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
ESP32 vs Raspberry Pi: The Microcontroller Mindshift 20.04.2026 23:03
Most developers think of microcontrollers like the ESP32 as just "smaller computers," but the reality is far more interesting. Unlike Raspberry Pis or other single-board computers (SBCs), microcontrollers operate on a fundamentally different tier—running real-time operating systems (RTOS) or even bare-metal firmware, with no traditional OS in sight. This episode dives into the architectural distin...
Microsoft's Phi: When Data Quality Beats Model Size 20.04.2026 21:31
Microsoft AI's Phi family of small language models challenges the assumption that bigger is better. This episode explores the data quality thesis behind Phi, its evolution from a coding model to multimodal, and the gap between Microsoft's research credibility and its product rollout turbulence. Episode #387272 — open it directly at myweirdprompts.com/387272
Why AI Coding Needs Two Brains 20.04.2026 23:04
Frontier models reason about code changes but struggle with the mechanical merge. This episode explores why the smartest AI coding tools split the job between a reasoning model and a fast, specialized 'apply' model — and why that architecture isn't going away. Episode #049716 — open it directly at myweirdprompts.com/049716
Why Open-Weight Models Are Winning 20.04.2026 20:24
Deep Cogito's Cogito v2.1 671B isn't just another LLM—it's a case study in how post-training on open-source bases is reshaping the AI landscape. This episode unpacks the strategy, architecture, and what it means for the future of model access. Episode #997420 — open it directly at myweirdprompts.com/997420
Profiling a Ghost Model 20.04.2026 18:36
When the source page returns a 404, how much can you responsibly say about an AI model? This episode wrestles with inference, uncertainty, and the limits of public knowledge in the age of opaque enterprise AI. Episode #369161 — open it directly at myweirdprompts.com/369161
Evaluating Enterprise AI: Palmyra X5 20.04.2026 20:14
What does it take to trust an AI model in healthcare or finance? This episode examines Writer's Palmyra X5, its million-token context window, and the challenges of evaluating proprietary models on behavior rather than architecture. Episode #627529 — open it directly at myweirdprompts.com/627529
The Structured Output Gap in Vision APIs 20.04.2026 23:45
Why can a model see a cereal box but not reliably return its coordinates in a consistent schema? This episode explores the hidden engineering challenges that separate general-purpose vision models from dedicated object detection tools, and what that means for your annotation pipeline. Episode #514151 — open it directly at myweirdprompts.com/514151
AI Model Spotlight: ** Aion-2.0 20.04.2026 21:03
Aion-2.0, a roleplay-focused AI model, comes from an unlikely source: AionLabs, an Israel-based venture studio specializing in AI-driven drug discovery. Built on DeepSeek V3.2, it boasts a 131k-token context window and unique reasoning token visibility—but its benchmark claims lack independent verification. Why would a pharma-adjacent lab release a storytelling model? We break down its architectur...
NVIDIA's Strategic Pivot: From Chipmaker to Model Builder 20.04.2026 21:35
NVIDIA's Nemotron 3 Super isn't just another AI model—it's a signal of the company's shift from hardware supplier to AI competitor. This episode explores what that means for the industry, the architecture behind the model, and why Jensen Huang called their earlier strategy a miss. Episode #683407 — open it directly at myweirdprompts.com/683407
The 30-Person Lab Outpacing AI Giants 20.04.2026 20:39
How Arcee AI built a frontier reasoning model with 30 employees and $20 million. This episode explores efficiency as a design principle, not just a constraint, and what it means for the future of AI development. Episode #281466 — open it directly at myweirdprompts.com/281466
Diffusion Models Take on Text Generation 20.04.2026 19:37
Inception Labs' Mercury 2 uses diffusion, not autoregression, to generate text in parallel. Does this architectural bet deliver on speed and reasoning, or is it just another model? We break down the benchmarks, the funding, and the real-world tradeoffs. Episode #864922 — open it directly at myweirdprompts.com/864922
Your Schema Is a Contract 20.04.2026 22:45
Why an hour of database planning saves weeks of remediation, how schemas function as unspoken contracts, and when denormalization actually makes sense. Episode #720219 — open it directly at myweirdprompts.com/720219
Why File Naming Conventions Are More Than Just Style 20.04.2026 24:47
In this episode, we dive into the often-overlooked world of file naming conventions and their critical role in development workflows. From snake_case to camelCase, we explore the origins, ecosystem preferences, and practical implications of each convention. Learn how case sensitivity across filesystems like ext4, APFS, and NTFS can lead to latent failures in CI/CD pipelines and why treating filena...
The Gold Standard Myth 20.04.2026 26:46
The gold standard is often romanticized as an era of monetary stability — but the historical record tells a different story. From its fragile beginnings to its catastrophic role in the Great Depression, this episode explores why the gold standard collapsed and what replaced it. We trace how money evolved from cattle to clay tablets to gold-backed notes, and why modern fiat currency isn’t as differ...
How the Dutch Invented Stock Markets 20.04.2026 25:37
How did a 17th-century Dutch trading company create the blueprint for modern stock markets? The answer lies in the VOC—the first publicly traded company—which turned risky voyages into tradable shares. From chaotic open-air trading in Amsterdam to today’s trillion-dollar exchanges, this episode traces the surprising origins of public markets, early market manipulations, and how democratization unf...
How Python Ate Wall Street 20.04.2026 28:31
Algorithmic trading isn’t just for hedge funds anymore. Python’s ecosystem—Pandas, NumPy, Backtrader, and Qlib—has collapsed the gap between institutional desks and independent quants. This episode explores how integrated tooling transformed finance, why AI is reshaping research workflows, and where reinforcement learning hits its limits. The real question: When everyone has the same tools, where...
Why Direct Currency Trades Don't Set the Price 20.04.2026 24:36
When you trade the Thai Baht against the South African Rand, why does the dollar decide the rate? This episode unpacks the strange arithmetic behind cross-pairs and why direct markets can't escape the dollar's shadow. Episode #205467 — open it directly at myweirdprompts.com/205467
How AI Models Track a Ship Seizure’s Ripple Effects 20.04.2026 21:03
Early on April 20, US CENTCOM fired on and seized the Iranian vessel *Touska* near the Strait of Hormuz—an escalation that upended prior forecasts just 72 hours before a ceasefire expiry. This episode breaks down how an AI forecasting council, composed of Claude, Kimi, and Grok, recalibrated its predictions in real time. The models agreed on some outcomes (no large-scale kinetic action within 24 h...
When OSINT Meets the Fog of War 20.04.2026 32:15
In an era where official channels contradict themselves and state blackouts leave populations in the dark, open-source intelligence (OSINT) has rushed in to fill the void. But is it clarifying the fog of war—or amplifying it? This episode examines the chaotic interplay between decentralized information ecosystems and modern conflict, from Iran’s near-total internet blackout to Israel’s flood of co...
Who Keeps Matplotlib Running? 19.04.2026 30:36
Matplotlib, the Python plotting library, powers millions of scientific charts worldwide—yet its maintenance relies on a surprisingly small team of just 15 people. In this episode, we explore how this critical open-source project is governed, funded, and sustained. From the Steering Council’s decision-making to the role of fiscal sponsors like NumFOCUS, we uncover the delicate balance of volunteeri...
When Diarization Fails Silently 19.04.2026 26:05
Speaker diarization errors are invisible but costly—mislabeling a call center agent or collapsing a courtroom transcript. This episode explores why the pipeline breaks and how to build robust systems that catch what the tools hide. Episode #742011 — open it directly at myweirdprompts.com/742011
How ADRs Solve AI's Institutional Memory Problem 19.04.2026 24:34
Software projects are full of decisions that look wrong in isolation—until you learn the hidden constraints behind them. Architectural Decision Records (ADRs) capture not just what choices were made, but why, when, and what tradeoffs were considered. Now, in the era of AI-assisted coding, ADRs have taken on new importance: they provide the institutional memory LLMs lack. This episode explores how...
Staking Out the Middle: UK's Post-Brexit AI Strategy 19.04.2026 22:18
The UK's new sovereign AI fund isn't just about capital—it's a calculated bet to carve out a competitive lane between US hyperscalers and EU regulation. This episode explores what 'sovereign AI' means for a mid-sized economy navigating post-Brexit identity, and whether targeted compute access and visa pathways can build domestic capability without getting crushed by either pole. Episode #753082 —...
How AI Flattens Your Voice in Emails 19.04.2026 21:41
AI is reshaping how we write emails, but its output often lacks the personal touch that defines authentic communication. This episode explores why AI-generated prose feels flat, how readers subconsciously notice the gap, and what strategies—from fine-tuning to prompting—can help preserve your unique voice. Dive into the technical and philosophical nuances of AI's homogenizing effect on writing and...
The Hidden Cost of Currency Volatility in Property 19.04.2026 25:40
When buying property abroad, currency moves can silently add tens of thousands to your cost. This episode unpacks how to assess liquidity, volatility, and macro drivers in thin currency markets, using an Australian expat's Jerusalem property purchase as a case study. Episode #728894 — open it directly at myweirdprompts.com/728894
Similar podcasts
Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.