Death of the Free Internet
History
Arcotherium:
“Until 2015, an optimistic and liberal (in the best sense) ethos pervaded the Internet. Major Internet platforms were open and had little content policy beyond bans on real-life criminal activity (stalking, selling illegal goods, child pornography, etc). Aside from these narrow and reasonable explicit exceptions, anyone could go online and converse with interested people around the world about whatever they wanted in public. These platforms had no real means to control what people talked about on them even if they wanted to, which they didn’t; to the extent an organization can have views all believed in free1 and open online discourse, which promised to make censorship impossible and end the hegemony of 20th century broadcast media.
In this, they were backed up by major non-tech institutions, such as the UN, the EU, the US State Department, and the New York Times2, all of which saw Internet freedom as a corollary of offline freedom, and one which would liberalize the world’s remaining non-democratic states in time. These institutions both supported Internet freedom rhetorically3 and spent hundreds of millions of dollars creating and disseminating technical tools to evade Internet censorship and preserve online anonymity.
But in just a few years, every single major platform had created strict, but incredibly vague and opaque, ideological content rules backed up by a colossal moderation/censorship apparatus with tens of thousands of human moderators and an entire new field of machine learning created from nothing. The Internet went from a space for people without institutional backing to get their views out to one with regular purges and demonetizations of heterodox figures and those associated with them, encouraged by those very same non-tech organizations that formerly championed Internet freedom. This was justified as a response to (massively overblown and mostly fictitious) Russian influence campaigns4, and as fighting nebulous “hate,” the definition of which could be shifted at will to cover whatever views or ideas the organizers classifying it wanted it too and to exclude those they didn’t, and “misinformation.”
The closure of the Internet occurred in two distinct phases: 2015-2019 (associated with the 2015 migrant crisis, Brexit, Donald Trump’s election, and the so-called alt-right) and 2020-2022 (associated with the lockdowns and the 2020 election). Almost all of the technical and organizational infrastructure and conceptual justifications were built out in the first phase, which successfully decapitated and destroyed the previously-independent, but by definition fringe, online ecosystem of ideas. In the second phase, the same weapons that previously targeted heterodox online discourse were turned against opposition to the COVID measures du joure and mainstream conservatives. The scale of censorship has no precedent; tens of millions of accounts suppressed or banned, billions of posts removed.
These purges are usually framed as a direct response to particular events or atrocities. For example Amazon ostensibly removed all Confederate flag merchandise from their web store in response to Dylan Roof’s 2015 massacre of black churchgoers. This is a mistake. Atrocities without a propaganda campaign from the press and political pressure from an opaque network of governments and quasi-government NGOs lead to no changes. There were no mass purges after Audrey Hale’s 2023 murder of elementary schoolers in Nashville motivated by anti-white hatred or the 2016 murder of five Dallas police officers by an ex-marine motivated by the BLM movement. On the other hand, press campaigns not centered around an atrocity, such as the mass deplatformings of YouTube personalities, lead to the same types of purges as the atrocity-based ones. It is not the atrocity that causes the changes, but the media and behind-the-scenes pressure campaign.
The late 2010s also featured suppression of heterodoxy offline as well. This combined physical attacks on the streets, as happened to Charles Murray, and the firing of public figures with non-Internet platforms, as with Bo Winegard and Bryan Pesta. Even moderates and liberals in already left-wing institutions were not immune, as the purges of those who valued pluralism and objectivity within the New York Times, ACLU, and many progressive activist organizations showed. As such, the closure of the Internet shut down all avenues for heterodox discourse during the Great Awokening. If you spoke from an institutional platform, you would be fired. If you spoke in person, you would be beaten or shot. And if you spoke online, you would be censored, suppressed, demonetized, and doxxed.
The closure of the Internet is the most important social phenomenon of the late 2010s.
…
The clearest pattern across all of these platforms was one of very rapid escalation. The closure of the Internet is one of the clearest examples of the slippery slope (or “no limiting principle”) in action; it only took a few years to go from deplatforming the fringe neo-Nazi website Stormfront to the entirely mainstream New York Post, from banning books on Holocaust denial to banning those opposing transgender maximalism, and from banning honestly discussing the effects of immigration in Europe to banning honestly discussing the lab leak hypothesis of COVID origins. It is one thing to defend freedom of speech or freedom to create a website in the abstract and quite another to defend Andrew Anglin calling to exterminate the Jews. And even the most unpopular political or intellectual expression is easier to stomach than something like r/watchpeopledie. But the free speech maximalists were right. Once the conceptual and technical architecture for identifying, deplatforming, suppressing, and censoring “hate” and “disinformation” was built, the institutions defining these terms quickly extended them to as many inconvenient facts, ideas, and positions as they could.
A whole new conceptual universe was constructed to justify and organize the closure of the Internet. For example, the words “disinformation” (false information shared maliciously), “misinformation” (false information7 shared inadvertently), and “malinformation” (information that is true but misleading) practically did not exist until 2016, when in response to Trump’s election they were rustled into existence by a coterie of governments, academics, and nonprofits.
The meaning of the term “online safety” shifted from concerns about people using the Internet to physically attack, stalk, and harass people in real life to horror at the idea of users seeing marginalized ideas and information. This led to the Orwellian naming of so-called “Online Safety Acts,” which rather than focus on preserving anonymity and pseudonymity, which protect the physical safety of people online, instead work to eliminate them. A similar definition shift occurred with “harassment and abuse” and “violence.” This reached its apogee with the slogan “silence is violence,” the totalitarian idea that one must not simply not oppose orthodoxy but actively and enthusiastically propagate it or be treated as a murderer and therefore legitimate target.
The Soviet concept8 of “hate speech9”, first imported at scale to the West during the 60s revolution, went from fringe to a common term in everyday discussion.
The first big foundational dataset in the field (using “a list of criteria founded in critical race theory”) was published in June 2016 by Zeerak Waseem and Dirk Hovy. Along with academia, Google Jigsaw spearheaded this creation of this field, publishing a much larger Wikimedia dataset in October 2016 and launching Perspective API as a baseline for the rest of the moderation/censorship industry (“empowering moderators to identify and address toxicity in comments”) to build on in February 2017. The first academic workshop on “Abusive Language Online” was held in August 2017. By December 2017, the field was fully institutionalized10, with Kaggle and Jigsaw launching the “Toxic Comment Classification Challenge,” which received over 4000 entries.
The last major piece of this conceptual universe was the word “platforming.” This word shifts the burden of action; rather than exchange of ideas being the default and censorship something that must be justified, censorship becomes the default and permission to speak the exception that must be argued for. A person or group or idea is no longer “censored” or “banned” but “platformed.”
(https://arctotherium.substack.com/p/the-closure-of-the-internet)
Discussion
What made it all possible: the Para-State
Arcotherium:
“Tech companies had, and used, the power to close the Internet, but they could not and would not have done it alone. The 20th century totalitarian party-states, like Nazi Germany, the USSR, or the PRC, with their tendencies to formalize and unify authority, had official government censorship departments with lists of ideas and topics that could not be mentioned, and a party line on what narratives are publicly acceptable. Western states today, and especially the US, do not. Instead, these functions are outsourced to an opaque web of nominally-private NGOs, media organizations, and academics. Jacob Siegel, pointing to the close links between these organizations, tech companies, and US intelligence, calls this network The Information State, and I prefer the term “para-state” for those organizations that are neither for-profit tech platforms nor officially part of the government.
This type of arrangement, with organizations separate from, but reliant on and often deputized by, the state determining publicly-permissible thought is common in Western history. For much of its existence the Catholic Church played a similar role. Today, the most institutionally similar regime outside the West is the Islamic Republic of Iran, with a clerical body that is not part of the formal state, the Supreme Council of Cyberspace, determining what is to be censored.
Why is it appropriate to call these organizations para-statal rather than private? First, because they are often public trusts, given tax privileges35 in exchange for operating in the public interest as determined by the state. Second, they were and are usually funded by governments— especially the EU, Germany, Britain, and the US—or by other tax-privileged public trusts. Third, they exercise power, such as blacklisting and censorship, usually exercised by states36. Fourth, they are often directly privileged and empowered by governments; for example, under the EU’s Digital Services Act, tech platforms must defer to registered “trusted flaggers.” These institutions are not part of the formal state, but they are part of the ruling regime.
In the US, the para-state is much more dangerous than the formal state when it comes to censorship. The formal state in the US is extremely fragmented and polarized, which means there are always power centers within the formal government to oppose the censors. Perhaps even more importantly, the US has strong memetic antibodies against formal state restrictions on speech, which makes it very easy to point at and mobilize opposition against. The para-state, on the other hand, is homogeneously leftist (aside from a handful of explicitly conservative organizations, as per Conquest’s Second Law) and unbound by Constitutional and many legal restrictions. The formal state also typically coordinates through public and explicit law or regulation rather than vague, opaque, and ever-shifting “codes of conduct” and “content policies,” which are much harder to oppose. This is less true in the rest of the world, where explicit state controls on speech are common. There, such organizations are often backed up by very heavy-handed state power.”
…
* Journalists
Before the Internet, the social role of the media was to uncover and disseminate information about the world. The Internet rendered this largely obsolete, as it allowed for much easier dissemination of information, which put traditional media in steep decline. The closure of the Internet gave the media a new lease on life as an integral component of the emerging censorship para-state. Rather than finding information, media outlets became sources of official truth to be cited by fact-checking organizations to be used by tech companies to provide lists of ideas, groups, and individuals to suppress, and to coordinate censorship campaigns with threat of bad press. To facilitate this, Meta and especially Google began replacing lost ad dollars with their own grant programs for recognized news organizations.
In some jurisdictions, this para-state relationship is even formalized, with the governments of Canada, Australia, the Nordics, the United Kingdom, France, California, New York, and Illinois giving officially-recognized journalists direct subsidies or tax exemptions, and the European Union and the United Kingdom exempting them from some online speech restrictions. This is in addition to publicly-funded broadcasters in the UK, France, Germany, Italy, Spain, Ireland, Portugal, the Netherlands, Belgium, Switzerland, Austria, the Nordics, Canada, Australia, and New Zealand, which enthusiastically participated in the closure of the Internet.
* NGO Enforcers
The last leg of the para-state involved in the closure of the Internet was the army of enforcers that was responsible for making sure dozens or hundreds of the different private organizations involved got with the program of mass censorship. NGOs can’t throw people in prison, so they operated through pressure campaigns, both in public in concert with media organizations, and behind the scenes, against private companies. These organizations used mafia tactics: censor the people and ideas we want you to censor, bringing people aligned with us into your organization and decision-making structure to do so in the process, or we’ll launch a PR campaign smearing you, with the threat of aligned governments and already-captured private institutions destroying your business to back said PR campaign up. Like the fact-checkers and disinformation experts, these enforcer organizations were and still are overwhelmingly tax-privileged nonprofit public trusts.
It is difficult to tell from the outside, but the most important of these organizations is probably the Southern Poverty Law Center (SPLC). The best account here is Patrick McKenzie’s “Notes on a non-profit indicted for bank fraud,” which can be summarized as follows (though I would recommend reading the entire thing): The SPLC ran a coordinated pressure campaign, starting in 2017 and escalating in 2020 and 2021, going after “Internet companies” which included not only tech companies like Facebook and Twitter, but also “banks, credit card processors, and any other financial infrastructure their enemies could touch” to remove “extremists” from the Internet and from financial infrastructure. The SPLC maintains a blacklist, the SPLC Extremist Files, to be used in financial infrastructure and elsewhere, one with quasi-official status used across many major companies and platforms, alongside the US Treasury Department’s OFAC. It’s worth pondering the word “extremist” for a moment. Even taken entirely literally—and the SPLC, of course, is not a neutral organization whatsoever—all it means is extreme relative to the mainstream/orthodox view. The SPLC thus does its best to purge any sufficiently radical social critique from both financial infrastructure and the Internet.
The SPLC campaign had two parts, a public pressure campaign in the media, and a much more consequential series of private, behind-the-scenes, meetings with industry.”
(https://arctotherium.substack.com/p/the-closure-of-the-internet)
Fact-Checkers
Arcotherium:
“In order to suppress “misinformation,” tech companies needed some source of truth. But no source of objective truth on all possible topics of discussion exists in the real world, and Facebook or Google or some government saying “we will decide what you’re allowed to talk about online” would not have survived scrutiny. So to determine what was officially true for the purposes of moderation, tech companies turned to third-party fact-checking organizations, which sprouted like weeds during the closure of the Internet. These groups attained total epistemic authority over the public square between 2015 and 2022.
Most official fact-checking organizations were and are certified by one of two groups: the International Fact-Checking Network (IFCN) and the European Fact-Checking Standards Network (EFCSN), which creates a chokepoint in the fact-checking ecosystem. There is significant overlap between the IFCN and EFCSN; many organizations are certified by both and EFCSN’s chair also sits on the IFCN’s advisory board. The EFCSN was founded in 2022 with the consolidation of several dozen pre-existing European fact-checking organizations under one umbrella. The EFCSN is funded mostly by the European Union, with additional contributions from Meta and Google, the last funding a truly Orwellian “prebunking at scale” project. The IFCN was founded by the Poynter Institute, a school of journalism, in 2015, after a $1M foundation grant. The Poynter Institute hired an ex-SPLC employee to create a list of 515 news websites to be blacklisted by advertisers, including mainstream conservative ones like the Washington Examiner, in 2019. The IFCN’s funding sources are very similar to those of the fact-checking organizations it certifies: tax-privileged foundations, Google and Meta, and the US State Department. “From its launch in 2015, the IFCN has received funding from the following organizations: Arthur M. Blank Family Foundation, the Duke Reporters’ Lab, the Bill & Melinda Gates Foundation, Google, Facebook/Meta, U.S. State Department, the National Endowment for Democracy, the Omidyar Network (Luminate), the Open Society Foundations, Fritt Ord Foundation, Duke Reporters Lab, Craig Newmark Philanthropies, and the Park Foundation.” If we add the European Union and the British and German foreign offices, this list covers the major funders of almost every fact-checking and anti-disinformation organization. The Trusted News Initiative (TNI) was founded in 2019 as an “alliance of news media, social media, and technology corporations” to “combat disinformation” and is closely linked to the British government via the BBC. Unlike the EFCSN or the IFCN, the TNI does not certify organizations; instead it coordinates incumbents, such as Microsoft, the European Broadcasting Union, Twitter, YouTube, Radio Canada, the Australian Broadcasting Corporation and Agence France-Presse, to move against “disinformation” as quickly as possible across as many channels as possible.
These three organizations exist to certify or coordinate fact-checkers with governments and platforms, but what of the fact-checking organizations themselves? The first five official fact-checking partners of Meta in 2016 were PolitiFact, Factcheck.org, Snopes, ABC News, and the Associated Press. As the ecosystem expanded and professionalised, a galaxy of smaller and more specialized groups sprang up, with the larger fact-checking organizations often using them as their own sources of ground truth. A few examples of these organizations include the Atlantic Council’s DFRLab (founded 2016, it is funded by the US State Department, National Endowment for Democracy, and UK Foreign Office, among others), the Global Disinformation Index, a UK nonprofit that builds advertiser blacklists (founded 2018, funded by the UK Foreign Office, the European Union, the German Foreign Office, the Open Society Foundation, and the US State Department, among others), and the Stanford Internet Observatory (founded in 2019, it was funded by the National Science Foundation, before being defunded when Republican politicians realized it kept labelling conservative arguments as misinformation).
Ironically, the best descriptor of fact-checking organizations uses their own terminology: they are professional purveyors and enforcers of malinformation, that is, true information that gives a false impression. There are many ways, even sticking strictly to factuality, to deceive, of which the most important to fact-checkers are decisions on which claims to fact check, how loosely to interpret claims (no one speaks with perfect precision all the time, so you can fact-check the same imprecise but broadly correct statement as mostly true or mostly false), which authorities to defer to as sources of ground truth (you can find credentialed experts to say almost anything), what context should be taken into account, and what the statement is ostensibly implying (so the same statement can be declared either true or malinformation).
Those trusting fact-checkers tend to be more correct than a random person on the total set of possible questions, but extremely and systematically misinformed on key issues of public importance, such as virulence of COVID-19 or the number of unarmed blacks killed by police.“
(https://arctotherium.substack.com/p/the-closure-of-the-internet)
Did Free Speech Suppresion Work ?
Arcotherium:
“The closure of the Internet did not accomplish what its architects intended. It failed to restore the epistemic authority of the academia/media/NGO network. It did not succeed at its top priority of keeping Donald Trump out of office. It has not stopped the rise of right-wing politics in Europe and the Anglosphere. And it did not keep the population compliant with the ever-changing winds of public COVID policy (RFK Jr is in the White House). The reason why it failed at all of these things is quite simple: all of them are downstream of people rationally responding to real-world issues, not bovine masses imbibing troughs of disinformation.
To quote page 249 of Jacob Siegel’s The Information State:
Mass censorship and manipulation did not convince vaccine skeptics to reconsider their position. Instead, it appeared to redouble their doubts. The same was true for efforts at deradicalizing Trump supporters. Perhaps it was not a coincidence that the crusaders against disinformation tended to come from fields like counterterrorism, journalism, academia, and epidemiology that shared records of failure in recent years. If information control failed to achieve its putative goals, it did keep many people employed, offering them back some of the authority that their professions had squandered.
By removing harmful ideas, the war on disinformation was supposed to rebuild people’s trust in their political system. Without mis- and malinformation polluting their minds, they would come to see the party and its affiliated institutions as legitimate and good. Instead, it accelerated a debilitating collapse of legitimacy across American society. In a Gallup poll from 2022, at the height of the information state’s power, Americans expressed record-low levels of confidence in their institutions. Only 23 percent said they had a great deal of trust in the presidency, down 15 percentage points from the previous year. Congress was at 7 percent. Newspapers hit an all-time low of 16 percent, and just 5 percent among Republicans. Curbing Russian influence was always more of an excuse than the real purpose, but it is worth pointing out that the closure of the Internet failed at that too. This was always inevitable, because Russian influence operations were approximately irrelevant on the major Western platforms in 2015 and so there was not much to curb. It did, however, make organic pro-Russian sentiment much more common on the domestic Western right. It did so mostly indirectly, via negative polarization, because Western rightists correctly saw “fighting Russian influence” as an excuse to suppress them and delegitimize the 2016 election. Russia was also painted, by the censors themselves, as a paragon of conservative values to justify the crackdown on the right, which naturally made disinterested rightists, not knowing much about Russia beyond what the media told them, drift towards it. Additionally, many Western rightists joined Russian social media platforms, most notably Telegram, after being banned on Western ones (Voskresenskii 2023).
More importantly, the vast majority of pro-Russian sentiment in the early 2010s was not found on the right, but on the left, a relic of the Cold War. Recall Obama laughing at Mitt Romney for declaring Russia a threat in 2012 or the German government’s reaction to Trump’s warning about becoming dependent on Russian gas. And this went double for the anti-establishment left. The purge of the right naturally made the anti-establishment left, and with it pro-Russian sentiment, more prominent. And in part because of its presence on both ends of the political spectrum, foreign policy (and economic) heterodoxy were more permissible than “hate.” Pro-Russian right-wingers were less likely to be banned and suppressed than accounts talking about migration or race or feminism or any number of other topics, provided they downplayed their domestic political views and focused on foreign affairs, and could even expect some degree of solidarity from the left. The closure of the Internet failed to “deradicalize” the emerging online right too. In fact, it backfired. The online right after the purges was more radical, more open to using violence, more anti-Semitic and more fascist. Why?
First, it's important to understand that most dissidents had, and some still have, fairly mainstream political views. Steve Sailer, for example, has argued civilly and rationally for decades for immigration restriction and better selection of immigrants, fewer foreign wars, and less affirmative action. Most dissidents circa 2014 departed from the institutional consensus on a few particular issues—most commonly related to the histrionic and hyper-aggressive variant of feminism common in 2014—but otherwise considered themselves liberals, libertarians, conservatives, moderates, or otherwise non-radicals. James Damore is an example: his only dissident view was that Google's technical hiring process picked up more men than women because men were, on average (he took pains to point out that this did not necessarily apply to individuals) more interested in programming, and not because Google was sexist and needed special hiring tracks for women. Otherwise, he was a self-identifying liberal.
There were certainly “extremists” (not that “extremism” is something that ought to be persecuted or necessarily wrong) who believed some sort of radical overhaul of society was needed, but they were a small and fringe minority within the many heterodox spheres of 2014 and 2015. The “moderates” who were basically OK with the 2012 or 2007 West were much closer to the popular mood than the more esoteric and eccentric radicals, and so far more popular. This popularity made them high-profile. And since they didn’t see themselves as dissidents, and so typically didn’t maintain opsec, they were easy targets and targeted first. With them gone, the radicals were able to take over the sphere.
And it’s not just a question of selection effects. The arrival of mass censorship vindicated the radicals. Many of the purges were framed as fighting online Nazis and anti-Semites or misogynists, and actually getting persecuted in the name of and nominally on behalf of the Jews or of women changed things. It’s one thing to joke about Jews running the world, and it’s quite another to watch one of your friends get fired and debanked by a campaign organized by the ADL. Without these purges, many would have soon moved on to another meme. But once under fire, they dug in. Their views crystallized and hardened. An anonymous poster recalled that “we discovered National Socialism from first principles.”
I also believe the closure of the Internet is essential to understanding why major Silicon Valley figures, who had been loyal Democrats for decades, moved to the Republican Party in the 2020s—many of them are liberal or libertarian-inclined and were horrified by seeing this happen first-hand.
Because the hegemonic leftist institutional consensus is wrong about many easily-observable features of reality, it cannot actually shut down resistance to its authority. Reality bats last. But it did succeed in obliterating the online intellectual ecosystem and with it any rigorous alternative or correction mechanism. So the closure of the Internet did not lead to the resurrection of the deeply flawed institutional consensus but a kind of epistemic wasteland of lowest-common-denominator populism and mind-numbingly stupid propaganda. Contrast Sean Last with Candace Owens.”
(https://arctotherium.substack.com/p/the-closure-of-the-internet)
Examples
Arcotherium:
“But fact-checkers and disinformation NGOs did not always stick strictly to factuality. Often they were simply wrong, not just misleading. A notorious example is Hamilton 68, founded by a former FBI agent and funded by the Alliance for Securing Democracy (ASD, founded 2017 in direct response to the 2016 election). Hamilton 68 had a proprietary list of 644 Twitter accounts that they claimed to be linked to Russian disinformation operations. Hamilton 68 would look at what these accounts were posting about, use this to claim that whatever the topic was was downstream of Russian influence, and then be cited as an authority on this by other fact-checkers and media organizations (“including NBC, CBS, ABC, PBS, CNN, MSNBC, The New York Times and the Washington Post,” plus “Snopes and Politifact”).
This “independent expert research” was then used as a pretext to suppress these discourses.
As it turns out, Hamilton 68’s list of Russian agents consisted of a handful of public Russian accounts (mostly connected to Russia Today) and hundreds of ordinary Americans with zero ties to Russia, mostly conservatives or otherwise right-of-center but including a few dissident leftists and others as well. Twitter knew this internally by 2017, but said nothing; this was only revealed in 2022 with the Twitter files. The Hamilton 68 episode shows how easily claims of foreign influence were weaponized to suppress domestic dissent, and the false narrative that Russian influence campaigns were a major component of Brexit and the 2016 presidential election were crucial to legitimizing the closure of the Internet. Other high profile failures include labeling the lab-leak hypothesis of COVID origins a debunked conspiracy theory in 2020 even though it was always plausible, suppressing a BMJ report on poor medical practices at a contractor working for Pfizer, labelling true allegations that the Biden administration was purchasing crack pipes for smokers as “mostly false,” or labelling claims of Biden’s evident senility as false every year from 2020 through 2024.”
(https://arctotherium.substack.com/p/the-closure-of-the-internet)