy ones. He also kept track of some of the problem donors. “There were some who were more interested in their own motivations, and I felt were being dishonest and manipulative, so I started advising recipients more.” Soon, those problem donors either left because their business dried up, or were kicked out because they got bad reviews.

Now, this is the internet; if you lose ground in one place, you can set up shop somewhere else. And that’s what Drew saw these KDR problem donors start to do by setting up their own Facebook groups. But the critical mass was still with KDR, so they didn’t get much traction. That is, until, one day KDR’s chat function stopped working due to what Drew assumed were “technical reasons”. Beth Gardiner wasn’t able to fix it immediately (Gardiner declined to be interviewed for this story). And so KDR’s unique sales point—the community that kept an eye out for one another—degraded, and its userbase, too, drifted to Facebook. 

It’s impossibly easy for a donor to create a Facebook group advertising himself in which he has absolute authority to exclude anyone or anything critical of him. There’s no easy way of knowing who’s behind each group, because you can’t look inside unless you’ve been accepted as a member. “It made it impossible for me to keep my running list of good and bad donors and recipients,” said Drew. “I couldn’t help to police, so to speak, because everything was happening in isolated and private conversations on Facebook.” Whereas donor ads and recipient requests were public on the KDR marketplace, they’re entirely controlled by group owners on Facebook. Drew could no longer see who, if anyone, might need help.

The accountability system in this network resembles international diplomacy more than a decentralized, community-led commune. Owners and moderators of the Facebook groups create alliances and share their blacklists, but this has created factions in the marketplace, and at times, all-out wars between coalitions of donors claiming their rivals are abusive. Users have learned to keep screenshots of conversations as evidence for any future disputes. Health misinformation is rife. Drew has been trying to regain his position as a trusted source; although he doesn’t advertise himself as a donor anymore, he has aligned himself with one Facebook marketplace, Sperm Donation USA, and maintains another, the Joe the Donor and Others Must Go group. The latter is his way of gathering warnings from recipients who have had bad experiences with a self-styled mega-donor, who has grabbed media attention by claiming to have attempted over 800 inseminations resulting in over 150 children. It’s the only way Drew can try to do what he used to do: keep recipients safe.

Communities require design specific to their needs. Solving the problem of matching sperm with egg isn’t just about health and fertility science. It’s also about protecting people by building a system that generates trust and accountability. But the internet is fluid, iterative and ruthless. Something like Facebook, promoted as a one-size-fits-all solution, is often not—but it’s difficult to interrupt the path of least resistance.

If sperm donation tells us anything about the internet, it’s that these enigmatic, squishy human outcomes are best developed in open, high-quality public spaces, with people who will own them. “A real site needs a moderator who isn’t a donor, and is willing to dedicate an extensive amount of time and energy maintaining it,” said Drew. “Sadly, it seems unlikely that anyone is going to dedicate the time, effort, and sweat into making something superior, to be able to beat Facebook Groups.” And other than the platform enacting a blanket ban on all sperm donor pages, he doesn’t see a clear step forward toward creating a safer, more cohesive community.

That leaves Katherine and Eric sailing into uncharted waters. They have only ever known Facebook. They’re not interested in any other route, nor do they have many options. And they’ve done as much research as they can. They’ve found a donor who matches their criteria—no regular contact with the future child, but with the option to meet after they are 18; no mental health issues in the donor’s family; looks like Eric; no STDs. Artificial insemination only. They like that he uses his real name, and shares trackable information. So far, they’ve had three unsuccessful attempts using samples shipped through the mail. I met them on their fourth round, when they came to New York to meet him in person. They were debating for the hundredth time what they’d do if their donor demanded to swap methods, from artificial insemination to intercourse. “I’ve had this conversation with Eric before,” said Katherine. “I’m like, you know, seriously, I’m just going to be laying there. The only reason I suggested it is because it could be done and over with.” Eric had always been more resistant to the idea, but with Katherine’s ovulation window soon closing for another month, “I’d probably be just like, fuck it, go for it,” he said. 

Katherine and Eric don’t need a saint; they just need some sperm. This stranger on the internet could give them the family they’ve always imagined. The stakes are high, and people make compromises. 🌳

Dr. Aleks Krotoski is a social psychologist, journalist, and New_Public contributing editor. Her beat is where our human boundaries bump up against what technology can do, and how we make sense of why.

Illustration by Josh Kramer.

A scholar, artist and hacker, Amelia Winger-Bearskin marries lessons from indigenous cultures with hopes for a human-centered, values-driven digital future. 

I first met Amelia when she gave a presentation in 2019 on how Iroquois wampum—a code woven into beads that acts both as currency and contract—prefigured both cryptocurrencies and the US Constitution. Since then, I’ve followed her vital podcast wampum.codes, which features convivial interviews with indigenous technologists about how they’re building digital art, tech, and spaces reflective of Native values.

After she published a fascinating exploration of the history and future of decentralized storytelling, I sat down with Amelia to understand what decentralized storytelling is, why it matters, what virtual worlds can learn from native traditions, and how it might shape our future. 

—Eli Pariser, Co-director, New_ Public

Eli Pariser: So let’s start with the basics: What is decentralized storytelling? 

Amelia Winger-Bearskin: I grew up as a young hacker in the ’90s, a 12-year-old girl with other 12-year-old girls who were hacktivists. So the dreams of decentralized storytelling, for me, started in those early days when I was first learning about the internet—when I’d go to school, and only one or two people at my whole school were online. And we would have these other lives, not just with our local friends but with people all over the world, some of whom we knew nothing about except their username.

But I’m Seneca-Cayuga Nation of Oklahoma, Deer Clan, which is a member of the six-nation Haudenosaunee—other people know us as the Iroquois. And we have a storytelling tradition that I believe is also decentralized storytelling.

My mom is a traditional storyteller from our tribe. Being a storyteller for the Seneca-Cayuga Nation is something that is a cross between being a historian, being a performer, being a creative writer, and being a leader within the culture.

So you need to be given the stories from elders and have the reputation to be able to receive them. But the understanding is that you’re not going to just tell the story the exact same way that the elder who gave it to you would. You’re supposed to add your own creativity, you’re supposed to add your own perspective, and you’re supposed to make it relevant to the generation that you’re telling it to. You’re an interpreter, but you’re also a repository of information. 

The way we tell our stories, key aspects of the stories are embedded within the landscape: “this is a story of Bear Mountain,” or embedded in handicrafts, like “this is the story of the cornhusk doll,” or the stories are actually told through song, or through design, or through imagery within basket weaving, or raised beadwork. 

Telling stories across multiple media is a strategy to make sure that your ideas become like a rhizome—they become sticky, and accessible to many generations. You have to make sure you can distill your information into a clear abstraction, and then embed it in as many ways as you can. So, if you’re a young child, and you only have a short amount of time with a storyteller, who may only pass through your village for a short period of time, and that storyteller tells you the story of a mountain that you see every day as you walk along and have adventures with your friends, that story will remain in your memory better. And that story may have scientific information around planting, around seasons, might have information around the stars—the scientific advances and discoveries that people generations before you have made.

Why is decentralized storytelling important for all of us, as a way of doing things, as a practice?

I think it’s important in two ways. 

One, it’s about future-proofing some of our core values for the seven generations to come. If we want to leave some of the insights from our ancestors that we are responsible for holding for future generations—I think it’s smart for us to build them into a decentralized format. That way we have them distributed in ways that are not just left to a monolithic interpreter, which is assuming that people in the future will understand what they need to understand through books, or films, or whatever is the archive that we imagine will be the storytellers to the future. 

And the second reason is, this new generation already is very fluent in decentralized storytelling, and they want to participate in stories, game worlds, and creative narratives more than they want to be told what to discuss through centralized media formats. So if you’re interested in engaging with that generation, I think it is really necessary to understand the core aspects of decentralized storytelling. 

I think we can learn a lot from young people around ways in which they use spaces and break spaces. They have the constraints that most young people have, which is they have tons of time, and they don’t have a lot of money, so they end up reusing outdated technology in really innovative and creative ways as well.

Decentralized storytelling involves a lot of skill and creativity. How about rules and responsibility?

Absolutely. We see that in a lot of game worlds, right? When I first started to get involved in the world of Minecraft with my son, I felt like, “This is just really open, what are you supposed to do?”

And he said, “Well, when you sign into the server, they will welcome you. And then they’ll tell you the rules, they’ll tell you what they’re doing. Like, right now we’re building this giant forest.” 

So I think we’re seeing those protocols written out quite explicitly in game worlds. And yet, we’re still having difficulty in some consumer software for social networks to make those same kinds of things explicit. We do know it creates a lot of burden on moderation and governance, which as humans we’ve been struggling with in the real world, just like virtual space. 

What do you wish that game world creators would learn from Native traditions of indigenous storytelling? What are the pieces that they most often miss? 

I do like that right now we have games and story spaces that take a long time to tell. Maybe a game could be 24 or 32 hours in player time, or longer if it’s an open world format, where people may go back again and again in order to just sort of hang out in the story world. I think that’s really beautiful. 

In a similar way, in my tribe, a story can be told every night for a winter, or for six hours at a day, and you come back the next day, it’s another six hours, and you don’t actually ever hear the “end” of the story, because it goes on. It becomes part of your life. I think that has a very ancient and grounding quality to it. 

But I think that indigenous protocol understands the value of each member—that their participation in the culture is absolute. People are not left out or left aside as quickly as I see in the gaming community. There still is not yet a design that allows people to have true safety within online spaces. I do think we’re close to being able to figure that out from a technological point of view. But it has to start being important to the people who are creating games that every member of their community feels safe and can participate in the same way.

For me that raises this question about scale. Do you see a tension between the size of an online story space, and whether people within it feel empowered to participate?

I recommend people check out the reporting over the years about EVE Online. Many social scientists have written papers about the communities that have evolved through many years of playing with thousands and thousands of members. 

I was doing a talk recently at Sundance Film Festival with the creative director of EVE Online. And he said, when he created this game, he would feel like a god creating a world. But now he just feels like a janitor, because the players are like, we demand this, we’ve organized and we voted. The players created their own tribunals, their own governance, and then they have their own systems of protests that they will do if they don’t like changes. 

So I think EVE Online is a really good example of a governance structure that truly cares about justice within their very massive multiplayer system.

Underneath what we’ve been talking about is a question about the economic models of storytelling. I’ve been thinking about the Marvel Cinematic Universe, which is a massive, centralized universe all built around Disney’s IP ownership stake, and its contracts with various actors and so forth. How do the economics of storytelling shape the kinds of stories that get told?

In the model you just mentioned, people have consolidated IP, splintered it into fragments and then shot it out to more locations, to make even more money for the centralized entity. That’s the way in which we have seen IP traditionally gain value. 

But as an audience member, or as a creator, or participant, I don’t see the younger generation being as excited about consuming more and more splintered nodes of primary IP in millions of different formats. 

I see them being more excited about participating. They’d rather see a video from someone like themselves, or rehash that meme and make their own meme comp and then follow this other person who makes another meme comp. There are a lot of exceptionally low-fidelity, low-quality deep-fried rehash meme comps that are so much more interesting and relevant to the people of their generation. They’re just cleaning the clock of traditional media in terms of relevance, and interest, and how unique and funny they are. 

We can continue to just say, we have to make money in the same ways that we’ve made it previously. Because we can’t imagine sharing. Or the next generation can find a way to communicate to older generations that they have to release the stranglehold on power, when it comes to storytelling. I’m rooting for the next generation.

What are your hopes for future generations? 

I hope that we have stories that contain our values, that are based on caring for one another, a deep understanding of each other’s experience. And I hope that we’re able to both see more of ourselves in the stories that we listen to, and also feel empowered to be storytellers in everything that we do. 🌳

Amelia Winger-Bearskin is an artist who innovates with artificial intelligence to bring positive outcomes for our land (turtle island). She is a Banks Preeminence Chair and Associate Professor of Artificial Intelligence and the Arts, at the Digital Worlds Institute at the University of Florida.

Eli Pariser is an author, activist, and entrepreneur focused on how to make technology and media serve democracy. He helped lead MoveOn.org, co-founded Avaaz.org (now the world’s largest citizen’s organization), wrote the 2011 bestseller The Filter Bubble, and co-founded Upworthy. He is currently Co-director of New_ Public at the National Conference on Citizenship.

Design by Josh Kramer. Story images by Amelia Winger-Bearskin.

Public domain images of wampum and wampum-making tools are from the National Museum of the American Indian.

I know exactly where I first experienced digital liberation, although my memory is entirely in the virtual; I can’t remember where I was sitting or what time of day it was, but this is true of most of my memories playing World of Warcraft. What I do remember is that we traveled the old fashioned way, on wolves, along the old roads where the undead prowl the hills and a sinister haze lingers in the air. We arrived at a great lake, and at the center of it stood an island and a ruined city called Caer Darrow, cast in the crimson shadows of the Western Plaguelands. We were there to tell a story.

I was still in high school, and WoW was a place I found respite in when other parts of my life were in rougher shape. By chance, I had seen a recruiting post on WoW’s forums for a new guild called the Dragonmaw Vanguard, and after attending some in-game public events, I joined. We were, among other things, a role-playing guild, meaning we collaboratively crafted deep and elaborate narratives for our characters, often spanning whole continents and worlds. Our guild also self-governed in a unique way; we took turns handing off one of roughly twelve administrator council positions, each managing various guild needs, from finance to battleground tactics, and of course, our guild’s collective lore. 

The journey to Caer Darrow was one of the first times I had been invited into the story-making process. It felt empowering: a kind of permission to be myself, with strangers in a strange place. It was in the aftermath that I remember one of the others messaging me to ask if my character might be gay. I said he was (which was easier than saying I was—I had not yet come out). They were elated at this, and said their character was queer as well, and so we began to plan the next chapter. From then on, by crafting stories with the community we built, I began to find my true self in a fantasy world; I began reaching toward freedom.

Now a decade has passed since my time inhabiting the world of Warcraft, and I’ve been reflecting on what that experience meant, wondering if that feeling of digital liberation is still possible. These days, my work in media and technology policy is concerned with the societal harms caused by tech giants, which seem to have become inescapable parts of our lives. These platforms have changed the social experience of the internet radically, including the queer experience. Certainly I, and many of my friends, delight in the endless flows of queer content on these giant platforms. It is easy to feel seen by mainstream culture, and to identify this feeling as something akin to liberation. But are amplification and representation enough? Liberation requires more—community, and care, and growth, and change.

Being queer and online, I often have a front-row seat to the floods of hate and abuse against queer people on platforms like Facebook, YouTube, Twitter, and others. I’m hesitant to generalize what I see anecdotally, but a series of reports in the last year from organizations including GLAAD in the US, Galop in the UK, and the Anti-Defamation League, have shown the horrifying extent of this online harm. Particularly distressing is the abuse toward LGBTQ+ youth, who face the highest suicide ideation and attempt rates among youth demographic groups, according to the Centers for Diesease Control and Prevention. As these platforms continue to grow, I wonder whether we have essentially automated despair and trauma. 

When I scroll through social media, I feel more like a serf in a feudal system, planting and harvesting that ephemeral substance: content.

It’s no revelation that these platforms were not built for us, are not governed by us, and do not truly serve us. But they have captured us. In the medieval fantasy world of WoW, ironically, our guild was collectivist—we shared responsibilities, resources, decisions—in order to co-create on our own terms, in our own ways. But when I scroll through social media, I feel more like a serf in a feudal system, planting and harvesting that ephemeral substance: content. I’ve noticed how rare it is now that I ever make decisions with other people on social platforms, much less practice self-governing. I see most of my friends as their content, and I, too, feel more like queer content in a machine. Can I become something else? 

It was another invitation that helped me find clues to a path forward. A friend had brought me into her book club on “eroding capitalism,” and we happened to choose The Dispossessed, Ursula K. Le Guin’s novel about an “ambiguous utopia.” Le Guin, who passed away in 2018, was a novelist more than an activist; yet her stories often reflected upon the ethical failures of our society, and what might be necessary to change. In The Dispossessed, anarchists have abandoned their capitalist planet, Urras, and established a colony on its barren moon, Anarres. The new society is built without property and governed by decentralized organizations (syndicates). But in the years since its revolutionary exodus, Anarres has slowly started to resemble the brutal society it rebelled against: more central control, less tolerance, more violence.

Anarres reminded me of the internet: both promised a radical break from the political structures of the past. I thought of how the late lyricist and internet activist John Perry Barlow famously declared the independence of cyberspace: “Governments of the Industrial World, you weary giants of flesh and steel, I come from Cyberspace, the new home of Mind. On behalf of the future, I ask you of the past to leave us alone.”—and how he later recognized his naiveté, remarking “We all get older and smarter.”  Like Anarres, the internet revolution has faltered. It, too, has become more centralized, less tolerant, and more violent. 

Le Guin contended that even within a society founded on principles of liberation and solidarity, there would still be a need to forge alternatives. The protagonist of The Dispossessed is Shevek, a physicist in Anarres whose scientific pursuits are being stifled by the increasingly close-minded society. Deeply worried for his home yet reluctant to elevate himself above others, Shevek chooses to peacefully rebel, and with his friends creates the Syndicate of Initiative, a collectivist organization dedicated to revitalizing free thought; a revolution within a revolution. With the help of the Syndicate, Shevek is able to crack the code to create a technology, owned by no one, that can instantly communicate across the stars; he calls it the Ansible.  

I wonder if we could revitalize the internet with similar acts of rebellion. I don’t think they need to be big. Our guild, after all, was a kind of syndicate: an entirely online community we built and governed on our own terms. We did so while operating within a harsh capitalist structure—World of Warcraft, and its developer, Blizzard Entertainment, have never been beacons of tolerance and inclusion. But that doesn’t mean we couldn’t build something radically different and good. 

If we are to imagine liberation in the age of massive tech platforms, the path forward cannot be more digital empires and domination; it must be anti-authoritarian and communal. I’ve found new rebellious projects, perhaps each a Syndicate of Initiative in its own way, that give me hope. The social app Lex is designed to “connect queer lovers and friends” while holding strong community ethics. Cuties, which started as a cafe but closed its doors during the pandemic, has evolved into a virtual community project to “provide resources and joy to the LGBTQIA+ community, giving special focus and attention to queer, trans, and gnc BIPOC.” Scuttlebutt is a decentralized protocol (like HTTP or RSS) for building social networks and communities free from corporate control. #ResistJam was an online game jam I got involved with in 2017 about “creating games that resist oppressive authoritarianism in all its forms.” These projects aren’t going to, say, overthrow Facebook anytime soon, and that’s not their purpose—but they are demonstrations of how we could get free, and they deserve to be supported and recognized for it. They are showing the way.

There is another scene in The Dispossessed that serves as a warning. Shevek meets a woman named Keng, an ambassador originally from Earth. We learn that her world—our world—has become an impoverished husk of itself, beset by a ruined climate and war. Contemplating the fate of Earth, Shevek asks, “Then Anarres, as you heard me speak of it—what would Anarres mean to you, Keng?” Keng replies: “Nothing. Nothing, Shevek. We forfeited our chance for Anarres centuries ago, before it ever came into being.”

A small invitation from a stranger in World of Warcraft was what set everything else into motion for me, allowing me to take my first steps toward finding myself, and building something liberating. We can still make similar invitations to people around us to build a better internet. The alternative is inaction. Unfortunately, inaction does not mean nothing happens; it means something else happens. The choice is ours. 🌳

Erik Nikolaus Martin works across several public-interest policy and technology areas. He has advised organizations including Democracy Fund and the Federation of American Scientists, served in the White House Office of Science and Technology Policy, and managed education programs at the game development software company Unity.

Illustration by Sam Sharpe.

Editor’s note: Names of interviewees have been changed.

After creating an elaborate but ultimately unsuccessful slide presentation to convince her parents that she would be a responsible user of Snapchat, 14-year-old Arwen in Virginia recently decided she no longer really cared about being on the platform. She had found a mix of other apps to fulfill her needs: Apple iMessage and FaceTime to stay connected to her close friends, TikTok to enjoy cultural content (“enemies to lovers” tropes are her favorite), and YouTube to learn to play guitar. She realized, she told me, that it’s better to delay creating a more public presence on the internet.

Part of the work of being a young person in 2021 means making daily decisions about how to engage on various platforms—and at times, actively messing with the larger structural forces of an industry that desperately wants to expand its foothold in their lives. From Google’s pervasive presence in schools and entertainment, Amazon’s foothold in e-commerce and cloud computing, to Facebook’s race to be the next WeChat, Big Tech’s push to acquire and consolidate data, and eyeballs, and pocketbooks has never been stronger. And at no time is the pressure and expectation to be extremely online more intense than during adolescence. Pandemic-related lockdowns have only amplified that reality, and it doesn’t appear to be changing anytime soon.

Wanting to understand how young people are navigating social media amid Big Tech’s ongoing digital land grab, I reached out to a handful of teens from across the United States to talk over the last few weeks of summer. With my colleagues Monica Bulger and Kiley Sobel, I also conducted online focus group interviews with fifty tweens and teens as part of the Joan Ganz Cooney Center’s By/With/For Youth Project last fall. Across all of these conversations, the themes of decentralization, multiplicity, and digital diversity kept reappearing in the way youth described their preferred ways of socializing online. Far from accepting the kind of centralized vision of interaction that would be required to sustain the “metaverse” (the tech futurist fixation of the moment), youth are busy segmenting and compartmentalizing their social engagement across different platforms according to their own needs. They recognize both their vulnerability to an industry determined to monetize them, and their power to push back. 

As with many cycles of technology, these practices aren’t entirely new; when I was studying youth and technology at the Pew Research Center, we wrote about the fluid nature of teens’ online identities and social lives as early as 2001. At the time, our team found that teens were leveraging the as-of-then entirely new affordances of digital media to experiment with multiple personas and layers of privacy and publicity in how they socialized online. It’s hard to overstate how radical this was at that moment; in the pre-internet world, those who assumed alternate identities or pseudonyms were seen as suspicious. But much of the exuberance for the freedoms of the early web was rooted in the ability to “try on” different identities. For the teens we interviewed, simply maintaining multiple email and instant messaging accounts was an early form of social decentralization that gave them the ability to be someone else—or many someone elses—online.

By 2013, our research at Pew found a new kind of fragmentation and decentralization happening; after several years of rapid adoption and fervent engagement by youth, many teens we spoke with were expressing waning enthusiasm for Facebook. They disliked the increasing number of adults on the site, got annoyed when their Facebook friends were sharing inane details, and  were drained by the “drama” that they described as happening frequently on the site. The stress of needing to manage a more public reputation on Facebook also contributed to their lack of enthusiasm about sharing content. Nevertheless, even as Instagram and Snapchat became preferred platforms for youth, teens felt they still needed to maintain some presence on Facebook in order to keep up with news and updates from their other networks.

The practices of today’s youth are potentially more seismic. Contrary to the goals of Big Tech’s investments, young people are shifting their time, content, and data away from mainstream social apps to a much more fragmented and less public kaleidoscope of communications and communities. More than previous generations, young people are now explicitly and thoughtfully controlling their boundaries of in-groups and out-groups. 

In some cases, they’re using multiple accounts (variously referred to as Finstas, burner accounts, and spam accounts, etc.). For their strong ties, they are largely socializing through interactions where there’s some shared trust framework and they can be at least partially shielded from whatever their current “public” is—whether that’s everyone at school, their extended family, or church group. This is sometimes done through social media and messaging apps—but sometimes it’s easier just to hang out on FaceTime for hours on end and avoid social media altogether.

Molly, a 17-year-old who lives in San Antonio, Texas told me the ways in which she and her friends use Instagram and Snapchat differently depending on their goals and moods. She described crafting an Instagram post as a “much more detailed” process, and that she primarily uses Instagram when she wants to see what others have posted or when she wants to make a statement. Snapchat, on the other hand, was where she communicated with her tight circle of friends and felt more free to do so—until recently, when she began to feel she “hasn’t really seen a purpose” in spending all of that time on her phone and would rather “be in the moment.”  

Kyle, a 16-year-old from New York, described the way he moves in and out of different platforms depending on who among his friends is using a given app. He sees certain rhythms across different apps during the school year versus the summer, when I spoke to him. “Our group chat for friends is on Snapchat. During the school year … people would use it for staying in touch during the day, but people aren’t really using it now.” Recently, he’s shifted the way he uses social media because he knows that college is on the horizon. He admitted to begrudgingly creating a Facebook profile because his older sister, who is already in college, said it would be a good way to show publicly that he’s involved in different clubs and sports. When I asked him about any trends he’d noticed, he mentioned the #MakeInstagramCasualAgain movement, which seeks to re-inject less filtered, less curated, and more lighthearted content into a platform where the pressure to post photos of gorgeous faces and places has taken the fun out of being social online.

They recognize both their vulnerability to an industry determined to monetize them, and their power to push back.

Some teens are opting to stay away from the pressures of posting altogether. Riley, a 16-year-old in Tampa, Florida, told me she has one Instagram account for following people from school and family, and another “private” account with 80 or 90 followers. She feels ambivalent about Snapchat: “I only got Snapchat because my friend has it. It’s like texting.” However, her favorite thing by far is just to “Facetime all day” with her best friends because they don’t go to the same school anymore. Otherwise, she spends a lot of time on TikTok, which she explained “is for entertainment, watching funny videos, but not really posting or talking.” 

TikTok is especially fascinating as a site of youth experimentation with social media practices. As Data & Society researchers Ireti Akinrinade and Joan Mukogosi detailed in a recent essay for Points, young people creatively deploy “strategic knowledge” to challenge the dynamics of the platforms they inhabit and achieve a range of social and political goals. The authors cited a teen who curated a “spam account” to watch ultra-conservative TikToks, which allowed him to then publicly expose and disrupt a troubling trend where users who had tested positive for Covid-19 were intentionally going into public spaces to infect others. 

Young people’s online behaviors shapeshift at a pace that’s perfectly appropriate for the rapid-fire evolution of adolescence, but wildly frenetic for anyone trying to understand where technology is headed. For instance, as scholar danah boyd noted in her book, It’s Complicated: The Social Lives of Networked Teens, young people will go to extreme measures to achieve the level of privacy or publicity they desire on social media—preferences that can oscillate depending on the time of day, type of post or perceived audience. And this variation can radically alter the way these users appear to behave online, creating signals that are easily misinterpreted. One teen boyd interviewed who didn’t want to be visible to family members, but did want to communicate with friends, would delete her social media account altogether every night, only to reactivate it for brief periods to check in when she knew her family would be offline.

This youthful creativity can be a threat to centralized platforms, many of which depend on generating user bases with enough data coherence to be attractive to advertisers. As such, it’s inevitable that these companies will double down on lucrative industries like cross-device tracking, data brokers, and internet-connected devices to congeal young people’s diverse social lives into reconstituted data packages. It’s something that Deana, a 13-year-old from Rio Vista, California who calls herself “an aspiring activist,” already understands is happening. Though she depends on social media for learning about the issues she cares about, she understands the platforms are making money off of her engagement. “If the product is free, you are the product,” she said in one of our focus group conversations last fall.  

The parents of these teens are already mired in a range of personal, financial and professional obligations that require relatively stable identities and leave them locked-in online. But members of Deana’s generation are likely to keep challenging the one-size-fits-all design of social media platforms. They are discovering their senses of selves—and in doing so, I believe, collectively laying the cultural groundwork for a more diversified social web: one that may genuinely put more power in people’s hands. If history is any guide, we should keep our eyes on the kids for signs of the next revolution. As Molly in Texas told me, “This is the age that [adults] have created. You get to evolve with it, why can’t we?” 🌳

Mary Madden is a veteran researcher, writer and nationally-recognized expert on privacy and technology, trends in social media use, and the impact of digital media on teens and parents. She is currently a Senior Fellow for the Joan Ganz Cooney Center at the Sesame Workshop.

Design and photos by Josh Kramer.

The scene through the chamber windows was soothing. Here on the top floor of the library, Laith Shanbezadeh could see the bright blue sky, cumulus clouds blowing eastward across the lake. It was rare that he got moments of solitude like this: no family, no climate migration officers, no school—just him, his thoughts, and the view. All it took was a lunch break during his first time on a jury.

Laith was still a bit unsure what he was doing here. Since the rest of his family had neither membership to any of NetCo’s co-ops nor access to their VPN, they had never served on a jury. And now he had been chosen, and at only age 16, too. It must have been a fluke that he was selected…

Laith was relishing the quiet when the animated jury assistant appeared from the table. “Hello, Laith. Would you like to join a virtual lunch table with Renee and Ali?” 

The assistant—the AI recorder, researcher, and reporter for the jury deliberations—dissolved back into the table when Laith accepted its offer. A moment later, real-time holograms of Ali and Renee appeared in their chairs. Rene, a digital patternist, was holoing from her home office in Uptown, and Ali, a college junior studying ceramics, from his campus civics building.

Laith had taken an early affinity to Ali and Renee. They both had a cool confidence about them. Renee seemed to have an inexhaustible knowledge of the city functions; he hoped to one day have a job like hers, constructing healthy virtual spaces and infrastructure. Ali, only a few years older than Laith, seemed so comfortable calling out some of the more senior jurors, and somehow did it with wit and charm. 

“So, how does this one compare to other ones you’ve been on?” Laith asked them cooly, trying to mask his curiosity.

Ali yawned and leaned back, briefly disappearing outside the holo feed’s field of vision. “I always brag to my classmates about how radical NetCo juries are. But when we have folks making all sorts of useless tangents, it definitely takes away from the experience.”

“Yeah—tough luck on the jury placement. Honestly, what insights can this motley crew provide about zoning?” Renee added, taking a bite of her onigiri.

“Yeah, but…”  Laith paused. Better not to reveal his interest in zoning. “…but at least we don’t have to eat a boring lunch alone.” 

Rene sighed. “I miss when you’d be dismissed cause they didn’t need you for a trial. NetCo doesn’t need to hear everyone’s opinion. Who really cares what Bob thinks about housing development?” 

“Now hold up there, Renee.” Ali, back on holo, feigned sternness. “Big Bob is an ‘OG techno-hippie.’ Of course we’d want his input!” This got Renee laughing. Regardless of its pertinence to the deliberations, Bob had taken every opportunity to remind the jury that he was a longtime civic tech influencer.

Laith chuckled cautiously. “But the whole point of NetCo’s juries to incorporate as many different people’s input into its design, right?”

Laith recalled that the Network Cooperative emerged after the Cloudbusting in the ‘30s. People formed co-ops to purchase server real estate and digital infrastructure, setting up VPNs and other web services, powered by resources shared by the co-ops. It was designed to give its members greater control over their data and digital communities. And the juries made sure everyone got a say in how the system was run. Supposedly.  

Ali put on a look of comical fear.  “Oh no, what would we ever do without our OG techno hippie?”  

“When Bob said ‘techno-hippies’ like him were to thank for democratic tech, I was done,” Renee snorted. “He definitely didn’t go to high school here.”

“Wait, what do you mean?” Laith asked.

“If he had, he’d have given his crew’s name,” Renee said. Laith stared quizzically. 

“Oh my god. You don’t know about the Pirate Crews?” Ali looked aghast. “Renee! Tell him.”

“Y’all making me feel like a senior citizen,” said Renee, motioning to the AI assistant to record her story.

Way before NetCo, when Renee’s mom was a senior at Park, the school district introduced AI mascots: bots to track social cohesion, improve test scores, monitor mental health, and offer recommendations on how to respond to student needs. Admins said the mascots were only about school spirit, but the mascots had huge influence over whether students passed or failed. 

The juries made sure everyone got a say in how the system was run. Supposedly.

The kids from the different high schools that hung out at the esports lounge downtown, called the AIs techno-fascist surveillance tools. They believed no one should have that sort of access to their lives without their consent or contribution, and formed the “pirate crews” to push back. The crews hacked into the district’s systems, and Renee’s mom’s crew, the Pirate Panthers, were the first to successfully run their AI on the district servers. Every time the district released a security update, the crews would crowdsource a workaround. 

“It went on for a semester until the district finally invited some of them to negotiate. But of course, that generation of students grew up. Some went into mainstream tech—AI, quantum computing, blockchain, cybersecurity—and others, like my mom, stayed local: focusing on civic technology, establishing municipal broadband, supporting entrepreneurs. Most of them stayed skeptical of surveillance by the government and private companies. But they also kept working together, and the result is the tangled hodgepodge of virtual infrastructure and cooperating tech organizations we call NetCo.” 

Laith was amazed. He imagined himself as a Pirate, making the web open and secure for folks in the city. “Are you a patternist for NetCo?”

Renee let out an ironic laugh. “Absolutely not! I’m with the city. NetCo is a disorganized collection of co-ops designed by folks trying to reclaim the glory of their techno-anarchist youth. We haven’t figured out democracy in the real world, why would adding technology make it better? They don’t need us to deliberate housing development; they need to hand this case to the planning department and give us our time back.”

Ali made a face. “I agree, NetCo brings in too many randos now. But I doubt a bunch of bureaucrats would do any better—”

“NetCo doesn’t include everyone,“ Laith blurted out. There was an uneasy silence. I guess I have to explain to them now, he thought. “I’m one of the few climate-displaced people that I know who’s ever been invited to a jury. We assumed CDP status precluded us from participating.”

 “You’re a CDP! That’s why you’re not holoing in.” Renee exclaimed.

A look of surprise, and perhaps disgust, seemed to cross Ali’s face, but it was hard to tell for sure through the hologram. “I’m surprised you got selected for a jury. I know CDP aren’t really interested in engaging with the community—”

“Woah,” Renee cut in. “That’s a huge assumption to—” 

“But it’s true—”

Laith knew the familiar stereotype, and felt his chest tighten—maybe he couldn’t be friends with these people outside after all. “It’s not disinterest! We never get asked and no one’s ever listening!” 

“Maybe you feel that way,” Ali huffed. “But automatic registration for government elections and NetCo juries when you’re 16—”

“But it’s not automatic if you’re older than 16, and to register or serve on a jury you have to be a resident. And if you move here and you’re older, there’s nothing automatic about it.”

“All of the CDPs I went to school with moved into a NetCo house after a year or two.” Ali replied.

“There aren’t homes for families like mine—extended families. Twelve of us share a three-bedroom apartment, and we’re typical. The new homes being built are for tiny, wealthy nuclear families; not homes that CDP families can both afford and stay together in. It’s clear they don’t care about what we want—”

“Okay, but NetCo analyzes your engagement when you use their VPNs, even if you aren’t officially serving on a jury. So at least they see your preferences and habits.”

“We get internet from the Bureau of Climate Migration, not NetCo. Also, we don’t want to be tracked by NetCo or the municipal broadband. The Pirates didn’t either.”

There was another silence, interrupted by the jury assistant’s chime. “Attention, jurors. The lunch recess will be ending. Would you like to save this recording?” 

“I forgot about the assistant. We can just delete this conversation,” Rene offered, sounding embarrassed.

“Listen—Laith, I obviously don’t know enough about CDPs.” Ali conceded. “But you just gave us some pretty good data. Maybe if we upload it, we can improve NetCo’s selection algorithms.” 

Laith didn’t say anything. He wished he shared Ali’s faith in the Pirates and the Network Cooperative they created. He wondered if Renee was right to be wary of NetCo’s idealism. Thinking of his family, he wondered if a system designed without CDPs in mind could truly adapt and include them? 

He took a deep breath and swiped the table, sending the recording to NetCo. 

As the other jurors began holoing in, Laith turned to the windows again, taking in the clouds over the lake, the gulls in the wind, the people on the shore. Maybe he was where he was supposed to be.

Lafayette Cruise is an urban planner and futurist. His practice engages projects at the intersection of urban planning and speculative fiction in order to imagine, plan, and build a more equitable, just, and sustainable future.

Design by Josh Kramer. Photo by Ciprian Boiciuc, via Unsplash.

It would have been in my kids’ middle-grade years—something like ages six, seven, eight?—when they first properly observed Instagram over my shoulder. We’d be on the couch, pasta boiling in the nearby kitchen, and they’d hang their little elbows around my neck or maybe fold into a half-sit on my lap. That was a developmental era with bright bursts of consciousness, big leaps in abstract reasoning, and a bare-bones sense of numeracy. But then—simple numbers turned out to be all they needed to understand social platforms. “Wow, Mama,” they’d say. “43 likes!” They had internalized Instagram’s deepest logic. More is nearly always more.

In the years since that little-kid parenting era, I’ve asked myself—the way so many of us have—whether more is what it’s cracked up to be. And I don’t just mean the self-evidently hollow offering of “likes.” There are all kinds of numbers that animate apps on the scale of Instagram or Facebook, reinforcing the algorithmic idea that bigger is better—or that bigger is inevitable, anyway. It took me a while to figure out that I needed something simpler, more intimate, and I got it in an unexpected way. I started using an app that’s officially marketed to families, and it became the digital social experience I’d been hoping for—not just as a parent, but as a person.

Notabli is not the kind of self-announcing refusenik tech targeted to folks who are outraged about the big social giants, who want to go rogue—or even just go weird or experimental. Notabli is an approachable and minimalist app for privately sharing, archiving, and easily printing digital images—led, as the app often notes, by an interest in “the kids you love!” It’s marketed as a convenient way to keep images of children far away from the public (or even the slippery semi-private) internet, and as a way to build the 21st century form of the family scrapbook. 

My friend-from-the-internet, software engineer Jesse Kriss, started using Notabli in 2014. He had quit using Facebook in 2009, and the prospect of a modest-scale sharing platform with strong guarantees of privacy and security appealed to him. The idea of digital connection “without all the baggage, complexity, and tradeoffs of a larger ecosystem” was so clearly desirable—even in those days, he told me. Back then, Jesse wasn’t even Notabli’s putative customer. He and his wife had no children at the time. He signed on anyway, to connect with friends who did have kids and, after his own child was born, to document his young family’s life together. 

It was at Jesse’s recommendation that I started using it to share images of my own kids’ lives a couple of years ago with a small network of people—maybe two dozen, tops. Kids fresh-faced and heading off to school. A quick video clip of a bike ride. A particularly good cake, baked by a 10-year-old. After a while, I occasionally shared some non-kid snapshots, just life-in-general. A view of the trees. A lonely streetscape during the pandemic. And then, eventually, I got off Instagram entirely. Because once I’d started noticing Notabli’s designed features—and my own behavior choices in response to them—I recognized that its virtues came from having less, rather than more.

Notabli offers a sturdy few interactive possibilities. In the app itself, there are likes, comments, and easy archiving, but zero of the glitter and charisma that make an app its own never-ending destination. There’s no third-party data harvesting, which is reassuring to parents, yes—but there are also no advertisements, no interactivity around comments, no direct messages, no ephemeral features like Stories, no tricked-out graphics. It’s a very slow, very placid feed. And the app isn’t even a necessity: You can sign up for a digest of images (or videos, or sound recordings) sent by email at regular intervals. (My parents, in their 70s and without a shred of interest in Facebook or Instagram or really even digital life in general, love the email bit: an uncomplicated and uncluttered delivery of digital media straight to their door.)

To a casual observer, Notabli might look like a banal glimpse of internet backwater, unlikely to compete with the social industry platforms and therefore, at best, a niche product for the extra-cautious. But like so many designs that are ostensibly marketed for family life in a narrow and even parochial way, there’s some signal present, something deeper that calls for sustained attention. I’ve come to think of Notabli and its “family”-branded counterparts as a proxy for another possibility—for technologies that support the smaller but heterogeneous networks in our lives, deployed at an adaptively manageable scale. The app delivers the simplicity and productive limitations that the symbol of family stands for. It’s a human-scale model for giving and receiving the records of our lives with the family-or-friend, neighbor-or-colleague communities we belong to, and want to build. 

Like Jesse, my enjoyment of Notabli has little to do with having or not having kids. We both find it to be an example of ever-elusive calm technology. “The smaller scale of content and connections means there’s no anxiety about keeping up,” he told me. “Even if I went two months without opening the app, I could easily catch up with everything I missed in under 30 minutes, even with videos.” Notabli is an unapologetically modest interaction. 

Okay, fine, but ask the average non-technical person about the idea of a smaller-scale and more humane social internet, and you’ll invite a scripted kind of response: That’d be nice, but this ship has sailed. You want efficient connection to your friends and family members? Then you go where the people are. It makes sense: most of us are too busy to find an alternative. And if you’re in the presence of book smart types, they’ll remind you: Big problems with capital—with monetized algorithmic “sharing”—aren’t addressable by individual human actions. We can be critical of these big platforms even while participating in them, they say. Make your big-platform account private and small; share only with designated “close friends,” and live with the rest. In fact, the logic goes, our participation at this point is probably out of our hands. Its entanglements in our lives have blurred the boundaries of work life and personal life. Opting out isn’t economically viable. 

I guess this is where I’ll admit that for me, the draw to Notabli and away from Instagram isn’t really sourced from a high-minded refusal so much as a sober realism—vivid and bracing in middle age—about human fallibility, the liquid slip of time, and the lie of quantifiable value. 

I find that Notabli’s limited scope—its designed less-ness—helps send me back to my closest relationships more of the time. My oldest and dearest friends, my extended family members, the lifeline mix of parents-and-children who share our schools and car rides and caregiving. I spend less of my life scrolling because I’m not trying to manage several oversized, ad-driven networks. So I’m checking in with that closer circle of people instead. And those people are going through stuff: navigating cancer and the insidious fissures in a marriage and the death of parents. I find the social bindery of more frequent interaction with a smaller set of people—the updates on the earrings they chose, or the “post”-Covid vacation planned and abandoned or, yes, the things their kids say—earns us both an open line of communication for getting through the big challenges together. These are the people who will be with me ten, twenty, thirty years from now. I want my technology to point me toward those relationships—connections with such tightly-woven histories that I can’t imagine their breaking apart, but whose ties can indeed become fragile under the slow pressure of more. 

When my kids marvel at 43 likes!, I can frame my response to them with a moderate, knowing acceptance of how the internet is, calibrating my screen time, hiding the metrics, or adding on whatever other tweaks might be possible. Those are good options, and ones that plenty of us choose. But I can also invite the necessary friction of adopting some different interaction in my digital life altogether, where an app like Notabli hands me some well-designed constraints, building less into its very architecture. I am missing out on the serendipity of random connection—high school friends alongside my neighbors—that the big platforms promise, but I’m nourishing a smaller network instead. It’s one choice among so many that I can take up or let pass me by. Instead of fighting uphill with the big apps’ relentless invitation to more, I’m enjoying addition—by subtraction. 

Here is the full loop of my interaction on Notabli these days: When I ask my 13-year-old daughter if I can take a picture or video of her playing the guitar, for example, she’ll say to me: Can you post it on Notabli? I post it, and we rarely go back to look at it again. We don’t track it for likes, even though (some two dozen) likes are available. We don’t get fed “suggested” content. We do get my daughter’s extended and inherited circle of relationships a little more closely connected to her life. That’s the point of the exchange for her: She is interested in something, and she is confident that a small set of people are interested enough in her to see it, to share it. She wants her grandparents to see these posts, and she especially wants my two best friends to see it—they live far away in Chicago, but they’re people who’ll be invested in my children forever. My daughter has taken this constellation of care in her mind’s eye and strengthened it with digital technology. We’ve learned to reinforce our relationships, not the draw to the machine. 

Author’s note: It’s one thing to write critiques of technology—what’s going wrong is relatively easy to spot, which is why good, strong jeremiads are everywhere. It’s often harder to write about instances where things are going right, like I’ve done here. But rest assured that I have no connection to Notabli, nor was I compensated by them in any way for my thoughts.

Sara Hendren, contributing editor at New_ Public, is a humanist in tech—an artist, design researcher, professor at Olin College of Engineering, and the author of What Can A Body Do? How We Meet the Built World. Her newsletter is undefended / undefeated.

Design by Josh Kramer. Photo by Leo Rivas, via Unsplash.