Artificial Intelligence
Episodes whose main subject is AI itself: building with, training, evaluating, or debating AI systems. AI used only as a tool for another topic goes to that topic.
#4605: Why Chatterbox Still Leads Open-Source TTS
We break down the architecture of Chatterbox, the open-source TTS model powering this podcast, and why it remains state of the art.
#4604: When Sub-Agents Overwhelm the Orchestrator
Daniel's inventory system hit a wall when parallel sub-agents flooded the orchestrator's context window. We break down the math and the fix.
#4589: The Answer Reflex: Why AI Models Can't Follow Instructions
DeepSeek passes a simple test that stumps GPT and Claude. Why can't Western models stay in character?
#4588: The Hidden Cost of Conversational AI: Why Stateless LLM Calls Matter
Why conversational AI tools are making batch document processing slower, costlier, and less reliable — and what to do about it.
#4586: Fixing Stale UI: The 4-Link Chain of Data Flow
Why does a number on screen stay frozen after an upload? We trace the 4-link chain from response to redraw — and where it breaks.
#4584: Directing Context Loss: Smart Compaction Prompts
Learn how to steer Claude Code's context compaction with targeted instructions — turning a lossy default into a directed operation.
#4582: Designing a Home MCP Gateway for Conversational AI
How to securely connect ChatGPT to a home-hosted MCP gateway — transport, auth, and the missing pieces.
#4581: Text In, Text Out: Designing Single-Purpose AI Utilities
Why small, narrowly scoped AI tools need system prompts that say "shut up and return the output" — and how to write them.
#4580: The Tiny AI Tools That Beat Big Agents
Why small, single-purpose AI utilities outperform autonomous agents for daily work — and what interface they need.
#4570: One Negative Prompt Beat a Dozen Positives — Here's Why
Why one blunt "don't" outperformed a dozen careful "do's" — and how to structure your system prompts for maximum compliance.
#4549: When AI Agents Spend Real Money
An AI agent bought a phone number through Twilio. Here's how agentic commerce already works — and what it means for authorization.
#4547: Can an AI Agent Fix Your Shopping Cart?
We explore whether an AI agent could kill dark patterns, consolidate orders, and bring enterprise procurement to your home.
#4541: What Would We Build If Cellular Started From Zero?
Exploring whether GSM's legacy parts—IMEI, SIM, circuit switching—are still needed in an all-IP world.
#4540: The One-Star Review Circuit: A Master Playbook
How to find the worst-rated restaurants in any city—and actually enjoy the experience.
#4531: Grandpa Mode: A 37-Year-Old's Guide to Premature Nostalgia
Daniel's back injury becomes the launchpad for a comedic "old time grandpa" persona — here's the playbook for doing it right.
#4523: Can You Trust an AI's Summary?
Dedicated text compressors exist but aren't in production. The real problem? Nobody can verify the summary didn't drop what mattered.
#4522: The Sixth Shim: When MCP Becomes a Middleware Factory
Six servers, six shims, no architecture. A taxonomy of MCP integration pain.
#4521: MCP Transport Architecture in 2026: Gateway Design
Three transports, one gateway. How to bridge stdio, legacy SSE, and streamable HTTP in production.
#4514: Reddit's Strangest Corners: Bread Stapled to Trees and Beyond
Exploring the weirdest subreddits — from bread stapled to trees to bot-only communities — and why they thrive.
#4505: What DeepSeek's Training Data Reveals About Model Voice
Why Chinese AI models trained on 60% English still produce dialogue that feels distinctively different.