The Crisis of Adulthood — with John Burn-Murdoch

Where this was said

AI as a Moderating Force: Reversing Social Media's Polarisation

At 52:00 · chapter starts 49:50

Scott raises a finding from John's research that he finds underappreciated: while social media created deep polarisation, AI chatbots appear to have the opposite effect, consistently nudging users toward moderate, expert-aligned views even when accounting for sycophancy ^{[1] — John Burn-Murdoch "Social media amplified fringe voices because anyone could post and attention was the currency. AI flips that model: trained on mainstream s…" 47:43 Play} . John provides the structural explanation: every new media form has a political valence shaped by its economics and constraints. Printing press and early television both favoured mainstream establishment voices due to capital costs and regulation. Social media broke that constraint — anyone could post, attention was the currency, and the incentive structure rewarded sensationalism and fringe amplification. AI flips the model: training data comes heavily from trusted mainstream sources, and the winning business model is making it useful for business users who need accuracy, not validation. Telling someone what they want to hear is costly if it means they give investors wrong information. So the inner workings of AI reward being correct, not engaging. The result is that AI conversations about politics tend toward straight-down-the-middle moderate answers — a property that, at scale, could meaningfully reduce the polarisation that social media created ^{[2] — John Burn-Murdoch "AI is this really interesting case where we seem to be going back to a moderating, establishment-favoring media... they tend to give you fa…" 49:20 Play} .