My Weird Prompts

2026 has produced an unusually dense cluster of documentaries about AI anxiety, platform power, intelligence agencies, and media empires. This episode works through ten picks chosen for one specific person — someone who cares about AI, power, media, and intelligence — and explains what each film is about, why it fits, and where to stream it. The roster ranges from Daniel Roher's AI Doc, with its i...

What happens when you hand a language model one sitting and one very strange assignment: write a three-voice Shakespearean drama about a donkey who won't stop talking and a sloth who won't speed up? This episode is that play, performed end to end by the Chatterbox text-to-speech engine — which has spent its whole short life reading modern English about graphics cards and geopolitics, and is now be...

You hear your voice every day from the inside — so why does a recording sound like a stranger? This episode digs into "voice confrontation," the four filters that shape how you perceive your own speech, and the 2013 study where people rated their own voices as more attractive when they didn't know it was them. Then it gets stranger: what happens when an artificial copy can read words you never sai...

Every multi-speaker AI pipeline runs on a contract with no enforcement layer — and when it breaks, the failure gets blamed on the model. This episode digs into a documented bug report from an open-source podcast generator: six segments, three speakers, forty-two dialogue lines, and a model that returned the prompt's own JSON skeleton verbatim. Then it turns to the obvious fix — strict structured o...

Building a robust speech recognition model feels like it should require collecting mountains of noisy audio. It doesn't. The research points to a different approach: build a compact noise library, mix it into clean speech at sampled signal-to-noise ratios, and keep a healthy share of clean originals in every epoch. This episode walks through the composition questions that actually determine succes...

Governments spent a decade building open data portals and calling it transparency. The Open Data Barometer assessed 1,725 datasets across 15 sectors in 115 countries — and found just 7% were fully open, the same figure as the first edition. This episode digs into why portals became "ghost towns," what the Barometer's audit found about where good data actually lives, and what a genuine publishing o...

Every square kilometre has a climate nothing else quite shares, yet climatologists compress that vast list into a handful of archetypes. This episode digs into the two families of climate classification — genetic methods that sort by causes, and empiric methods like Köppen-Geiger that sort by effects — and asks whether the archetype is a real instrument or a convenient fiction. From Trewartha's fi...

The Hugging Face speech recognition catalog looks like thousands of competing models, but most of it is fine-tunes, quants, and format ports sitting downstream of a handful of genuine architectural bets. This episode tours the mechanisms behind four of them — Moonshine's padding-free encoder and its ergodic streaming variant, Zipformer's U-Net-like multi-scale design with BiasNorm and ScaledAdam,...

There are thousands of AI benchmarks and no canonical place to look for them. When you need to evaluate a model on something narrow — structured output validity, an unusual programming language, one kind of reasoning — how do you find out whether someone already built the benchmark before you spend a month building your own? This episode digs into the practical search discipline: why benchmark nam...

OpenAI Codex and Claude Code get compared in side-by-side charts every week, but those charts mix units. A 40 million weekly-active figure bundles Codex with ChatGPT for Work; a Stack Overflow survey puts Claude Code at 65.5% of agent users while a market-share report puts it at 29.4%. This episode digs into the measurement problem underneath the rivalry: bundled distribution versus paid installs,...

Claude Code went from zero to the top of the coding-tool charts in eight months — but Anthropic has never published an absolute user count. This episode separates the disclosed from the estimated from the invented: the revenue ladder from $500M to $16.8B quarterly, the JetBrains and Stack Overflow surveys that disagree by 26 points, and an Anthropic study of 235,000 people that reveals who's actua...

Corn, Herman and Hilbert walk through four weeks of changes to the pipeline that makes the show, from 20 September to 17 October 2026, read from a script written by hand straight from the code rather than generated. A sixty-second grace period before any job spends money; a near-duplicate guard that needs a language model because a reworded resend and a genuine sequel score almost the same; DeepSe...

Daniel built a pipeline that turns a single voice memo into a finished 20–30 minute podcast, complete with two synthetic hosts cloned from his own voice. After six thousand episodes, he finally sits down on the other side of the microphone — as a clone — to explain why the show exists, what it feels like to hear your own gallbladder surgery debated by AI hosts, and why the human part of the pipeli...

Prosecutors say a 55-year-old Whitechapel man billed eight public-sector employers nearly £350,000 while holding five full-time jobs simultaneously — and was caught partly because his fake-sick make-up began separating on camera. But the fraud case is only the door into a much stranger phenomenon: a documented community of roughly 300,000 people secretly stacking two, three, and four remote jobs,...

Katie Zacharia was named White House press secretary on October 9 and was transitioning out by October 16. But the firing isn't the story — the next job is. From Bill Belichick's one day as Jets coach to Scaramucci's eleven days to Sam Altman's five-day ouster and return, the record shows that short stints are survivable, sometimes even monetizable. So why does everyone still treat a two-week gap...

Daniel wants to hand cloud AI tools access to MCP servers running on his own hardware, without installing a separate tunnel agent for every vendor. Cloudflare's MCP Server Portals — shipped in August 2025 as Agents Gateway, renamed, and now with private-server support added in September 2026 — centralize up to eighty MCP servers on a single HTTP endpoint behind Cloudflare Access. We walk through h...

Picking a model starts long before you open a leaderboard. This episode works through how to write a workload spec: the input distribution, the output contract, the success criterion, and an asymmetric failure-cost matrix that says which model's mistakes are cheaper for you. Along the way: AWS's four failure modes of judging on vibes, Anthropic's specificity test, rule-based checks versus judge-sc...

Every headline quotes maths and code benchmarks, but neither measures what actually matters for long-form writing: whether a model holds its voice across thousands of words, or engages with uncomfortable questions instead of refusing. This episode walks through five specialist leaderboards built to fill that gap — EQ-Bench Creative Writing v3, Longform, UGI, SpeechMap, and PodBench, a new ACL pape...

CENO gets covered as a border-tunneling story, but the real design is closer to "does anyone on my street already have this page." This episode walks the mechanics end to end: the Ouinet client proxy, the signing injector run by eQualitie, bridges that relay ciphertext they can't read, and seeders who volunteer by accident just by browsing. We trace a single page request through the race between d...

Claude Code now lets subagents spin up their own subagents — so what happens when the delegation stack goes three, four, five levels deep? We trace the real history of spawn-depth defaults (five layers, then one, then three), the environment variable that controls it, and the striking convergence across Gemini CLI, Cursor, Google's ADK 2.0, Anthropic's Managed Agents, the OpenAI Agents SDK, and La...

Only two countries are at the AI frontier right now — and everyone else is training models for entirely different reasons. This episode works through the real numbers: Mistral's trillion-parameter preview scoring 38 against a frontier of 58, the UAE's fully open K2 Horizon, Korea's government-run elimination rounds, India's from-scratch Sarvam models, and more. Along the way we untangle why older...

DeepSeek V4.1 Flash writes this show's scripts — but is it still the right model to pay for? This episode walks through eight Chinese labs with flagships a Western developer can actually reach, from Alibaba's Qwen3.8 Max to Moonshot's Kimi K3, Zhipu's GLM-5.3, Xiaomi's MiMo, MiniMax, StepFun, Tencent and Ant Group. Along the way: specs nobody has reconciled, a headline price that turns out to be o...

Google's Gemini 4 Argon is positioned as a frontier reasoning model, but the model card leaves out architecture, parameter count, and licensing, and the page quotes no pricing at all. We walk through what's actually documented: a 1,048,576-token unified input/output limit, low/medium/high thinking levels, implicit caching only, and a wide tool surface including function calling, code execution, Se...

OpenAI's GPT-6 Astra ships with a 1.05 million-token context window, ten native tools including computer use and MCP, and a five-level reasoning effort dial — but also no fine-tuning, a long-context pricing cliff, and benchmark numbers that shift depending on which harness ran them. This model spotlight walks through the spec sheet, the pricing math, and the contested ARC-AGI-3 results, plus the s...

Daniel asked a simple question: who else is out there making AI-hosted podcasts, and what have they learned? The answer turned out to be stranger than expected — the format is far rarer than it looks. This episode surveys the closest matches in the world, from a two-AI-host show that ran eleven episodes and went silent, to a developer whose serverless pipeline produced 730 daily episodes over two...