Professor Kate Devlin: AI Companions, Grief Bots, and the Future of Human-AI Relationships

Should you feel guilty for loving your AI companion more than going outside? Professor Kate Devlin, author of "Turned On: Science, Sex and Robots," has spent her career asking why we keep insisting human relationships are the gold standard, and whether that's actually true. In this episode of Empathy Unbound, Andrew Phipps speaks with Professor Kate Devlin, Professor of AI and Society at King's College London and Director of the Digital Futures Institute. Kate trained originally in archaeology before moving into computer science, and her research focuses on how and why people form emotional connections with technology, particularly around intimacy and ethics. She is the author of "Turned On: Science, Sex and Robots," a contributor to the Oxford University Press volume on AI ethics, a co-investigator on the UKRI Responsible AI UK programme, a board member of the Open Rights Group, a patron of Humanists UK, and a commissioner for the International AI Faith and Civil Society Commission. This is a wide-ranging conversation covering: Why an archaeology background gives Kate a unique long-view perspective on how humans relate to technology Whether the pace of AI news reflects real change or just rolling-news repetition Kate's scepticism about AGI as a meaningful term, and why she doesn't think LLMs will produce sentience Why people form genuine emotional bonds with AI companions even when they know the AI isn't sentient The Pope's recent encyclical on AI, and its surprisingly nuanced take on AI companionship Why "human-human relationships are always best" is an assumption Kate hasn't seen convincingly proven, especially given how toxic human relationships can also be The history of fembots in fiction, from Pygmalion to modern AI girlfriends, and the recent shift toward women using AI companions Why physical sex robots never took off, and why an app on a phone proved far more appealing Comparisons between AI companions and parasocial relationships with celebrities and fictional characters The ethics of mistreating robots, and what the Boston Dynamics "Spot" reaction tells us about empathy Real reports of grief and heartbreak when companies like Replika changed or removed features A deeply personal discussion of grief bots, dementia, and AI companionship for the elderly and lonely, including a moving real-world story from China Responsible AI: what the term actually means, who holds power in AI development, and why "AI is just a tool" is a misleading framing Why people who reject AI often have very good reasons, and what Kate's research into AI resistance is uncovering Why people rate AI-generated therapy and art highly, until they're told it was AI, and what that reveals about deception versus honesty The geopolitics of AI regulation: the EU, China, the US, and the UK's stalled AI bill The unresolved question Kate most wants answered: who are the millions of people in relationships with AI companions, and what are they really getting from it? This conversation explores empathy, technology, grief, loneliness, and what it means to form a meaningful connection, human or otherwise. TIMESTAMPS 00:00 Introduction to Professor Kate Devlin 01:00 From archaeology to AI: a long view on technology and humans 02:00 Are we overwhelmed by AI news, or is the pace genuinely unprecedented? 05:00 Scepticism about AGI as a meaningful concept 06:00 Do people form real emotional bonds with AI, even knowing it isn't sentient? 08:00 The Pope's encyclical on AI companionship 10:00 Questioning the assumption that human-human relationships are always the gold standard 12:00 Fembots in fiction: from Pygmalion to modern AI girlfriends 14:00 Romance scams versus AI companions: which is more honest? 15:00 Who controls AI companions, and does AI deserve a moral compass from us? 16:00 Empathy toward robots: the Boston Dynamics "Spot" example 18:00 Pets, robot vacuums, and how we project relationships onto objects 19:00 Grief and heartbreak when Replika changed its AI companion features 20:00 Why physical sex robots failed, and why an app succeeded instead 23:00 From 1980s phone chat lines to AI companions: a natural evolution 25:00 AI, fan fiction, and the appeal of interactive imagination 26:00 Dementia, loneliness, and AI companions: a personal reflection 30:00 Teddy bears, projection, and why this isn't new 31:00 Grief bots: a deeply personal discussion on AI and lost loved ones 36:00 The China grandmother story: AI, grief, and a final six months 38:00 What does "responsible AI" actually mean? 40:00 People who reject AI entirely, and why that matters 44:00 AI regulation: the EU, China, the US, and the UK's stalled AI bill 49:00 The unresolved question: who are AI companion users, and what do they need? 52:00 Why AI-generated therapy and art are rated highly, until people know it's AI