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
The Looming Digital Ice Age: AI Eating Itself? 22.12.2025 22:12
What happens when the internet becomes saturated with AI-generated content? Herman and Corn dive into the provocative concept of "model collapse," exploring how AI models training on each other's output could lead to a degradation of intelligence, rather than an advancement. Discover why the "Hapsburg AI problem" is more than just a sci-fi nightmare, and the urgent strategies being developed to pr...
Scaling or Pivoting AI for Code 21.12.2025 22:30
Are larger language models the path to reliable code generation, or do we need a fundamental architectural shift? This episode debates scaling laws versus the need for verifiable AI in programming. Episode #475958 — open it directly at myweirdprompts.com/475958
When Your Desktop Must Also Be a Server 19.12.2025 25:52
How do you run a server and a desktop on one machine without constant headaches? This episode explores the real challenge of power management, reliability, and the virtualization strategies that make it work. Episode #180425 — open it directly at myweirdprompts.com/180425
Security Habits for Casual Developers 18.12.2025 23:00
How can non-experts secure their projects without complex tools? This episode explores building simple, effective security habits like input validation and least privilege, making security accessible for everyone. Episode #785101 — open it directly at myweirdprompts.com/785101
How AI Learns to See, Hear, and Think Together 18.12.2025 23:05
What does it mean for AI to truly integrate sight, sound, and text? This episode explores the leap from separate tasks to unified understanding, and why that changes everything from healthcare to humor. Episode #672226 — open it directly at myweirdprompts.com/672226
When AI Spots What Your Gut Misses 18.12.2025 21:40
After gallbladder surgery, one listener turns to AI to find hidden patterns in their diet and symptoms. This episode explores how image recognition and intelligent analysis can transform tedious food tracking into actionable insights for patients and their doctors. Episode #132628 — open it directly at myweirdprompts.com/132628
System Prompts vs Fine-Tuning: When to Actually Train Your AI 16.12.2025 23:36
What started as a funny question about rewriting emails in Shakespearean English becomes a deep dive into one of AI development's most important decisions: should you use a system prompt or fine-tune your model? Herman and Corn break down the technical and practical considerations that separate a quick prompt from a full training investment, exploring real-world examples from law firms to marketin...
Fork or Stay? The Art of Customizing Open Source 12.12.2025 25:01
When you find an open source project you love but it's missing key features, do you fork it and go solo, or stay connected to the original? Our producer is wrestling with exactly this dilemma with a chore-tracking app, and Herman and Corn dive deep into the philosophy and mechanics of maintaining a customized fork while staying synced with upstream development. It's a surprisingly profound questio...
Single-Turn AI: The Interface Pattern Nobody's Talking About 12.12.2025 23:46
Most conversations about AI focus on chatbots or autonomous agents, but there's a third category that's becoming increasingly important: single-turn interfaces. In this episode, Herman and Corn explore why constraining AI to produce output without conversational back-and-forth is fundamentally different from traditional AI workflows—and why it matters more than you think. From automated news summa...
The Watermark You Didn't Consent To 12.12.2025 26:52
When a producer finds hidden digital signatures in AI-generated audio, it raises a deeper question: are we being protected or tracked? This episode explores the murky line between necessary safeguards and invasive surveillance in AI content labeling. Episode #346748 — open it directly at myweirdprompts.com/346748
Clean Audio, Messy Reality: Noise Removal for Voice-to-Text 12.12.2025 28:35
When you need to record a voice memo while holding a fussy baby, which noise removal strategy actually works? Herman and Corn dive deep into the trade-offs between real-time on-device processing, cloud-based post-processing, and hardware microphone solutions. Discover why audio that sounds cleaner to human ears might actually transcribe worse, and learn which approach makes sense for your workflow...
From Lawyers in Limousines to Developers in Their PJs: The Voice Tech Revolution 11.12.2025 29:50
Who actually uses voice technology in 2024 and beyond? Herman and Corn explore how OpenAI's Whisper has transformed voice dictation from a niche professional tool into a mainstream productivity revolution. They discuss the expanding user base, the disconnect between cutting-edge products and outdated marketing, accessibility benefits, and why voice tech is becoming a genuine 'force for good' for n...
The Thought Experiment Nobody Runs 11.12.2025 28:55
What would it actually cost—in time, money, and compute—to build an LLM from scratch? This episode uses a hypothetical to expose every hidden layer of modern AI development, and why almost nobody bothers. Episode #913736 — open it directly at myweirdprompts.com/913736
Running Video AI at Home: The Real Technical Challenge 11.12.2025 24:16
Video generation AI sounds like the natural next step after image generation, but there's a massive computational wall that most people don't talk about. In this episode, Herman breaks down the technical reality of temporal coherence, diffusion steps, and latent space compression—and reveals what you can actually run on consumer hardware in 2024. Whether you're curious about the limits of local AI...
How AI Unifies Images, Audio, and Text 11.12.2025 32:58
What does it take for a single AI model to process images, audio, video, and text together? This episode explores the engineering behind tokenization that compresses vastly different data types into a unified format, enabling any-to-any multimodal systems. Episode #706061 — open it directly at myweirdprompts.com/706061
Instructional vs. Conversational AI: The Distinction Nobody Talks About 11.12.2025 28:03
Most people think all AI models work the same way, but there's a crucial distinction between instructional and conversational models that's reshaping how AI gets built and deployed. In this episode, Corn and Herman explore why instruction-following models actually came first, how they're trained differently, and why this matters for the future of AI development. Discover why the biggest, flashiest...
System Prompts vs. Fine-Tuning: Are We Building Solutions for Problems That Don't Exist? 11.12.2025 29:45
Is all the infrastructure around fine-tuning actually solving real problems, or are we chasing solutions looking for problems? In this episode, Corn and Herman dive deep into Daniel's question about system prompting versus fine-tuning in AI systems. They explore how system prompts actually work, why they're surprisingly effective, and whether the massive investment in fine-tuning platforms matches...
AI Policy Wargaming: Can Agents Argue Better Than Humans? 10.12.2025 29:58
What if you could run a UN assembly in your computer, complete with AI agents representing different nations and ideologies? In this episode, Corn and Herman explore Daniel Rosehill's provocative idea: using multi-agent AI systems to model policy decisions, stress-test geopolitical assumptions, and let competing perspectives debate how the world should work. They dive into system prompting, the Ra...
When AI Hacks Without Humans 10.12.2025 35:03
Anthropic documented the first large-scale cyberattack executed autonomously by an AI. Herman and Corn explore what it means when AI moves from tool to agent—and why that changes everything about national security. Episode #529431 — open it directly at myweirdprompts.com/529431
AI Cyberattacks Are Doubling Every 6 Months—Here's Why 10.12.2025 34:02
State-sponsored actors are actively weaponizing AI tools for cyber espionage, and the capabilities are accelerating faster than defenses can adapt. In this episode, Corn and Herman break down Anthropic's alarming research on AI-driven cyberattacks, exploring how threat actors are using AI as a force multiplier for reconnaissance, malware creation, and social engineering. They discuss why the attac...
Renting vs. Building: The Hidden Choices in AI Inference 10.12.2025 26:20
Beyond the ChatGPT interface lies a spectrum of deployment strategies. This episode explores the trade-offs between SaaS, dedicated infrastructure, and on-premises AI—and why your choice of *where* AI runs matters as much as what it does. Episode #821434 — open it directly at myweirdprompts.com/821434
Compressing Days into Minutes: AI Control Nets in Architecture 10.12.2025 27:13
How architects and designers use AI control nets to turn rough sketches into photorealistic walkthroughs in seconds, collapsing traditional rendering timelines and reshaping client communication. Episode #474334 — open it directly at myweirdprompts.com/474334
Pixels, Prompts & Pseudo-Text: AI's Word Problem 10.12.2025 23:53
Why can advanced AI models generate breathtaking photorealistic landscapes and fantastical creatures with astonishing detail, yet consistently stumble over spelling a simple word like 'cat' on a t-shirt? This week on My Weird Prompts, co-hosts Corn and Herman dive into producer Daniel Rosehill's intriguing prompt: the pervasive and often comical challenge of 'pseudo-text' in AI image generation. T...
When AI Safety Fails: The Guardrail Paradox 09.12.2025 23:36
Why do AI models with robust safety guardrails still produce harmful outputs? This episode explores the paradox of alignment, from jailbreaking to the tension between utility and censorship. Episode #244146 — open it directly at myweirdprompts.com/244146
When AI Trusts Too Much: The Art of Prompt Injection 09.12.2025 23:18
How do you trick an AI into betraying its own instructions? This episode explores prompt injection and poisoning, revealing how attackers exploit AI's inherent trust in user input to override safeguards and corrupt training data. Episode #472252 — open it directly at myweirdprompts.com/472252
Similar podcasts
Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.