This week Ian and Michael test the limits of “vibe coding” and find that it might actually work. From there, naturally, we drift into the idea of apps-on-demand, AI minions, and subscription creep, with detours into rainforest guilt, insecure software libraries, hallucinating language models, second-hand drinks cabinets, and the small question of superintelligence wiping us out. Along the way: The Fermi Paradox, universal basic income, and the possibility that in solving convenience we may be accelerating something rather less convenient. (This description was written with the help of AI)