“Who will monitor the monitors”?

John O. McGinnis

Elon Musk’s takeover of Twitter may be the most practically consequential event for free expression in decades. He is clearly committed to disseminating the widest possible range of views and has the technological acumen to do so in the most effective way. Unfortunately, while effective free expression relies both on technology and law, it also depends on the broader social culture, and his innovations and success are threatened by the decline in our culture of free expression.
That is very unfortunate, because never have technology and American law so much favored the free exchange of information. The internet has disintermediated and thus opened up ever more decentralized avenues for expression. One no longer needs an institution, like a radio station or newspaper, to publicize one’s ideas or access the widest range of opinion and analysis.
When I was young, it was much harder to obtain that range. Three television networks rationed the news in half-hour programs with a remarkably similar white-bread view of the world. Newspapers offered more varied perspectives but still were limited in their pages of coverage. Magazines of opinion like National Review and The New Republic widened the national range of thought but appeared only twice a month and rarely engaged one another across the political divide.
The scarcity of offerings at the time even affected my career choices. On leaving college, I wanted to be an essayist, but thought that the prospect of landing a job at one of a small number of outlets was hopeless. I went to law school instead.
Contrast that world with ours. Sites like this one have allowed me and others to fulfill our youthful dreams. But more importantly for society, ideas and previously buried facts pour onto the internet by the hour. That information then reaches the audience in a convenient form, whether in postings, podcasts, or videos. But Twitter (and to a much lesser extent Facebook) still plays the essential role of a clearinghouse, allowing short-form summaries of longer-form content. It also makes it easier for people of different views to quickly engage with one another. Twitter itself now also permits the expansion of the content that was previously delivered in 280-character bu-rsts. These threads—a form of bullet points—have the-mselves become a vibrant new method of information exchange. For instance, a scholar of Chinese politics has been using these methods to illuminate what is going on behind the scenes in that totalitarian nation. Its distributed intelligence also gets out in front of legacy media: it has been describing the astounding latest advance in AI, ChatGPT, well in advance of major newspapers. And soon, it appears, we may not even need the threads.
Of course, Twitter permits the distribution of nonsense as well. But it also provides the tools for users to avoid people and messages they do not want to see. They can block those who want to follow and comment and their messages. They can mute those whose messages they wo-uld like to hear from less. Thus, while Twitter creates a forum for exchange, each of us customizes that forum. Nevertheless, for anyone who wishes, it allows them to get outside the ideological bubble that many would otherwise inhabit in their neighborhoods and workplaces.
I have focused only on information exchange about matters that affect politics and policy, but the internet and Twitter in particular also offer riches on science, the arts, and almost any other subject conceivable. The old information order catered mostly to people who shared interests in common with many others, because the narrow bandwidth of print and television made that kind of publication the only profitable avenue. But now it is much easier to gather a specialized audience cheaply. The result has been a cornucopia of information nuggets of every shade and function.
The technology associated with the internet benefits from a legal regime in the United States that has never been more resistant to government censorship. Our law is absolutely clear that government cannot restrict speech outside of a very few categories such as fighting words, libel, and speech that presents a clear and present danger of leading to substantive evil like a riot. Sadly, even many supposedly free nations have far more speech-restrictive regimes. Europe is already threatening Musk, but technology permits Twitter to apply different rules in different places.
The Decline of Free Speech Culture
Nevertheless, our waning free speech culture obstructs the new age of information exchange. The great and good have greeted Musk’s takeover by redoubling their call for Twitter and other avenues of exchange to prevent “disinformation.” While that sentiment has not been translated into law as it has in Europe, it nevertheless hobbles Twitter, as Musk recognizes. Not all information people provide is true, but a core idea supporting free expression is that the assessments of truth, particularly about policy and politics, are best made in a decentralized manner. Even if we cannot be confident the people will sift the wheat from the chaff, a premise of the First Amendment is they are more likely to do so than some centralized authority, given the biases of those who control that authority.
Even though the First Amendment does not restrict private actors, the cultural concern about the danger of centralized authority does not disappear when an authority other than government makes the decisions in the form of content moderators. That danger becomes acute when Twitter or Facebook has substantial market power. Those who are banned from Twitter have few other avenues nearly as good to allow others to access their work on the internet. Consider this analogy: if elders of a small town four hundred years ago had enough power to persuade people to ostracize those with whom they disagree, free speech would not be effective even if the state did not formally censor. That example shows that the exchange of ideas requires not only favorable law, but a favorable culture, one in which people widely accept the dissemination of ideas they dislike and even consider to be false. In such a culture, there is confidence that truth will win out in the long run.
Musk faces a culture increasingly hostile to this view. The mainstream me-dia in particular has emphasized the danger that Mu-sk’s takeover will lead to a disinformation dystopia as he changes Twitter’s monitoring process. That stance is not surprising when one considers that Twitter is a competitor of established institutions. If Musk succeeds in making Twitter an ever better boulevard for free and decentralized expression, legacy journalism is at risk, because it will be easier to find better experts and more vibrant and original opinion. Twitter is an engine of disintermediation, and institutions resist disintermediation for existential reasons.
Such hostility to upstart methods of communication has characterized legacy media’s reaction to technological innovation in the past. After the French Revolution, for instance, there was an outpouring of new pamphlets on current events. But the Paris Book Guild reacted by petitioning the government to require that any book published in France receive the imprimatur of one of the publishers in the booksellers’ guild:
We request, sir, that you glance over it and lend all your influence to our demands. From these abuses of the freedom countless persons who can barely read have established and maintain shops in every quarter of the capital, hanging over their door their name and the title of Bookseller, which they have no scruple about usurping. France will soon be infected by the sale of bad books if everyone is free to do business as a bookseller.
The media also has a political valence—to the left. Twitter on the other hand empowers people of all ideologies. Thus, just as the mainstream media tends to favor restrictions on spending for political speech at election time because such funding can counteract its one-sided influence, so it is largely hostile to a forum that will disturb its power to set the national conversation.
Academics are almost uniformly on the left, which helps explain why a good number are threatening to ditch Twitter in favor of Mastodon, another site for information exchange. But like those celebrities who threaten to leave the nation when a candidate they oppose wins the presidency, the follow-through will likely be limited, because that site is technologically inferior and has a much more limited reach. One can also expect partisan attacks, like a recent one from the White House, because Twitter’s power of disintermediation reduces the hold that left-leaning institutions like the mainstream press and academia have on the national conversation.
This political bias explains why the media has given little attention to the decision of Twitter’s previous regime in 2020 to suppress any mention of Hunter Biden’s laptop on the grounds that this was disinformation. The claims about the laptop and its contents have been verified. Worse still, it has been shown that among the reasons for the suppression was the undue credulity that apparently left-leaning Twitter monitors gave to guidance by the government. Moreover, Bari Weiss has recently shown that Twitter itself decided to mute a variety of conservative and other heterodox voices so that their messages would not get much attention. Managing the presence of tweets on the basis of their ideological content undermines Twitter’s essential promise as a place of free exchange—where the people themselves do that kind of amplification and muting by their decisions on retweeting information.
These incidents underscore the danger of closed systems of monitoring for disinformation where the monitors have substantial market power. The aphorism quis custodiet Ipsos custodes? translates today as “who will monitor the monitors”? Who will make sure that monitors themselves are not victims of disinformation—perhaps willing victims if the disinformation accords with their preexisting biases? Biases can be powerful if monitors are drawn from an ideologically orthodox bubble of elite-university graduates.
Musk has an excellent plan for addressing this problem. Building on an initiative already in the works at Twitter called Community Notes, he is creating a structure where individuals outside the company can add context to surround tweets if they are false or misleading.
This approach is far better than having some designated monitors at Twitter suppress “disinformation.” First, the method of referring information to a relevant community assesses whether that information is likely to be false in a more decentralized manner, counteracting the biases that a small group may have. Second, rather than eliminating the information, Community Notes provides context and additional information. It is better to combat disinformation with good information than with suppression, which does not eliminate the thought in the body politic, but merely sends it to lurk underground where serious refutation is less likely.
Most importantly, the premises of Musk’s concept are more consistent with a free, democratic culture. The answer to bad speech is more speech. And the decision about whether the speech is false is best made not by a centralized group of experts, but by more decentralized mechanisms, even if experts may gain more influence by virtue of an authority that must be continually earned.
Elon Musk may well lose money from his buyout of Twitter. But no one is more likely to succeed in advancing our culture of free speech, because he has the resources, the technological expertise, and, above all, the impulses of a free, if eccentric, man. The friends of liberty should wish his venture every success.