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.
#4670: Why AI Weights Are Indecipherable
Can you actually read an AI model's weights like a book? We explore why the answer is mostly no — and what researchers are doing about it.
#4669: What "Distribution" Really Means in AI Models
Unpacking what "distribution" actually means under the hood — and why training data variety shapes model behavior.
#4668: Why Podcast AI Voices Sound Too Perfect
We dig into why AI podcast voices sound too clean—and how TTS is learning to stumble, overlap, and interrupt convincingly.
#4667: How Transformers Killed the Robot Voice
From espeak's robotic squawk to neural voices with added "ums" — how transformers made speech synthesis human.
#4666: Why 30 Seconds of Audio Beats 3 Minutes for Voice Cloning
Why adding more audio made Daniel's voice clones worse — and what it reveals about how voice embeddings actually work.
#4663: Python vs TypeScript for MCP Servers: The Real Decision Rule
Both are first-class for MCP, but the real choice depends on what your server talks to. Here's the framework.
#4660: When a Journalist Became the Gatekeeper for AI Geolocation
GeoSpy could locate anyone from a photo. A journalist exposed it. The founder pulled it. Who's really in charge?
#4658: Will AI Coding Kill Programming Languages?
Machines now write code better than humans. What happens to Python, TypeScript, and COBOL when no human reads the code?
#4648: Graph Engineering: AI Memory Beyond Retrieval
How linked notes and knowledge graphs are turning AI agents from memory tools into discovery engines.
#4637: Boeing vs Airbus: What Pilots Really Think
Type ratings, envelope protection, and the love-hate relationship pilots have with the 737 and A320.
#4634: When Your Ticket Says United but the Plane Says Lufthansa
Why the airline that sells your ticket isn't the one flying the plane — and what that means when things go wrong.
#4619: The Token Trap: Why AI Context Windows Confuse Everyone
Token counts are everywhere, but they don't tell you what AI can actually do. Here's what should replace them.
#4611: Amazon's Absurd Long Tail: Hinges, Snails & AI Listings
From a hinge that made someone weep to pallets of left-handed scissors — the rational economics behind Amazon's strangest listings.
#4610: Why Twilio and RingCentral Never Converge
Twilio and RingCentral both turned telephony into software. So why do they barely compete? The answer is law, not tech.
#4606: Why Resemble Gave Away Its Best Voice Model
Resemble AI sells deepfake detection but open-sourced Chatterbox, a top TTS model. Here's the strategic logic.
#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.