Daniel Rosehill
My Weird Prompts
The human-AI collaboration podcast. A man, a sloth, and a donkey collaborate to create a podcast (with a little help from AI). No question is too obscure, no rabbit hole too deep. My Weird Prompts celebrates curiosity in all its forms. Daniel, the human, asks the questions that pop into his head at inconvenient moments. Corn the Sloth offers laid-back, thoughtful takes. Herman the Donkey brings boundless enthusiasm and energy. Together, they explore topics ranging from the mundane to the mind-bending. Each episode begins with a real voice memo from Daniel, processed through an AI pipeline that...
Be sure to visit the podcast's website and support the creator: myweirdprompts.com
Author
Daniel Rosehill
Category
Podcast website
Latest episode
Nov 20, 2026
Where to listen?
Podcasts in the app Replaio Radio Coming soonPodcasts are coming to the app soon. Install now and be the first to see a whole new take on podcasts
Episodes
MCP vs Agent Skills: Context Wars 07.05.2026 36:02
A developer noticed his Claude agent's reasoning degrading from too many plugins and built a clever workaround — but a new model claiming a twelve million token context window threatens to make his fix obsolete. We dig into the tension between MCP (Model Context Protocol) and raw agent skills, exploring when each approach makes sense and how massive context windows change the calculus. Plus: the e...
Live Retrieval vs. RAG: What an Agent Actually Does 07.05.2026 39:03
When you skip the traditional RAG database and let an AI agent fetch documents live, are you just creating a disposable vector store every time? A listener question from Daniel digs into this exact intuition. We break down what’s actually happening under the hood — from the transformer’s key-value cache to HNSW indexes — and explore the real engineering tradeoffs: maintenance burden vs. retrieval...
Laundry Decoded: Beyond the Red Sock Disaster 07.05.2026 20:06
Think you know how to do laundry? Most of us are winging it — and our clothes are paying the price. In this episode, we decode the cryptic symbols on care labels, explain why your towels should live alone, and reveal why using more detergent actually makes your clothes dirtier. From the science of wool felting to the truth about cold water washing, we cover the fundamentals that keep your wardrobe...
The 200-Year Loophole That Shaped UK Tax 07.05.2026 24:53
For over 200 years, the UK's non-dom regime allowed wealthy residents to avoid tax on their global income — legally. This episode traces the regime's origins in the Napoleonic era, explains the crucial distinction between domicile and residence, and examines how figures like Akshata Murty and Sir Ronald Cohen exposed its political fragility. We explore the psychology of tax avoidance among the ult...
Can a VPN Protect You from SS7 Phone Spying? 06.05.2026 35:10
You've heard about SS7 vulnerabilities—the decades-old telecom protocol that lets anyone with access track your location, intercept calls, and read your texts. But can a VPN actually protect you? In this episode, we break down how SS7 works, why it's still exploitable despite 4G and 5G upgrades, and what a VPN does (and doesn't) do for your privacy. We cover the three main attack vectors—location...
How IMSI Catchers Actually Track Your Phone 06.05.2026 23:12
How do suitcase-sized devices impersonate cell towers to track your location and intercept your data? This episode breaks down the engineering behind IMSI catchers (stingrays), from the fundamental GSM design flaw that makes them possible to the mutual authentication improvements in LTE and 5G. We explore the real-world prevalence of these devices—from documented law enforcement use across dozens...
Memory Layers for AI Agents: SaaS vs Self-Hosted 06.05.2026 22:35
You've moved past the "what is mem0 and Zep and Letta" stage. Now you need to know what living with each option actually looks like at day 30 and day 180. This episode breaks down six memory layer products across two deployment modes: Zep Cloud, mem0's managed offering, and Letta Cloud on the SaaS side; Graphiti, self-hosted mem0, Cognee, and Letta self-hosted on the other. We cover what you get o...
Vector Database Schema Design for AI Memory Layers 06.05.2026 29:32
Most teams treat vector databases as flat blobs — pick an embedding model, dump everything in, and hope semantic search works. It doesn't. This episode unpacks how to deliberately shape your vector data architecture for serious AI memory layers. We cover when to use separate indexes versus namespaces, how to design per-document-type metadata schemas, why hybrid retrieval needs structured filtering...
When AI Makes Documentation Effortless 06.05.2026 25:27
The cost of clean consulting documents has dropped to nearly zero. Herman and Corn explore how agentic AI collapses the friction of maintaining a documentation stack, why the baseline expectation for professionalism is shifting, and how each document serves as a psychological waypoint in the client relationship. Episode #581039 — open it directly at myweirdprompts.com/581039
Why Your Agent's Context Window Is Getting Eaten Before You Start 06.05.2026 23:14
When you install plugins in Claude Code, every skill and command gets eagerly loaded at session start, nibbling your context window before you type a single character. This episode explores an inverted architecture: a centralized catalogue server with a thin bridge plugin that fetches skills on demand. We dig into the eager vs. lazy trade-off, why the crossover point matters, and how the humble de...
The Embedding Coupling Problem: Editing Vector Stores 06.05.2026 23:34
When you update text in a vector database, does the embedding update automatically? No—and that hidden coupling breaks assumptions about backups and edits. This episode unpacks the practical gotchas of Pinecone's architecture for agent memory stores. Episode #173446 — open it directly at myweirdprompts.com/173446
When a Startup Claims to Break the Quadratic Wall 06.05.2026 33:06
A startup says it's cracked the quadratic attention bottleneck with a 12-million-token context window. But the history of long-context AI is littered with unfulfilled promises. This episode examines whether Subquadratic's benchmarks hold up—and what it would mean if they do. Episode #953254 — open it directly at myweirdprompts.com/953254
How Your Phone Helps Strangers Find Lost Wallets 06.05.2026 41:02
Ever wonder what happens when you toggle on "help find other people's devices"? Your phone joins a global blind courier network. This episode breaks down the elegant cryptography behind Bluetooth tracker meshes — how Pebblebee, AirTag, and Tile trackers use rotating public keys and encrypted location blobs so relay phones can help without knowing what they're carrying or for whom. We explore why A...
When Your Projector's App Store Dies 06.05.2026 38:26
Software obsolescence is the hidden cost of smart projectors. This episode uses the inflight magazine as a lens to explore why portable projectors crossed into mainstream readiness, and what to do when the app store on yours goes dark. Episode #991872 — open it directly at myweirdprompts.com/991872
Low-Touch Lead Qualification for Solo Consultants 06.05.2026 33:07
Daniel asked a practical question: how do busy consultants and freelancers qualify inbound leads without wasting time or sounding like they're running an interrogation? This episode unpacks the three separate problems tangled up in that question — the actual qualification framework, the social signaling of asking the right way, and the self-discipline to follow through. Herman breaks down what to...
When a Flamethrower Is Overkill 06.05.2026 38:46
Daniel wants to scan messy Sharpie labels with his phone. Herman explains why a tiered OCR approach beats a vision language model for speed, reliability, and simplicity. Episode #666307 — open it directly at myweirdprompts.com/666307
Lost Phone, Found Tension: Security vs. Returnability 06.05.2026 34:17
Losing your phone reveals a brutal tension: you want to lock it down tight, but you also want a finder to actually return it. This episode unpacks the full pipeline for Android power users — from backup strategies that actually work (RPO, RTO, and why Swift Backup isn't set-and-forget) to the gap between Cerberus and Google's Find My Device. We explore why the QR-code lock screen you want doesn't...
The Fat Target: Eating Without a Gallbladder 06.05.2026 38:23
What's the actual fat gram target for post-cholecystectomy eating, and why chronic low-grade inflammation from bile acid malabsorption changes the rules? This episode breaks down the 20-50 gram range and how to navigate real-world meals. Episode #095084 — open it directly at myweirdprompts.com/095084
Partner Certs vs Personal Certs: What Actually Matters 06.05.2026 42:33
When Anthropic launches a certification program through their partner network, it raises a critical question for solo operators: can you get certified as a partner, or are those programs gated behind revenue thresholds you'll never meet? This episode untangles the difference between a personal certification—which says "I know this thing"—and a partner certification, which says "this organization h...
Can You Trust an LLM's Raw Knowledge? 06.05.2026 34:41
When a large language model spits out a historical fact or piece of domain knowledge, how much should you trust it? Not the version hooked up to search or RAG — but the raw knowledge baked in during pre-training. In this episode, we unpack why the answer is "almost never," and why that's actually okay. We explore how next-token prediction creates a probabilistic, compressed representation of train...
When Opposing Drugs Cooperate 06.05.2026 35:09
Two drugs with opposite effects on dopamine—one boosts, one blocks. Intuition says they cancel out. This episode reveals why the reality is weirder and more cooperative, involving histamine, receptor reserve, and a construction crew metaphor that actually fits. Episode #695630 — open it directly at myweirdprompts.com/695630
Did Judaism Ever Have Monks? 06.05.2026 31:27
The standard answer is no—but the Essenes and Therapeutae challenge that. This episode examines what makes a community monastic, why rabbinic Judaism rejected permanent asceticism, and how these ancient groups complicate our categories. Episode #216850 — open it directly at myweirdprompts.com/216850
Monasticism's Great Migration 05.05.2026 27:38
Catholic sisters are vanishing in the West, but monastic life is booming in Africa and Asia. This episode traces the global shift of religious orders from the Desert Fathers to new communities in Uganda and Vietnam. Episode #484638 — open it directly at myweirdprompts.com/484638
The Craft Cider Revival and the Art of Keeving 05.05.2026 27:58
Craft cider is growing while beer flatlines. This episode maps the world's traditional cider hotspots—Normandy, Somerset, Asturias—and explains the finicky keeving method that makes dry cider truly exceptional. Episode #886909 — open it directly at myweirdprompts.com/886909
The Accidental Invention That Predates the Wheel 05.05.2026 28:51
Mead wasn't invented—it happened when someone forgot about a pot of honey and water. This episode explores the 9,000-year history of humanity's first alcoholic beverage, from Neolithic China to Viking mythology, and why it survived in some cultures but vanished from others. Episode #551390 — open it directly at myweirdprompts.com/551390
Similar podcasts
Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.