# History of Social Media Social media is where the [[Surveillance State]], [[Surveillance Capitalism]], and political radicalization converge into a single system. The platforms were built to maximize engagement; radicalization is a feature, not a bug, of engagement optimization. The algorithmic turn of 2012–2016 is when social media went from communication tool to political weapon. [[Elon Musk]]'s Twitter acquisition made explicit what was always implicit: these platforms are instruments of political power, and whoever controls the algorithm controls the information environment. The arc from [[LifeLog]] → Facebook → algorithmic radicalization → the [[January 6 Capitol Attack]] is one of the vault's most important tech-to-politics pipelines. This page traces the full trajectory. ## Precursors (1990s–2002) The social internet existed before social media. Bulletin board systems (BBSes) in the 1980s, Usenet newsgroups, and Internet Relay Chat (IRC) in the early 1990s were text-based, decentralized, and ungoverned — closer to pamphleteering than broadcasting. GeoCities (1994) let users build personal websites organized into themed "neighborhoods," creating a proto-social web of user-generated content. Yahoo acquired it for $3.6 billion in 1999 and eventually destroyed it — an early example of the acquisition-and-degradation cycle that would define platform capitalism. SixDegrees.com (1997) was the first social network with both user profiles and friend lists — the structural DNA of every platform that followed. It shut down in 2001, having proved the concept without finding the business model. Friendster launched in 2002 and demonstrated that social networking could attract mass audiences — at its peak it had over three million users — but collapsed under scaling problems and management dysfunction. The precursor era established that people would voluntarily publish personal information, build social graphs, and spend hours engaging with user-generated content online. The business model that would monetize this behavior — [[Surveillance Capitalism]]'s extraction of behavioral data for predictive advertising — had not yet been invented. [[Google]]'s 2002 decision to monetize search-residue data through targeted advertising provided the template. ## The MySpace Era (2003–2006) MySpace launched in August 2003 and became the most-visited website in the United States by mid-2006, surpassing [[Google]]. Its competitive advantage was customization — users could modify their profile pages with HTML and CSS, embed music players, and create visually distinctive personal pages. The platform became the center of the independent music scene and youth culture. [[Rupert Murdoch]]'s News Corporation acquired MySpace for $580 million in July 2005 — a deal that seemed visionary at the time and catastrophic in retrospect. News Corp treated MySpace as a media property to be monetized through display advertising rather than as a social platform to be scaled through network effects. By 2008, Facebook had surpassed it. By 2011, News Corp sold MySpace for $35 million. The MySpace trajectory is a case study in how legacy media companies misunderstand platform dynamics — and a reminder that platform dominance is not permanent. ## Facebook's Rise (2004–2012) TheFacebook launched on February 4, 2004 — the same day DARPA cancelled [[LifeLog]], the military life-recording program. [[Mark Zuckerberg]] built it in a Harvard dorm room as an exclusive social network for Harvard students. The exclusivity was the growth engine: Harvard only, then Ivy League schools, then all colleges, then high school students, then everyone. Each expansion wave created FOMO-driven adoption. Facebook's early product was a profile page and a chronological feed of friends' updates — a digital yearbook that updated in real time. The platform was useful, addictive, and seemingly benign. The transformation began in September 2006, when Facebook introduced News Feed — the first major algorithmic curation of social content. Users no longer had to visit each friend's profile to see updates; the algorithm selected and surfaced content for them. Users initially revolted (a protest group titled "Students Against Facebook News Feed" gained 700,000 members within days), then adapted. The pattern — introduce algorithmic mediation, absorb backlash, normalize — would repeat across every platform. [[Peter Thiel]]'s $500,000 angel investment in Facebook in 2004 — the first outside funding — connected the platform to the [[PayPal Mafia]] and the broader [[Silicon Valley]] venture ecosystem. Facebook's May 2012 IPO valued the company at $104 billion. It had one billion monthly active users. The significance of the Thiel connection runs deeper than investment returns. Thiel's worldview — articulated in his famous line that "competition is for losers" and his investment thesis that the most valuable companies are monopolies — shaped Facebook's strategic posture. The platform pursued monopoly through network effects, acquisition of competitors ([[Instagram]] in 2012, [[WhatsApp]] in 2014), and the cultivation of switching costs (your social graph exists on Facebook; leaving means losing your connections). The [[Antitrust]] implications are direct. ## Twitter and Real-Time (2006–2015) Twitter launched in July 2006 with a radically different model: 140-character public messages, a reverse-chronological feed, and an open follow graph (you could follow anyone without reciprocal permission). The constraint bred a new form of discourse — concise, public, and fast. Twitter became the platform of record for breaking news, political commentary, and real-time public conversation. The Arab Spring uprisings of 2010–2011 cemented Twitter's reputation as a tool of democratic mobilization. Protesters in Tunisia, Egypt, [[Libya]], and [[Syria]] used Twitter to coordinate, document, and broadcast their movements to the world. Western media celebrated social media as a liberation technology — a narrative that aged poorly as the same platforms proved equally effective for authoritarian surveillance, far-right radicalization, and coordinated [[Disinformation]] campaigns. Twitter's reverse-chronological feed was its brand identity and, in retrospect, the feature that kept it relatively honest. You saw what the people you followed posted, in the order they posted it. The algorithm didn't decide what you should see. This would change. ## The Algorithmic Turn (2012–2016) The period between 2012 and 2016 is when social media became a different thing. The shift was from chronological feeds — you see what your friends post, in order — to algorithmic feeds — the platform decides what you see, optimized for engagement. The business logic was straightforward: algorithmic curation increases time-on-platform, which increases ad impressions, which increases revenue. The political consequences were catastrophic. Facebook's algorithm optimized for engagement signals — likes, comments, shares. Content that provoked strong emotional reactions (outrage, fear, moral indignation) generated more engagement than content that informed or entertained. The algorithm learned that [[Engagement Farming]] works best with emotionally provocative content, and it served more of it. This is the structural mechanism behind the [[Outrage Industry]]: platforms didn't create outrage content, but they systematically amplified it because it performed well by the metrics the algorithm optimized for. Instagram, acquired by Facebook in 2012 for $1 billion, adopted algorithmic feed ordering in 2016. Twitter introduced an algorithmic timeline in February 2016 — prompting the #RIPTwitter backlash from users who understood that algorithmic curation would fundamentally change the platform's character. [[YouTube]]'s recommendation algorithm, refined throughout this period, optimized for watch time — and researchers documented that the algorithm reliably recommended progressively more extreme content because extreme content held viewers' attention longer. The algorithmic turn is the inflection point. Before it, social media was a communication tool with the biases of its users. After it, social media was an amplification machine with the biases of its optimization function. The distinction matters: individual users choosing to share outrage content is a human behavior problem; a platform systematically amplifying outrage content because it maximizes engagement metrics is a structural design problem. ## Radicalization and Manipulation (2016–2020) The 2016 U.S. presidential election revealed the political consequences of engagement-optimized platforms. The [[Internet Research Agency]] — a Russian government-linked operation based in St. Petersburg — conducted a systematic [[Disinformation]] campaign across Facebook, Twitter, Instagram, and YouTube, creating fake American personas and groups designed to inflame political divisions. The operation reached an estimated 126 million Facebook users. The [[Cambridge Analytica Scandal]] (revealed in 2018, operations dating to 2014–2016) exposed a different vector: [[Cambridge Analytica]], working for the Trump campaign, harvested Facebook profile data from 87 million users through a personality-quiz app and used it for [[Psychographic Profiling]] and [[Microtargeting]] of political advertising. The scandal demonstrated that [[Surveillance Capitalism]]'s data-extraction infrastructure could be directly weaponized for political manipulation. Facebook's own internal research, portions of which leaked to the public through various channels before and during the Frances Haugen disclosures, found that the platform's recommendation algorithms drove users toward extremist groups. The mechanism was simple: if a user joined a group expressing mild political discontent, the algorithm recommended more extreme groups because those groups generated higher engagement. [[QAnon]] spread through this exact pathway — algorithmic recommendation funneling users from mainstream conservative content into conspiracy communities. YouTube's recommendation algorithm followed the same pattern. Researchers documented "rabbit holes" in which a viewer starting with mainstream conservative content would be algorithmically guided through progressively more extreme material — from Republican politics to far-right commentary to white nationalist content — because each step deeper held the viewer's attention and generated more watch time. The consequences extended beyond American politics. In Myanmar, Facebook was the primary communication platform for a population newly connected to the internet. The platform's engagement-optimization algorithms amplified anti-Rohingya hate speech and coordinated incitement to violence. A UN fact-finding mission concluded that Facebook played a "determining role" in the genocide. In Christchurch, New Zealand, a white supremacist livestreamed a mass shooting on Facebook in March 2019 — the platform failed to prevent the broadcast and copies spread across platforms faster than content-moderation systems could remove them. ## Platform Power and Backlash (2020–2023) The [[January 6 Capitol Attack]] in 2021 crystallized the platform-power question. Social media — particularly Facebook groups, Twitter, and fringe platforms like Parler — had been used to organize, promote, and livestream the assault on the Capitol. In the days following, [[Donald Trump]] was banned from Facebook, Twitter, Instagram, YouTube, and other platforms — an unprecedented exercise of corporate power over a sitting president's communications. The [[Deplatforming]] of Trump split the political response along lines that revealed the incoherence of both parties' positions on platform power. Liberals who had spent years arguing that platforms needed to do more content moderation celebrated. Conservatives who had spent years arguing that platforms were private companies free to set their own terms of service suddenly demanded government regulation of platform speech. The [[Section 230]] debate — whether platforms should be liable for user-generated content — became a rare point of bipartisan interest pursued for contradictory reasons. Frances Haugen, a former Facebook data scientist, testified before Congress in October 2021 and released tens of thousands of internal documents (the "Facebook Papers") documenting that the company knew its algorithms harmed teenagers, amplified extremist content, and undermined democratic discourse — and chose profit over safety. [[Meta]] (the company renamed itself in October 2021, a rebrand that fooled approximately no one) faced [[Antitrust]] hearings, regulatory threats, and advertiser scrutiny. [[TikTok]]'s rise during this period introduced a new variable: an engagement-optimization algorithm so effective that it made Facebook and Instagram's algorithms look quaint. TikTok's "For You" page — an entirely algorithmic feed requiring no social graph — demonstrated that a platform could achieve maximum engagement without the user needing to follow anyone at all. The algorithm *was* the product. The national-security concerns about TikTok's Chinese ownership (ByteDance) added a [[Surveillance State]] dimension to the platform-power debate. ## The Elon Musk Era (2022–Present) [[Elon Musk]]'s acquisition of Twitter for $44 billion in October 2022 — and its subsequent rebrand to X — made explicit what had always been implicit: social media platforms are instruments of political power, and ownership of the platform is ownership of the information environment. Musk gutted content-moderation teams, reinstated previously banned accounts (including [[Donald Trump]]'s), replaced the verification system with a paid subscription (effectively selling credibility), and algorithmically boosted his own posts and those of accounts he favored. Advertisers fled — at various points, ad revenue dropped by over 50%. The platform's function shifted from a public communications utility to a political instrument aligned with Musk's personal ideology and political relationships. The Twitter/X transformation prompted migration to alternatives: [[Bluesky]] (founded by former Twitter CEO [[Jack Dorsey]], built on a decentralized protocol), Threads (launched by [[Meta]] in July 2023, leveraging the Instagram social graph), and Mastodon (a federated, open-source alternative). None achieved Twitter's pre-Musk centrality as a public square. The fragmentation of the social media landscape — from a few dominant platforms to a constellation of ideologically sorted alternatives — may itself be the most politically significant outcome. ## AI-Generated Content and the Post-Social Era The flood of AI-generated content — text, images, video — into social media feeds beginning in 2023 represents a qualitative break. When a significant fraction of content on a platform is generated by AI rather than humans, the platform is no longer "social" media in any meaningful sense. The engagement-optimization algorithms that were designed to surface the most engaging *human* content now surface the most engaging *content*, regardless of whether a human created it. The economics of [[Engagement Farming]] shift from human content creators competing for attention to AI-generated content competing for attention — a race to the bottom of the uncanny valley. The implications for [[Disinformation]] are direct: AI-generated deepfakes, synthetic text, and bot-generated engagement make it progressively harder to distinguish authentic human communication from manufactured content. The platforms that were already struggling to moderate human-generated [[Disinformation]] are now attempting to moderate machine-generated disinformation at scale — a technical challenge they are losing. ## Why This Matters for the KB Social media is the vault's primary convergence zone — the place where [[Surveillance Capitalism]]'s data extraction, the [[Outrage Industry]]'s emotional manipulation, [[Disinformation]]'s epistemic pollution, and the [[Surveillance State]]'s monitoring capabilities all operate simultaneously on the same infrastructure. **The engagement-to-radicalization pipeline.** The algorithmic turn of 2012–2016 created a structural incentive for platforms to amplify emotionally provocative content, which reliably produces political radicalization. This is not a conspiracy; it is an optimization function. The same engagement-maximization logic that serves a user increasingly extreme political content also serves them increasingly extreme diet content, increasingly extreme fitness content, and increasingly extreme conspiracy content. Radicalization is a side effect of engagement optimization — and a predictable one that internal research at Facebook documented and that management chose not to address. The connection to [[QAnon]]'s algorithmic spread, the [[January 6 Capitol Attack]]'s social-media organization, and the broader [[Stochastic Terrorism]] dynamics the vault tracks is structural. **The LifeLog-to-Facebook pipeline.** The temporal coincidence of [[LifeLog]]'s cancellation and Facebook's launch — and the functional equivalence between DARPA's life-recording ambition and Facebook's actual product — is the vault's most provocative tech-to-politics connection. Whether or not there is a causal link (there is no credible evidence of one), the structural point stands: what the government proposed to build as a surveillance program, the private sector built as a consumer product, and the public voluntarily populated it with more personal data than any intelligence agency could have compelled. The [[Surveillance State]] didn't need to impose [[Total Information Awareness]]; [[Surveillance Capitalism]] made it voluntary. **Platform power as political power.** [[Elon Musk]]'s transformation of Twitter into a personal political instrument — and [[Mark Zuckerberg]]'s capacity to shape the information environment of billions — demonstrates that platform ownership is a form of political power that existing democratic frameworks were not designed to constrain. The [[Section 230]] debate, the [[Antitrust]] hearings, and the [[Deplatforming]] controversies are all symptoms of a political system that has not yet developed the institutional vocabulary to govern platform power. The vault's tracking of [[Techno-Authoritarianism]] and [[Military-Industrial Complex 2.0]] operates in this space. ## See Also - [[Surveillance Capitalism]] — the economic logic - [[Surveillance State]] — the state-surveillance interface - [[LifeLog]] — the DARPA precursor - [[Total Information Awareness]] — the surveillance architecture that social media made voluntary - [[Engagement Farming]] — the platform optimization dynamic - [[Outrage Industry]] — the amplification economy - [[Disinformation]] — the epistemic pollution - [[Stochastic Terrorism]] — the violence pipeline - [[QAnon]] — algorithmic radicalization case study - [[Cambridge Analytica Scandal]] — data weaponization case study - [[Internet Research Agency]] — state disinformation case study - [[Microtargeting]] — the political application - [[Psychographic Profiling]] — the individual-inference layer - [[Section 230]] — the platform liability framework - [[Deplatforming]] — the content moderation controversy - [[Antitrust]] — the monopoly power framework - [[January 6 Capitol Attack]] — the convergence event - [[Elon Musk]] — the platform-as-political-instrument case - [[Mark Zuckerberg]] — the platform-as-infrastructure case - [[Facebook Timeline]] — the detailed corporate timeline - [[Donald Trump]] — the deplatforming subject - [[TikTok]] — the algorithmic-feed-as-product model - [[Bluesky]] — the decentralized alternative - [[Techno-Authoritarianism]] — the endpoint