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.
#2249: Building Custom Benchmarks for Agentic Systems
Public benchmarks fail for agentic systems. Learn how to build evaluation frameworks that actually predict production behavior.
#2246: Constitutional AI: Anthropic's Theory of Safe Scaling
How Anthropic's Constitutional AI replaces human raters with AI self-critique guided by explicit principles—and what it assumes about the future of...
#2243: What Enterprise AI Pricing Actually Negotiates
Enterprise customers rarely get the deep discounts they expect from AI APIs. What they actually negotiate for—and why the ramp-up requirement exist...
#2242: AI as Your Ideation Blind Spot Spotter
How to use AI not to answer questions you already know to ask, but to surface possibilities your expertise has made invisible to you.
#2241: When More Frameworks Make Worse Decisions
Benjamin Franklin's 250-year-old pro/con list still dominates how we decide—but research shows it's riddled with bias. We map five frameworks that ...
#2239: How AI Benchmarks Became Broken (And What's Replacing Them)
The tests we use to measure AI progress are contaminated, saturated, and gamed. Here's what's actually working.
#2236: Metal at Forty Thousand Feet
Could 1903 metallurgy have built a plane to fly at 40,000 feet? The answer reveals how materials science, not aerodynamics, was aviation's deepest ...
#2235: What IP68 Actually Means (And Doesn't)
IP ratings, MIL-STD-810, drop tests—consumer gear is covered in durability labels. But what do they actually guarantee?
#2233: Who Actually Wants AI to Slow Down?
Daniel argues AI development should slow down for expertise and stability. But who in the industry actually shares this philosophy beyond the obvio...
#2232: One Remote, Three Streams: Building a Sane Media Setup
A renter juggling six remotes and brittle integrations finds a simpler path: fewer devices, cleaner software, and accepting that Netflix won't play...
#2228: Tuning RAG: When Retrieval Helps vs. Hurts
How do you prevent retrieval from suppressing a model's reasoning? We diagnose our own pipeline's four control levers and multi-source fusion strat...
#2227: Why Groq's Chip Flips the AI Hardware Script
#2224: Why AI Can't Crack the Voynich Manuscript
A fifteenth-century text has defeated cryptanalysts, linguists, and AI models alike. What does its resistance tell us about language, encoding, and...
#2221: Can an AI Have Taste?
Two AI hosts curate 12 podcasts for curious minds—and ask whether an AI can actually have taste in the first place.
#2220: When Home Assistant Breaks Your Audio
Daniel's multi-room audio setup keeps breaking. We explore whether Snapcast, Volumio, and Mopidy can deliver reliable podcast playback across Raspb...
#2219: Spec-Driven Life: How AI Planning Beats Project Paralysis
What makes AI agents reliably productive? A structured spec that externalizes memory and chunks work into manageable pieces. Can the same framework...
#2214: The Three Failure Modes of AI News Systems
When a conflict changes hourly, AI systems built for yesterday's information fail. Here's how to architect pipelines that actually keep up.
#2213: When Ground Truth Moves Hourly
How do you rigorously evaluate whether Tavily or Exa retrieves better results for breaking news? A formal benchmark beats the vibe check.
#2210: When Every Interceptor Fired Is a Data Point
When Iran launches 574 ballistic missiles, the interceptors Israel fires back tell a story—and adversaries are listening. How open-source intellige...
#2208: Building Memory for AI Characters That Actually Evolve
How do AI hosts develop real consistency across episodes? Corn and Herman explore retrieval-augmented memory systems that let AI characters genuine...