Composed, written, produced and performed by Albee Wilson and Ben Berman.
No AI or samples were used in the making of this song.
Shout out Oscar.

(Fact check for the claim: "fact checks don’t work": Fact checking - defined as  “systematically publishing assessments of the validity of claims made by public officials and institutions with an explicit attempt to identify whether a claim is factual” (Walter et al. 2020) gives individuals the resources to identify and discard misinformation whilst discrediting unreliable speakers. There are however three main issues to this approach: psychological, epistemic and practical. 

There is evidence that factchecking reduces misperceptions, agreement with misinformation and improves factual accuracy (Porter et al. 2019; Cobb et al. 2013; Walter 2020, Nyhan et al. 2020). Though, whilst fact checking does lead to the acceptance of a correction by an individual, it is by no means sufficient in removing the effects of this misinformation on their beliefs and behaviour (Johnson and Seifert 1994; Lewandowsky et al. 2012, Thorson, 2016).  More concerningly the ‘directionally motivated reasoning’ theory puts forward evidence that individuals are more accepting of misinformation that is consistent with their prior worldviews and are worse at accepting corrections which contradict it (Flynn et al. 2017; Walter et al. 2020). There is also mixed evidence of backfire effect where an attitude-dissonant correction exacerbates the effect of the misinformation (Nyhan and Reifler 2010) perhaps when issue salience is high (Flynn et al. 2017), but results are elusive (Ecker and Ang 2019; Wood and Porter 2019). As a result, we must remain sober to the realities of what factchecking can accomplish for defending democracy. 

 A further important consideration is the epistemic role of factchecking. Firstly, objectivity and unbiasedness may be an unachievable goal. Secondly, fact-checking is itself a communicative act and in a Foucauldian sense factchecking itself contributes to constructing ‘regimes of truth’ and ‘social reality’. Thirdly, it is epistemically very hard to state that a causal or future-regarding claim is true (Suomalainen et al. 2025). Finally, selecting which claims to check necessarily introduces some subjectivity or at least a consideration as to how possible it is to check (Suomalainen et al. 2025). This line of argument is perhaps over-philosophical in the face of the very tangible threat of misinformation to democracy. However, it must at least be kept in mind if we are evaluating whether factchecking is in the public interest.

More pressing then is the issue of practicality. It takes more effort to debunk a false claim than make one, and with the internet and AI, it becomes even harder. Indeed Wack, Duskin and Hodel (2024) find that fact checking during the 2022 US election was significantly limited by coverage, speed and reach. Given the resource constraints facing the factchecker and the inexhaustible amount of information online, editorial decisions have to be made over what is and what isn’t checked. Hopefully, AI can assist fact-checkers in this, but it is highly unlikely that this fundamental asymmetry will be changed. Moreover, even a successful factcheck might not actually reach its audience. There is increasing distrust in the media, and fact checkers which undermines the effectiveness of corrections as individuals find it easier to ignore and discount fact-checkers as biased (Brandtzaeg and Følstad 2017). Similarly, directionally motivated reasoning can lead individuals to seek out attitude-congruent viewpoints and avoid dissonant ones (Flynn et al. 2017). Overall, it is unclear how to expose people to fact checks, incentivise people to share them, and get them to accept the fact-checkers authority.

These challenges undermine some of the hope of fact-checking’s impact on democracy, but that is not to say that fact-checks are not in the public interest. There is evidence that fact-checking practices could be used to defend journalism against discrediting (Pingree et al. 2018), improve misinformation discernment (Guess et al. 2020)  and increase self-perceived misinformation identification (Goyanes et al. 2025). So long as fact-checkers remain sensitive to transparency, accountability, unbiasedness and their epistemic responsibility then it is reasonable to think that it will be strongly in the public interest. Finally, if its practices, collaboration, government support and public trust continue to improve then it should also have a stronger impact in the fight to protect democracy.)