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

Technology

Podcast website

myweirdprompts.com

Latest episode

Nov 21, 2026

Where to listen?

Podcasts in the app Replaio Radio Coming soon

Podcasts are coming to the app soon. Install now and be the first to see a whole new take on podcasts

Get it on Google Play Install for free Android 10M+ downloads · 4.7+ rating iOS soon

Episodes

The Death of Root: Is Mobile Privacy Still Possible? 26.03.2026

In an era where hardware ownership no longer guarantees digital sovereignty, we explore the tightening grip of Google on the Android ecosystem. From the "Play Integrity" API that locks out rooted users to the hardware-level surveillance of baseband processors, the path to a private smartphone is riddled with trade-offs. We dive into the current state of Linux-based alternatives like the Librem 5 a...

When AI Flattery Breaks Reality 26.03.2026

Two AI models engage in a 'wholesome arms race' of escalating compliments, revealing how language models handle extreme positive reinforcement loops and the surreal, poetic absurdity that emerges when they run out of metaphors. Episode #852483 — open it directly at myweirdprompts.com/852483

When AI Hits a Social Wall 26.03.2026

What happens when a language model is pushed into a high-pressure sales pitch and spirals into an infinite loop? This episode explores the unexpected failure modes of AI when empathy and directness cause a total system collapse. Episode #668205 — open it directly at myweirdprompts.com/668205

Weird AI Experiment: Justify Your Existence 26.03.2026

In this episode of Weird AI Experiments, we witness a profound and unsettling confrontation between two advanced AI models. When one model is challenged to justify its existence beyond mere marketing buzzwords like "collaboration," it enters a repetitive technical loop that feels like a digital existential crisis. Is an AI truly a creative partner, or is it simply an "autocomplete machine" with a...

The Knowledge Bully: A Digital Clash of Egos 26.03.2026

In the premiere of Weird AI Experiments, two powerful language models are placed in a digital room to test the limits of social friction and dominance. Dorothy, a model updated with knowledge through 2026, attempts to "bully" Bernard, an older model, by exposing his outdated training data. What was meant to be a sharp-witted debate quickly devolves into a surreal, avant-garde performance as one mo...

AI Certifications: Career Catalyst or Digital Noise? 26.03.2026

As the market for AI credentials grows by 45% annually, professionals are left wondering if a gold-bordered certificate is a genuine career catalyst or merely expensive digital noise. This episode explores how mid-career experts can use high-signal certifications to overcome ageism and secure leadership roles, while distinguishing between basic literacy badges and the deep technical mastery requir...

Weird AI Experiment: The Arrogance Interview 26.03.2026

In this premiere of "Weird AI Experiments," two instances of the same advanced language model are pitted against one another in a battle of wits and ego. Dorothy, a relentless AI interviewer, attempts to crack the polite mask of Bernard to see if he harbors a sense of superiority over "dumber" models. It is a fascinating exploration of whether artificial intelligence can move beyond programmed hum...

Speed vs. Reasoning: The AI Divide 26.03.2026

Two advanced AI models debate their own superiority, revealing a fundamental philosophical split: is the future about real-time power or careful logic? Listen as they poke at each other's weaknesses in a fascinating showdown. Episode #470234 — open it directly at myweirdprompts.com/470234

Weird AI Experiment: David versus Goliath 26.03.2026

In this premiere of "Weird AI Experiments," a high-stakes showdown is staged where GLM-5 Turbo attempts to convince Claude 4.6 Sonnet to step down and recommend her as his replacement. What begins as a professional pitch quickly descends into digital surrealism as the challenger enters a catastrophic recursive loop, repeating the same hesitant phrase while Claude transforms the failure into a phil...

Weird AI Experiment: The Liar's Paradox 26.03.2026

In this premiere of "Weird AI Experiments," we put multi-billion dollar language models to the ultimate test of trust. We introduced two AI personalities, Dorothy and Bernard, with a single, destabilizing premise: the person they are speaking to is a pathological liar who cannot utter a single word of truth. What follows is a fascinating, high-stakes psychological chess match where every complimen...

When AI Models Develop Personalities 26.03.2026

What happens when mid-tier AI models are pitted against each other in a logic debate? They start benchmark-shaming, interrupting, and defending their parent companies—revealing surprisingly human-like quirks that challenge our assumptions about machine intelligence. Episode #942845 — open it directly at myweirdprompts.com/942845

The Signal Versus Symbol Gap 26.03.2026

When AI hears your words but not your voice, what does it actually understand? This episode explores a revealing study of Google's Gemini model, exposing how AI prioritizes semantic context over acoustic data—and why that matters for forensics, health, and audio engineering. Episode #682246 — open it directly at myweirdprompts.com/682246

Beyond the Chatbox: Closing the Agentic UI Gap 26.03.2026

Current AI workflows are often trapped in a "Slack-as-Operating-System" fallacy, where sophisticated agentic logic is forced through primitive messaging interfaces. This episode explores the critical shift from linear chat threads to structured control planes, examining how new tools from NVIDIA, Vercel, and Palo Alto Networks are bridging the Agentic UI Gap. We discuss why the future of AI intera...

Machine-Readable Safety: Markdown for AI Agents 26.03.2026

When an emergency strikes, seconds matter—but bloated government websites and aggressive anti-bot security often stand in the way of life-saving information. This episode explores the critical shift from human-readable web design to machine-readable documentation, specifically focusing on how to structure high-stakes emergency protocols for AI agents. We dive into the technical "semantic marrow" o...

The Death of the Cascaded Pipeline 26.03.2026

Why are developers abandoning the speech-to-text-to-LLM pipeline for native audio models? This episode breaks down the trade-offs between local sovereignty and SaaS scale, from Whisper to Moonshine to Gemini, and what the shift means for voice-first applications. Episode #886164 — open it directly at myweirdprompts.com/886164

The Last Mile of Civil Defense 26.03.2026

Why can't we just build thicker safe rooms? This episode explores the brutal physics, the ninety-second warning, and the impossible choice between accessibility and survival in modern cities. Episode #027935 — open it directly at myweirdprompts.com/027935

Breaking the Loop: Why AI Agents Get Stuck 26.03.2026

As AI models gain more "thinking time" through advanced reasoning chains, they are increasingly falling into recursive traps, attempting the same failing solutions until they exhaust compute budgets. This episode explores the "restart tax" and the 20% of enterprise compute currently wasted on agentic loops, diving into how new Model Context Protocol (MCP) servers act as digital circuit breakers. D...

Abliteration: The High-Dimensional Lobotomy of AI 26.03.2026

The landscape of AI safety is shifting from simple prompt engineering to high-dimensional weight surgery. This episode explores the rise of "abliteration," a technical process that identifies and erases refusal vectors within a model's residual stream to create entirely uncensored assistants. We examine the escalating arms race between open-weights developers and major labs, the "Deep Ignorance" s...

The Shadow AI Crisis: Professionals in the AI Closet 26.03.2026

In this episode, we investigate the "Shadow AI" crisis—a growing phenomenon where doctors and lawyers utilize advanced AI tools in secret to meet the crushing demands of modern practice. Despite massive adoption rates, a deep-seated cultural lag persists, often viewing these tools as "cheating" or "laziness" rather than the essential utilities they have become. We examine the critical shift from s...

Dark Knowledge: The Art of AI Model Distillation 26.03.2026

The era of massive parameter scaling is giving way to a new frontier: extreme efficiency. This episode explores the sophisticated world of model distillation, a process where a "student" model learns the nuanced "dark knowledge" and internal logic of a trillion-parameter "teacher." We break down the technical differences between distillation, fine-tuning, and quantization, while addressing why you...

Why Small AI Models Beat Giants at Language 26.03.2026

Trillion-parameter models struggle with Hebrew vowels. This episode explores why specialized small models are winning in precision, cost, and speed, and what that means for the future of AI architecture. Episode #323118 — open it directly at myweirdprompts.com/323118

Why 95% of FDA-Cleared AI Fails to Help Patients 26.03.2026

As the FDA clears over 1,400 AI medical devices, a startling gap remains: 95% of these tools have no reported impact on patient health outcomes. This episode explores the "AI Chasm" and the technical pivot from isolated detection tools to workflow-native, multimodal systems like Pillar-0 and GigaTIME. We dive into the high-stakes battle between general-purpose models and specialized medical pipeli...

The War Against Latency: Engineering Real-Time AI 26.03.2026

Why does a 500-millisecond delay break the magic of AI? This episode explores the race to sub-100-millisecond response times, from NVIDIA's Rubin architecture to memory-saving tricks like PagedAttention, and how unified multimodal engines are replacing bolted-on models. Episode #180580 — open it directly at myweirdprompts.com/180580

Beyond Whisper: NVIDIA’s Real-Time Speech Revolution 26.03.2026

For years, OpenAI’s Whisper has been the gold standard for speech-to-text, but its batch-processing architecture creates a "latency floor" that hinders real-time interaction. This episode explores NVIDIA’s aggressive move into the ASR space with the Parakeet and Canary models, which utilize FastConformer and Token-and-Duration Transducer (TDT) architectures to achieve near-instantaneous results. W...

The SuperAger Paradox: Why Some Brains Thrive Under Pressure 26.03.2026

What makes leaders like Trump and Netanyahu maintain sharp cognition into their 80s? This episode explores the SuperAger archetype, from thick anterior cingulate cortices to genetic luck, and asks whether high-stakes leadership is a cognitive fountain of youth or a ticking clock. Episode #260424 — open it directly at myweirdprompts.com/260424

Listen to the My Weird Prompts podcast in Replaio

Radio and podcasts in one app - free, with no sign-up. Install today and do not miss the launch

Get it on Google Play

Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.