Surveillance Capitalism and Privacy

The ancient Greeks had a word for private citizens who stayed out of public life: idiotai, from which we get "idiot." Privacy was a mark of low status — you were private because you had nothing important to say. Fast forward to the digital age, and we've completed an extraordinary inversion: privacy has become a luxury good, available mainly to the powerful, while the rest of us live in what amounts to an open-air panopticon.¹

The mechanism is subtle because it works through convenience. You sign up for Unroll.me to clean your inbox; it's free, it's slick, Joshua Malina endorses it. What you don't know is that the service is owned by Slice Intelligence, a market research firm that rifles through your email to find Lyft receipts and sells the anonymized data to Uber. The CEO expressed that it was "heartbreaking" that users were upset. A co-founder took a different approach: "Do you really care? How exactly is this shocking?"¹

She's not wrong that we've come to accept this. But acceptance isn't the same as informed consent. The psychologist Michal Kosinski demonstrated that seemingly trivial Facebook "likes" — brands, celebrities — can reliably predict intelligence, personality traits, and political orientation. Cambridge Analytica then boasted that exactly these techniques were "instrumental" in getting Trump elected. We think of our digital behavior as private because it feels private — we're alone with our phone, in bed, on the toilet. But every swipe is a data point feeding an inference engine that knows us better than we know ourselves.¹

The Asymmetry

The real scandal isn't that data is collected. It's the information asymmetry. Data-mining companies know everything about us; we know almost nothing about what they know. And as this asymmetry deepens, the powerful have co-opted the very language of privacy to shield themselves from accountability.

Mark Zuckerberg bought four houses surrounding his Palo Alto home to ensure his own privacy — while removing users' ability to remain unsearchable on Facebook. The Trump White House cited "privacy concerns" to cut off public access to visitor logs. When the NSA was found to have swept up conversations between Israeli officials and American members of Congress, the response was revealing: Pete Hoekstra, a years-long NSA defender who had called Snowden a traitor, was suddenly outraged — not at mass surveillance in general, but at the specific instance where his people were surveilled. "Maybe unprecedented abuse of power," he tweeted, apparently without irony.²

Glenn Greenwald nailed the pattern: surveillance is always fine when it targets people unlike you. The principle of privacy only becomes sacred the moment your own conversations appear in someone else's database. This isn't hypocrisy in the ordinary sense — it's a structural feature of how surveillance power works. Those who wield it can always find a reason why this particular case is different.

The Platform Layer

Aral Balkan put it bluntly on the Web's 28th birthday: "We didn't lose control of our personal data — it was stolen from us by Silicon Valley." Tim Berners-Lee's letter that year identified the problem but named no names, perhaps because Google is one of the biggest contributors to web standards at the W3C and both Google and Facebook fund the Web Foundation. The elephants in the room stood silently in the wings.³

The business model is what Balkan calls "people farming." The product is free; you are the product. But this framing, common as it is, actually understates the problem. You're not the product — you're the raw material. The product is the behavioral prediction model built from your data, sold to whoever will pay. And the feedback loop between data accumulation and capital accumulation means the farmers keep getting richer and the crops keep getting more thoroughly harvested.

This extends beyond the obvious players. When robots.txt files — the gentleman's agreement by which websites tell search engines what they can and can't index — are studied across the top million websites, a clear pattern emerges: more sites block Baidu and Yandex than block Google. The default deny rules that technically apply to all crawlers effectively create a world where only Google can index the full web. Webmasters' preference for Google becomes self-fulfilling: Google has "better results" partly because it's the only search engine allowed to see most of the web.⁴

The academic world mirrors this. Academia.edu, which is not affiliated with any academic institution despite the domain name, has built a vast repository of scholarly profiles and papers. But it's vendor lock-in dressed as openness — no public API, login required for most features, and no way to export your data to a competing service. As one speaker put it: "On the Internet, the opposite of 'open' is not 'closed' — the opposite of 'open' is 'broken'."⁵

Pseudonymity as Self-Defence

The privacy problem looks different when you're not a corporation but an individual trying to think in public. Scott Alexander's 2021 essay "Still Alive" is the clearest account I've read of what's actually at stake when pseudonymity breaks down. Alexander blogged under a pen name for years — not because he was hiding anything sinister, but because he was a psychiatrist whose career nearly ended when interviewers found his earlier blog. Therapists are supposed to be blank slates for patients to project onto; a public intellectual life is professionally disqualifying. So he rebuilt his online presence under a pseudonym, and it worked — until the New York Times decided to write a profile and insisted on using his real name.⁶

The part that stings isn't the doxxing itself but the framework the Times applied. In their view, they started with the right to reveal his identity, and he had to earn anonymity by proving he was sympathetic enough — serious enough death threats, vulnerable enough life circumstances. Alexander's counter is devastating: "There's no dignified way to answer any of these questions except 'fuck you.' Just don't kick me in the balls." The burden of proof is backwards. You shouldn't have to demonstrate that exposure would literally kill you in order to keep your name off a newspaper. A 2020 poll found 62% of Americans afraid to express political beliefs, up ten points in three years, with the sharpest decline among "strong liberals." When the public square becomes a place where any newspaper can casually destroy your life, the square empties — and the only voices left are professional opinion-havers backed by institutions that can absorb the reputational risk.

What connects this to the surveillance capitalism story above is the asymmetry of consequences. The Times reporter who nearly doxxed Alexander wasn't malicious — it was Tuesday for him, a routine story. The policewoman who harassed Mohammed Bouazizi wasn't staging a political act — it was Tuesday for her too. The damage done by powerful institutions to individuals rarely comes from malice. It comes from indifference, from the institutional inability to recognise that what's routine for them is catastrophic for the person on the other end.

Terror Capitalism: The Endpoint

If the Silicon Valley model is "people farming" — extracting behavioral data from consenting-by-default consumers — then what's happening in China's Xinjiang region is the endpoint of that logic: a state that treats an entire ethnic minority as a data source, and an industry that profits from the extraction.⁷

The Integrated Joint Operations Platform (IJOP) uses AI to monitor checkpoints, facial recognition cameras, biometric databases, and social media across Xinjiang's Uyghur population. The system doesn't just watch — it profiles and predicts. It generates risk scores based on whether someone prays regularly, has a relative abroad, travels outside their registered police precinct, or has unauthorized Islamic content in their WeChat history. Those scored "unsafe" are sent to detention centers, and from there to internment camps that the government calls "transformation through education" facilities. Since 2017, more than a million Turkic Muslims have moved through this system.

The anthropologist Darren Byler calls it "terror capitalism" — the justification of counterterrorism creating a space in which tech companies can build, experiment with, and refine surveillance systems at a scale and intensity unavailable in liberal democracies. From the perspective of China's security-industrial establishment, "the principal purpose of Uyghur life is to generate data." The Uyghur region has become an incubator: dozens of Chinese tech firms — Hikvision, SenseTime, Meiya Pico, Leon Technology — developed their products on Uyghur subjects and now market them globally. CloudWalk finalized a deal with Zimbabwe to build a national facial recognition program. A Leon Technology spokesperson cited "unlimited market potential" in the Muslim-majority nations along the Belt and Road.

What's chilling about Byler's account is how the technology works not just through what it catches, but through what it threatens. Not using social media is suspicious. Destroying a SIM card is suspicious. Not carrying a smartphone is suspicious. Uyghurs buried phones in the desert, hid SIM cards in frozen dumplings, tried to destroy data cards in secret. They adapted their behavior, and eventually their thoughts, to the system. As one businessman told Byler: "Uyghurs are alive, but their entire lives are behind walls. It is like they are ghosts living in another world."

The connection to Western surveillance capitalism isn't metaphorical. The same facial recognition techniques, the same behavioral prediction models, the same business logic of extracting data from populations to sell predictions — it's the same industry, operating under different political constraints. The difference between Silicon Valley and Xinjiang isn't the technology or even the business model. It's who has power over whom, and what limits exist on how that power is used. When those limits erode, the technology is ready.

Privacy as Class Divide

Hannah Arendt wrote that privacy was once "a state of being deprived of something." Now, Congress votes to allow ISPs to sell browsing data, and talk immediately turns to premium products people could pay for to protect their habits. Can't afford it? As one congressman told a constituent: "Nobody's got to use the internet." Except that practically, everybody does — tech companies have colonized the public square. We use Facebook to organize protests, pose questions in debates, support candidates. We're paying a data tax for participating in democracy.¹

The philosopher Paul Ohm warned in 2009 of a "database of ruin" that would eventually connect every person to compromising information. As Zeynep Tufekci put it: "People can't think like this: I didn't disclose it, but it can be inferred about me." When a peeping Tom looks between the blinds, you know what's been seen. When a data firm cracks open your inbox, you may never find out what it learned — or how it was used.

The trajectory is clear. Orwell imagined that in a totalitarian state, the proles would be the ones spared surveillance while the powerful were watched. We've built the exact opposite: the powerful demand privacy while ordinary citizens are transparent to corporations and governments alike. Stepping into the White House is now more private than the rash you Googled.