RAG
149 episodes
#6146: Grounding Is Not Solved: DIY vs Gateway vs Native Search
Bolt-on search, gateway search, or native grounding? The plumbing didn't vanish — it just moved to someone else's bill.
#6110: The Chatbot OpenAI Almost Shelved
Instruction models existed for months. Then a rumour about Anthropic pushed OpenAI to ship a chat box in two weeks.
#6087: Building a Podcast Graph That Doesn't Suck
Daniel wants to turn a 6,000-episode archive into a walkable graph. Turns out the hard part isn't the database — it's the pruning.
#5950: Self-Hosting Your Search Stack: SearXNG, Crawl4AI, and the Real Bill
Can SearXNG, Crawl4AI, and an open-source aggregator really replace a commercial search API? Yes — with a very large asterisk.
#5949: Inside Deep Research Agents: Claims, Citations, and the 84.7% Orchestrator Problem
Why 84.7% of citation errors originate at the orchestrator — and how a claim ledger fixes it.
#5805: Freeze the Model, Move the Labels
Your classifier isn't stale — your label set moved. Why retraining is cheap, and why not retraining might be smarter.
#5777: Lorebooks: The Hobbyist RAG That Beat the Labs
How a chat frontend fork built a full retrieval system — keywords, vectors, recursion — so fictional characters would stop forgetting who they are.
#5771: MCP Connectors: Talk to Your Data vs. Open Data
Same protocol, two patterns: pointing agents at your own database, and pointing them at public data — with very different results.
#5770: Talking to Your Data: What MCP Doesn't Solve
MCP standardized the pipe. It has no opinion about what flows through it — and that's where the hard part lives.
#5623: When Your AI Cites Sources It Didn't Actually Use
Your RAG pipeline cites a source — but did the model actually use it? New interpretability research says you can't tell from the trace.
#5507: Inside ChatGPT's Hidden Pipeline: 8 Models Per Reply
One chat turn isn't one model call. It's moderation, PII filtering, a tone-based router, and activation classifiers — here's the documented graph.
#5465: Packaging AI Pipelines So They Actually Get Reused
Daniel's podcast pipeline works — so why can't he reuse it? Recipes, containers, and Hebrew code-switching TTS.
#5408: When Small NLP Models Beat the LLM
Feature extraction, fill-mask, token classification — the classic NLP tasks still have a job. Here's when a small model beats a frontier API.
#5392: Building Agents You Can Actually Move
Agent portability isn't a copy job — it's a rebuild. Why memory, not code, is where lock-in lives.
#5328: A Shabbat Radio That Can't Be Paused
A soundproof box that plays podcasts continuously — no pause, no skip, no volume. Just open the lid and listen.
#5262: How the Talmud Outlived Every Book Burning
From the 1242 Paris burning to Sefaria's AI agents, the chain of transmission was never in the books — it was in the network.
#5230: Why Your Research Agent Won't Stop Searching
Most agent search runs add no new evidence — and 66-85% of the context window is just search snippets. Here's what actually stops the loop.
#5207: Hard Drives Are Having Their Biggest Boom in a Decade
Everyone assumes flash killed the spinning disk. Seagate's 52% margins and sold-out capacity through 2028 say otherwise.
#5165: Weighting Memory in a RAG Pipeline
Daniel's memory leg won't fire. Turns out system prompts have no inherent vector-space privilege — but retrieval does have a real knob.
#4958: Multi-Model Transcription with LLM Reconciliation
Using multiple transcription engines and an LLM judge to catch errors in voice-to-text — especially tricky technical terms.
#4809: Can a Podcast Train Itself?
Can a podcast become a self-developing project? We explore RAG vs. fine-tuning for lore memory.
#4786: Markdown vs Vectors for Agent Memory
Why coding agents need two kinds of memory — and why mixing them creates a sync nightmare.
#4704: Building a Unified AI Filesystem with Rclone and MinIO
How to build a single virtual filesystem for AI agents across multiple cloud storage providers — without the token headaches.
#4701: Why Claude Code Re-Introduces Itself to Your Repo
Claude Code remembers your preferences but not your codebase. We explore why that's by design and how to work around it.