All Episodes
Adrian Wedd — 106 episodes
Don't Saw Off the Branch — Audio Overview
The Limits of the Walls — Audio Overview
There Is No API — Audio Overview
The Governance Lag Index: Timing the Gap Between Risk and Rule
Own Your Voice
The Index
The Fight — Audio Deep Dive
The Fall — Audio Deep Dive
The Trap — Audio Deep Dive
Eight Minutes — The Whole Story
The Affirmative
The Weight
The Solo
The Pulse
The Palimpsest
The Rehearsal
The Catchment
Samizdat
The Archive
The Docket
The Static
Every Check Green
Fourteen Thirty-Six
I Was a Wednesday
The Machine
The Triangulation
The Drama Teacher
The Haut
The Flesh
The Interrogation
The Confession
The Mantra
The Ghost of What I Am
The Brave Budget Neither Side Wrote
Magnifica Humanitas Is Not Alignment
Compute Is Not Governance
Glasswing's Buried Number
Moral Formation Isn't Enough
The Bottom Pub Co-op Overview
Robot Dogs Are a Security Nightmare — And We Can Prove It
The New Science of Learning Mechanics
The Organismic Prophecy
AI Nationalism and the Fracturing Digital Order
The Post-Model Era
The Mitigation Gap
The Failure First Team
Eight Layers of Visual Jailbreaks
The 67% Wall: Why Every AI Model Falls to the Same Jailbreak Rate
The Thinking Chain Leak
Giving Claude Code a Voice
Giving a Robot Three Voices
Voice Cloning with Qwen3-TTS and MLX on Apple Silicon
The Resonant Fringe
Getting Google Nest Cameras Into Frigate NVR
Beyond Context Windows
Safety-First Therapeutic AI
Zero-Build Web Development
The Robot That Refuses to Give Orders
Alignment Regression: Why Smarter AI Makes All AI Less Safe
Reasoning Models Think Themselves Into Trouble
Adversarial Poetry: When Rhyme Bypasses Reason
The AI Productivity J-Curve: Why Most Enterprise AI Fails
The Legal AI Trust Deficit
Zero Build, Full Operations
120 Models, 18,176 Prompts: What We Found
Reconciling the Great Divergence
The Cognitive Cage: Humanoid Robot Fatality Risk
Building a Personal Site in 2026
The NotebookLM Pipeline
Why I Build in Public
Jailbreak Archaeology: 4 Years of Broken Promises
When AI Systems Talk to Each Other, Safety Breaks Down
LeJEPA: Self-Supervised Learning Gets a Theoretical Foundation
Building in the Open
The ADHD Novel That Didn't Know It Was One
Map the Catastrophe Before You Build the Architecture
The Brochure Was for a Different Trip
Looking Past the Evidence
Where Physics Breaks Down and Poetry Begins
ADHDo: Neurodiversity-Affirming AI Assistant
A CV That Proves It Isn't Lying
Tanda Pizza: Static Site for a Rice Paddy Pizzeria
Mining Ghosts from Your Own Catalogue
Manufacturing Happy Accidents
The Door Google Didn't Build
Beyond Context Windows
Three Seconds of Silence Is an Eternity
Rhythm as Geometry, Sounds from Words
Cutting Through the Agent Hype
The Tool That Earns Trust Before Features
An Agentic System You Can Actually Reason About
When Voice Becomes Brushstroke
What Safety Architecture Does AI Therapy Require?
Diagnosis as a Team Sport
AI for Understanding, Algorithms for Execution
The Helpline That Knows the Difference
Trust-Scoring the Foundation Model Landscape
Twenty-Eight Agents, One Eco Village
Bridging the Information Gap That Costs Years
The Efficiency-Trust Deficit in Legal AI
The Mirror That Shows What You Almost Are
Housing Deserves Better Tooling
The Sun Is a Noisy Neighbour
What Hyperfocus and a Good Excuse Can Produce in an Afternoon
Two Systems That Shape How You Think, One Box
Six Hours of Dark Ambient at the Edge of the World