3.2.1 émotion podcast
Conversations on the structural failures of digital connection, and what rebuilding the emotional layer of technology looks like in practice.
Why your EQ score is backward
EQ treats emotional intelligence as a personal score on a bell curve. The data says it’s a shared infrastructure — and we’ve been judging drivers without paving any roads.
A deep dive on why treating emotional intelligence as a personal EQ score is structurally backward. We trace the 1995 Goleman bestseller that turned EI into a quotient by deliberately mimicking the IQ brand, unpack why alexithymia (~1 in 10 people, structurally unable to identify their feelings) is an infrastructure failure rather than a personal deficit, look at the 2025 WHO Commission on Social Connection (1 in 6 people experience persistent loneliness; ~871,000 deaths a year) and the May 2026 Foken & Dunn longitudinal study on AI-companion crowding-out, and decode the 4-gesture roadmap (feel, express, understand, meet) that 3.2.1 Emotion proposes for building EI as shared infrastructure instead of measuring individuals against a bell curve. Augmentation, never substitution: the AI is the microscope, not the destination.
Open episode →AI companions are making us lonelier
The Foken & Dunn longitudinal study (n=2,149, 12 months) shows AI chatbots ease loneliness in the moment but deepen it over a year — by crowding out the real human friction that actually cures it.
A deep dive on the May 2026 Foken & Dunn longitudinal study tracking 2,149 adults across the UK, US, Canada, and Australia for 12 months. Two directional arrows in the data: lonely people turn to chatbots (expected), and then they report greater isolation four months later (the trap). We unpack the four ingredients of real connection — vulnerability, reciprocity, presence, safety — and why AI can only fake the last two. Then we decode the architectural divide between substitution AI (success = you don’t need humans anymore) and augmentation AI (success = you close the app and message a real friend). Three concrete rules for using AI as an instrument, never a partner.
Open episode →How emotional signatures solve swipe fatigue
Why algorithms predict matches no better than random chance, what Gottman’s Love Lab actually measured, and the case for matching by emotional signature before face.
A deep dive on why 78% of dating-app users feel emotionally exhausted, the four communication patterns Gottman uses to predict divorce with 90% accuracy, and how attachment theory plus the pursue-withdraw cycle explain the chaos of modern dating. We then look at the architecture of matching by emoDNA / emotional signature instead of a half-second swipe on a photograph.
Open episode →Why text messages ruin relationships
The 25-point overconfidence gap: senders think they're understood 88% of the time, receivers decode the tone right only 63%. The misunderstanding is a feature of the medium.
A deep dive on the math of plain-text communication. We unpack the 2005 Kruger study, why emojis are a "crayon drawing of emotional reality", and the case for treating digital messages as multi-sensory emotional scenes that travel well, not fast, including the .emo file format, temporal messaging, and the augmentation-never-substitution rule for emotional AI.
Open episode →Loneliness is a digital design failure
Why the most connected society in human history is also the loneliest, and the cost of confusing visibility with intimacy.
A case that loneliness is not a personal failing but the structural outcome of platforms optimised for engagement over empathy. We map the public-health cost (mortality risk equivalent to smoking 15 cigarettes a day), why visibility is not intimacy, and what an emotional infrastructure for digital connection would actually require.
Open episode →