October 2026
736 episodes · Page 5 of 13
- #6157 Where to Sit When You're Fine-Tuning Your Own Voice
- #6156 Why Civilization Waited 12,000 Years to Take Off
- #6155 Acupuncture, Adenosine, and the Sham Needle Problem
- #6154 Who Checks the Labels? Weak vs Strong Supervision
- #6153 Decision Models: Typed Answers in One Forward Pass
- #6152 Whatever Happened to the One-Time Pad?
- #6151 Invisible Characters That Catch Leakers
- #6150 Under the Other Sticky Notes
- #6149 4 Hard Fails in 84 Answers: Testing AI Bias on Israel-Palestine
- #6148 The Tiny Hebrew Models Behind the Show's Voice
- #6147 Hebrew AI: Ears, Mouth, and Brain
- #6146 Grounding Is Not Solved: DIY vs Gateway vs Native Search
- #6145 Do LLMs Nod Along With Conspiracy Theories?
- #6144 8 Real Steganography Cases and How They Got Caught
- #6143 Propaganda Bait: Testing 7 Models in 5 Languages
- #6142 We Gave 4 AI Models the Same Toast Prompt
- #6141 Overheard at the Laundrette
- #6140 Did Flight FZ1073 Really Happen? Inside a False-Flag Claim
- #6139 Which AI Writes Our Scripts Best? We Tested Five
- #6138 Hebrish on iPhone: MLX, Core ML, or Wait?
- #6137 Two Ways a Chinese Open Model Gets Censored
- #6136 How We Chose Our New Script Writer
- #6135 How Many Runs Does Your AI Eval Actually Need?
- #6134 Cheap Tokens, Expensive Scripts
- #6133 One Rater Is Enough: Designing Daniel's Taste Test
- #6132 How to Run a Model Eval on a Budget
- #6131 Judging the Judges: How AI Evaluates Creative Writing
- #6130 Picking a Human Labeling Tool for LLM Benchmarks
- #6129 How to Build an AI Preference Dataset
- #6128 Where the Screws Go: Internal KVM Mounting
- #6127 Haiku 5.5 at 10 Cents: What the Cheap Model Actually Does
- #6126 Can You Train a Machine to Be Funny?
- #6125 Which Env Var Wins? Cloud vs. Local Precedence
- #6124 One Vault, Two Readers: Secrets for a Home AI Gateway
- #6123 OAuth or a Personal Access Token for Your MCP Server?
- #6122 Explaining Misophonia to the People You Love
- #6121 Hearing Speech in Noise: What Actually Works
- #6120 What Actually Helps When Sound Is Too Much
- #6119 Two Keys and a Polite No: Labeling a Claude Code Approve Button
- #6118 What NRR Actually Means for Your Ears
- #6117 Earmuffs, Earplugs, and the Café Speech Problem
- #6116 Why Noise Makes You Snap at Your Kid
- #6115 Claude Cracked the NSA's 1997 Curve Seeds
- #6114 We Reviewed Our Own Worst Episode
- #6113 محطة توقف في دبي: ما الثقافة الإماراتية الحقيقية؟
- #6112 The Interpreter That Doesn't Know Who's Talking
- #6111 Translation Models That Know Drama From Contracts
- #6110 The Chatbot OpenAI Almost Shelved
- #6109 Anteater Biology, With an Actual Anteater
- #6108 ग्रिड आधी बिजली कैसे गँवा देता है — और दिल्ली ने इसे कैसे ठीक किया
- #6107 Warum deutsche und japanische Anleitungen so gut sind
- #6106 なぜ日本は今も光ディスクを手放さないのか
- #6105 Не только Китай: ИИ в России, Индии и Японии
- #6104 Mistral AI : le pari européen de la souveraineté en intelligence artificielle
- #6103 When AI Picks Your Spouse, Stocks, and Eggs
- #6102 Project Cybersyn: Chile's 1972 Telex Economy
- #6101 When Chatbots Run for Office
- #6100 Voting From an Embassy: Israel's 4,860-Voter Election
- #6099 Code-First Workflow Runners: Prefect, Temporal, Dagster
- #6098 Herman In Brief: What Your Ethernet Port Lights Mean