Humans are Tribal and Judgy, and AI is Exploiting It (with Yarrow Dunham)

Why do kids form biases almost instantly? Why do people punish unfairness even when it costs them? And why do social media and AI seem to make all of this worse?In this episode of The Rip Current, I sit down with Yale psychologist Yarrow Dunham to unpack his many years of research into how humans form groups, enforce fairness, and turn tiny assumptions into lifelong beliefs. We talk about children, tribalism, polarization, altruistic punishment — and what happens when these ancient instincts collide with modern technology and generative AI.This conversation explains a lot about why the world feels broken — and why it doesn’t have to stay that way.

The Rip Current by Jacob Ward is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

00:00 — How Fast Bias Forms (Even in Kids)Yarrow Dunham explains how children develop group preferences almost instantly — and why bias doesn’t require ideology, history, or teaching.02:05 — The “Minimal Group” Experiment ExplainedWhy simply assigning people to meaningless groups reliably creates favoritism, memory distortion, and preference.04:45 — Is Bias Innate or Learned?What research with infants suggests about early-emerging social preferences — and why “innate” is the wrong shortcut.07:10 — Why Humans Are Wired for CooperationHow long-term reciprocity with non-kin sets humans apart from other animals — and why group loyalty evolved.09:55 — How Big Can a “Tribe” Be?From hunter-gatherer bands to modern identities: nested groups, concentric loyalties, and flexible belonging.12:40 — When Bias Becomes DangerousWhy liking your group doesn’t automatically mean hating others — and what turns neutrality into hostility.14:30 — The Surprising Power of Expected CooperationA key finding: bias toward out-groups collapses when people expect to work together — even before contact.17:10 — Why This Matters for PolarizationHow declining cross-group interaction fuels political and social division — online and offline.19:25 — Kids, Fairness, and Punishing UnfairnessWhy children will pay a personal cost to enforce fairness — even when they’re not directly involved.22:10 — Altruistic Punishment and Moral OutrageHow fairness enforcement connects to adult politics, ideology, and “voting against self-interest.”25:05 — Fairness vs. MeritocracyWhy kids start out egalitarian — and how societies train them to accept inequality over time.27:45 — Status, Race, and Group PreferenceHow high-status groups override in-group bias — and what research shows in the U.S. and South Africa.30:40 — The ‘Default Human’ ProblemWhy systems (and societies) treat white men as the baseline — and the real-world consequences of that bias.33:20 — What Social Media Gets Exactly WrongHow algorithms amplify group identity and hostility — creating a perfect polarization machine.36:05 — Why AI Feels Like It’s “On Your Side”How generative AI triggers ancient social instincts by mimicking agency, affirmation, and belonging.38:50 — The Danger of Sycophantic AIWhy flattery and agreement are design choices — and how they short-circuit growth, challenge, and truth.41:40 — The Feedback Loop That Makes Bias WorseHow AI trained on human bias reflects it back as authority — reinforcing mistaken beliefs at scale.44:30 — Can AI Reduce Bias Instead of Amplifying It?What psychology suggests about indirect contact, imagined cooperation, and redesigned systems.47:10 — What Actually Works to Reduce BiasEqual-status cooperation, shared goals, and why exposure alone isn’t enough.50:05 — The Real Fix Is StructuralWhy individual goodwill isn’t enough — and how institutions shape who meets whom.52:40 — Final Takeaway: Bias Is FlexibleThe hopeful conclusion: group boundaries can be redrawn quickly — if we choose to design for it.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theripcurrent.substack.com/subscribe