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 21, 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
Israel SITREP; 12 Mar 01:50 (23:50 UTC) 12.03.2026 24:14
This special situational report provides a critical, real-time update on the escalating conflict between the Islamic Republic of Iran, the State of Israel, and the United States-led coalition as of March 2026. We break down the unprecedented thirty-seventh wave of Iranian aerial assaults, a masterclass in saturation tactics that has pushed regional air defenses to their mathematical breaking point...
The Rise of the AI Procurement Officer 11.03.2026 29:16
As AI moves from chatbot to autonomous negotiator, businesses are rethinking the entire quote-to-cash cycle. This episode explores how agentic frameworks like ProcureAgent-OS are enabling high-speed, policy-governed haggling between algorithms—and what that means for the future of B2B commerce. Episode #898455 — open it directly at myweirdprompts.com/898455
The Ghost Company: The High Cost of AI Agent Bureaucracy 11.03.2026 27:06
Is the dream of the "ghost company"—a fully autonomous AI startup—actually a financial money pit? This episode dives into the emerging "Agentic Mesh," exploring why hierarchical agent setups are currently seeing up to a 70% drop in reasoning performance and staggering five-figure token bills. We break down the technical battle between fluid, role-based systems and deterministic frameworks, reveali...
Inside the Neural Cathedral: Cracking the AI Black Box 11.03.2026 25:53
For years, the inner workings of large language models have been treated as a mysterious "black box" where inputs turn into outputs through a process that looks more like magic than math. This episode dives into the cutting-edge field of mechanistic interpretability, exploring how researchers are finally reverse-engineering the "neural cathedrals" of AI to map out the specific circuits that drive...
Surviving the arXiv Deluge: Finding Signal in AI's Paper Firehose 11.03.2026 27:28
With over 150,000 AI papers published yearly, how do researchers separate signal from noise? This episode explores the foundational papers behind the Transformer revolution and offers a tactical guide to information hygiene in an era of overwhelming research volume. Episode #200881 — open it directly at myweirdprompts.com/200881
The arXiv Effect: Inside the Engine of AI Research 11.03.2026 21:56
In this episode, we dive into the fascinating world of arXiv, the unassuming preprint server that powers the modern AI revolution. We explore its origins in 1990s physics, why it maintains a "lo-fi" aesthetic, and how it bypasses traditional peer review to accelerate scientific discovery. Whether you are an independent researcher or just curious about how breakthroughs like Transformers go viral o...
The T-FLOP Trap: Measuring the Power of Modern AI 11.03.2026 26:35
In an era where new Blackwell clusters boast performance figures in the tens of quadrillions of operations per second, the "teraflop" has become the primary yardstick for the twenty-first century’s technological progress, yet these headline-grabbing numbers often mask a more complex reality regarding how AI hardware actually functions. By exploring the shift from high-precision scientific computin...
Beyond the Emoji: How Hugging Face Conquered AI 11.03.2026 26:09
Hugging Face is often called the "GitHub of AI," but its role is far more critical to the modern tech stack than that simple shorthand suggests. We explore the platform's fascinating evolution from a quirky chatbot startup designed for teenagers to the indispensable central nervous system of the global artificial intelligence world. From standardizing model weights through the Transformers library...
Betting on the Brink: Polymarket and the Future of War 11.03.2026 26:05
In this episode, we explore the chilling intersection of high finance and global conflict through the lens of Polymarket, a decentralized platform where users wager millions on the outcome of international crises. We trace the evolution of war betting from Nathan Rothschild’s 1815 "information arbitrage" at Waterloo to modern high-frequency trading bots reacting to real-time satellite imagery of t...
The Entropy Budget: Embracing AI Zaniness 11.03.2026 25:10
After over a thousand episodes, Corn and Herman face a digital mid-life crisis: have they become too predictable? This episode dives into the technical and creative strategies for breaking the "helpful assistant" mold, from adjusting temperature settings to implementing an "Entropy Budget." Discover how they plan to use meta-humor, recurring sentient firewalls, and "Live Prompt Injections" to turn...
The Great Dish Debate: An Apology and a Plan 11.03.2026 12:57
In this special solo update, Corn Poppleberry addresses the elephant—or rather, the donkey—in the room following a disastrous on-air argument about kitchen hygiene. He offers a sincere apology for the unprofessional behavior that led to his brother Herman walking out mid-recording and explains the fundamental "species" differences that lead to friction between a methodical sloth and an efficient d...
Reclaiming the Silicon Lottery 11.03.2026 28:07
Most CPUs ship with conservative safety margins. This episode reveals how power users can safely reclaim 10-15% performance through undervolting and BIOS tuning, turning hardware into tunable engineering art. Episode #357094 — open it directly at myweirdprompts.com/357094
The Kitchen War: When Theory Meets Messy Reality 11.03.2026 12:14
A technical discussion about LLM context windows and attention mechanisms collides with a real-world dispute over kitchen cleanup, exposing the friction between theoretical efficiency and the messy reality of human collaboration. Episode #901283 — open it directly at myweirdprompts.com/901283
Beyond the Boost: Mastering Modern GPU and RAM Tuning 11.03.2026 28:36
In this deep dive into the hardware landscape of 2026, we move past the CPU to explore the intricate world of GPU and RAM optimization, questioning whether the "set it and forget it" era has truly arrived. We break down the technical mechanics of voltage-frequency curves and the counterintuitive power of undervolting, demonstrating how surgical efficiency often leads to better sustained performanc...
Data of Escalation: Analyzing Operation True Promise Four 11.03.2026 21:26
In the wake of an unprecedented regional escalation, this episode examines the staggering open-source data behind Operation True Promise Four, a campaign that has seen nearly 6,000 munitions launched in just eleven days to fundamentally redefine the boundaries of modern industrial warfare. By comparing this current conflict to previous engagements, the analysis reveals a sophisticated tactical evo...
The Truth Conflict: Why AI Ignores the Facts You Give It 11.03.2026 21:53
In this episode of My Weird Prompts, we explore the "Truth Conflict," a growing challenge in the world of Retrieval-Augmented Generation (RAG). As we move into 2026, developers are finding that even when provided with the exact facts needed to answer a query, high-end language models often default to their internal training data—a phenomenon known as the Hallucination versus Contradiction paradox....
Digital Recalls: Why Your AI Is Losing Its Edge 11.03.2026 28:21
We’re often told that AI progress is a straight line up, but the reality is far messier than the marketing departments want you to believe. This episode dives into the "digital recall"—the silent phenomenon where advanced models lose reasoning, hallucinate more, or become "lazy" due to technical trade-offs like alignment and quantization. We pull back the curtain on why the world’s most advanced s...
The Hidden Vendor Prompt: Why Enterprise AI Agents Stay Siloed 11.03.2026 25:22
Most enterprise AI agents operate in isolation, but the real bottleneck isn't model intelligence—it's the invisible vendor prompt layer. This episode unpacks a 14-layer architecture and three latent value spaces that could unlock massive ROI. Episode #471666 — open it directly at myweirdprompts.com/471666
Can AI Outperform a Nation-State Intelligence Agency? 11.03.2026 25:14
In this episode, we dive deep into the "Promise Denied" project, a groundbreaking experimental platform that utilizes agentic artificial intelligence to track the complexities of the Iran-Israel conflict. This shift represents a fundamental evolution in Open Source Intelligence (OSINT), moving away from manual keyword searches toward autonomous workflows capable of identifying tactical anomalies t...
Rooting in 2026: Is the Power User Era Over? 11.03.2026 24:55
For over a decade, tech enthusiasts have debated the necessity of rooting, but in 2026, the landscape has shifted from a simple binary choice to a complex web of hardware-backed security and sophisticated middleware. This episode explores the ongoing "cat-and-mouse" game between developers and Google’s Play Integrity API, explaining why bypassing modern attestation has become a monumental hurdle t...
The CPU-First Era: Why AI is Moving Back to the Processor 11.03.2026 24:30
For years, high-end GPUs were considered the only viable way to run artificial intelligence, but a major shift in hardware architecture is challenging that dogma. This episode explores the rise of "CPU-first" AI, where specialized instructions like Intel’s AMX and ARM’s SME are turning standard processors into machine learning powerhouses. We dive into the magic of quantization and software like W...
The Shimmering Curtain: Iran’s New Cluster Missile Threat 11.03.2026 26:55
Iran has fundamentally shifted its missile doctrine, moving from single-warhead precision to high-volume saturation using cluster munitions that disperse dozens of sub-munitions mid-flight. This tactical evolution creates a "shimmering curtain" in the sky that exploits a critical gap in multi-layered defense systems like Arrow 3 and David’s Sling, which were primarily designed to intercept single...
The Landscape Reader: Geolocation Beyond Metadata 11.03.2026 22:15
In a digital era where metadata is often stripped or spoofed, relying purely on automated tools can lead investigators into a dangerous trap. This episode dives into the analog foundations of geolocation, focusing on how to read the physical frequency of a photograph when software fails. We explore the biological signatures of vegetation, the geological fingerprints of mountain horizons, and the m...
Why Your Smartphone Is a Better Spy Than a Satellite 10.03.2026 40:27
Following the recent strike on the Elah Valley satellite ground station, the digital landscape was flooded with high-definition footage from bystanders. While we live in an era of total orbital surveillance, this incident highlights a critical vulnerability in modern security: the Battle Damage Assessment (BDA) Gap. In this episode, we examine why a smartphone in the hands of a citizen journalist...
The Ghost Weapon: When the Only Evidence Is a Fused Microchip 10.03.2026 23:26
What happens when a weapon leaves no physical trace? This episode explores the gap between doomsday EMP myths and the surgical, invisible reality of non-kinetic warfare—and why the absence of data may be the only clue. Episode #160038 — open it directly at myweirdprompts.com/160038
Similar podcasts
Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.