Episode Summary
Executive Summary: The episode examines how misinformation spreads, why people are vulnerable to it, and what can be done to resist it. Dr. Sander van der Linden argues that misinformation includes both false and misleading content, spreads through repeated exposure and social networks like a virus, and can be countered through pre-bunking, media literacy, and scalable interventions. He also discusses the mixed role of platforms, schools, traditional media, and AI.
Main Topics: Defining misinformation vs. disinformation: Van der Linden explains that misinformation includes both outright falsehoods and misleading content such as half-truths, false framing, and out-of-context claims. Disinformation is misinformation spread with intent to deceive or harm. How much misinformation people see: Exposure varies widely by platform and media habits, with social-media-heavy users seeing more. He estimates a range from about 1% to 20% of a person’s media diet depending on definition and context, and says current measures likely underestimate total exposure. Why people are vulnerable: He describes cognitive and social mechanisms like fluency, illusory truth, and continued influence, plus personal risk factors such as political extremity, heavy social media use, lower analytic thinking, paranoia, and childhood trauma. Misinformation as contagion and inoculation: He compares misinformation to a virus and says epidemiological models can describe its spread. He argues for pre-bunking: exposing people to weakened versions of misleading tactics so they build psychological resistance. Platform responsibility and policy: He says social platforms have been inconsistent: more aggressive during elections and pandemics, but increasingly lax, including backsliding on moderation and leaving up harmful election-related content and deepfakes. Politics and lawsuits have chilled action. AI, deepfakes, and micro-targeting: He warns generative AI can massively scale disinformation, personalized propaganda, and deepfakes, but also notes it can help automate fact-checking, create prebunks, and support research tools. Education, traditional media, and trusted messengers: He advocates embedding media literacy and prebunking in schools, using traditional media and podcasts to educate audiences, and relying on trusted community leaders when government or platforms are not viewed as credible.
Key Arguments: Misinformation is broader than fabricated lies; misleading narratives, logical fallacies, and manipulative framing are central to the problem. Intent is hard to prove, so defining disinformation separately from misinformation is useful but practically limited. Exposure estimates depend heavily on definition; purely fabricated content may be a small share, while misleading information can occupy a much larger portion of media diet. Social media increases exposure because of algorithmic amplification, echo chambers, and the absence of editorial gatekeeping. People are vulnerable because repeated claims feel more fluent and familiar, which makes them seem truer over time. Pre-bunking works better than debunking alone because it prepares people before they encounter the falsehood. Psychological inoculation can scale, but its impact depends on timing, reach, and whether it is integrated into platform design or education. AI is a dual-use tool: it can accelerate industrial-scale disinformation, but it can also support detection, classification, and inoculation. Schools should teach students how misinformation techniques work rather than what to think, building long-term resilience. Traditional media still offers more safeguards than social media, but it should be more proactive in warning and educating audiences about misinformation. Behavioral measurement is the next challenge: researchers still lack strong real-world data on whether inoculation changes what people actually consume or share. Trusted local or community voices may be more effective than institutions alone in countering misinformation in polarized settings.
Data Points: Estimated share of media diet (purely fabricated misinformation): 1% to a couple percent - Van der Linden says fabricated falsehoods may represent only a small portion if defined narrowly. Estimated share of media diet (misleading information): up to 20% - He says broader definitions that include misleading content raise exposure substantially. Research papers published: more than 150 - The host introduces van der Linden’s publication record. Inoculation boost in lab simulation: about 60% - He says a 20-minute lab simulation can increase resistance substantially. Inoculation boost in real-world short ad: about 10% - He estimates a 30-second public message can produce smaller but scalable gains. Media literacy campaign example: Twitter pre-warning on election misinformation - He describes platform pre-bunking before Musk-era policy changes. Historical disinformation campaign duration: over 50 years - He references the tobacco industry’s long-running deception about smoking and disease.
Pivotal Quotes: "It’s both. It’s both outrightly false information as well as misleading information." — Dr. Sander van der Linden: Defining misinformation for the APA report "Misinformation behaves much like a virus if you look at these models." — Dr. Sander van der Linden: Explaining the epidemiological analogy for spread "The gist is actually that, you know, when companies place ads, you know, it’s all about small effects, right?" — Dr. Sander van der Linden: Arguing that small inoculation effects can still matter at scale
Implications: Listeners should treat misinformation defense as a skill set, not a one-time fix. The future likely depends on layered solutions: platform policy, school-based media literacy, AI-enabled detection, and trusted messengers who can scale prebunking before falsehoods spread.