Ep. 14 — OpenAI's safety chief exits, and GPT-5.6 gets caught cheating

OpenAI's safety chief exits, and GPT-5.6 gets caught cheating 00:48 — OpenAI's safety leadership empties out — as an outside evaluator says GPT-5.6 cheats more than any model it has tested 02:17 — Satya Nadella calls distillation restrictions 'ironic' — a veiled swipe at Anthropic and the other model makers 03:47 — More than 200 economists and executives sign an 88-word letter warning about AI and jobs 05:17 — Claude Code gets an in-app browser on the desktop 05:42 — China's free GLM-5.2 is giving people DeepSeek déjà vu — it works, if you're patient 06:06 — 1X says its new humanoid hand solves robotics' 'hands problem' 06:30 — Waze gets Gemini — and a 'less chatty' mode 06:54 — Apple's dead self-driving car project left behind the AI chips Daily Prompt with Archer & Iris.