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 18, 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 Predictable Structure of Estranged-Parent Emails 13.11.2026 22:24
An unopened email with a one-word subject line can still be predicted sentence by sentence. Exploring a listener's question about screening messages from an estranged father, this analysis unpacks the structural gaps that produce formulaic correspondence. It examines how DARVO, the grievance ledger, and gifts with strings function like a legal brief rather than a conversation. Drawing on research...
Why Siblings Experience Different Childhoods 13.11.2026 23:15
Two siblings can grow up in the same troubled house and come away with entirely different childhoods. Sparked by a listener's conflict with his peacemaking brother, this episode examines the research on nonshared environments, differential parental treatment, and family triangulation. It unpacks why "golden child" and "scapegoat" labels are shakier than they seem, how a 30-study meta-analysis expl...
When Distance Is How You Honour a Parent 13.11.2026 28:47
The fifth commandment's codified text—from Maimonides' Mishneh Torah to the Shulchan Arukh—forbids humiliating a parent in the moment, not maintaining contact with an abusive one. Prompted by a listener facing decades of scolding letters and grievance lists, this conversation unpacks rabbinic rulings from the Maharik, the Rema, and Rabbi Dov Linzer's 2018 responsum. Together they show how Jewish l...
What Access Point Firmware Updates Actually Change 13.11.2026 28:43
Clicking update on a Ubiquiti access point hides a surprisingly detailed changelog. Examining firmware 8.8.14, 8.7.11, and 6.7.54 reveals that AP updates are dominated by wireless bugfixes, WiFi 7 Multi-Link Operation, and 6 GHz regulatory table changes for eight countries—while security CVEs actually live in separate UniFi application bulletins. The code itself is a four-layer stack: an aging Ope...
Chrome, Firefox, and Verifiable Privacy Claims 13.11.2026 29:31
Trusting a tech giant like Google over a smaller privacy vendor comes down to a bet on incentives, but verifiability matters more than company size. This breakdown compares what Chrome collects by default, why Google's Privacy Sandbox failed, and how Firefox's built-in tracker blocking actually works. It also examines Proton VPN's independent Securitum audits and why skepticism of Romanian VPNs li...
Who Trains on Your AI Chats and What Happens Next 13.11.2026 21:12
Consumer tiers of OpenAI, Anthropic, and Google Gemini all train on chats by default, while API and business tiers do not. Even with opt-outs enabled, feedback buttons, safety reviews, and separate Codex settings can still expose conversations. The bigger privacy risk isn't models memorizing secrets—it's five-year retention windows and human reviewers. Also covered: how OpenRouter publishes provid...
How Self-Consuming AI Models Decay Generation by Generation 13.11.2026 26:10
Even with infinite samples and a flawless learning procedure, an AI model trained recursively on its own outputs still degrades. Model collapse isn't just stasis—each generation resamples and amplifies its predecessor's errors, starving the rare tails of a distribution while the average gets richer. From Shumailov's three compounding error sources to Briesch, Sobania, and Rothlauf's logic-expressi...
Software Flaws That Became Permanent Infrastructure 13.11.2026 21:56
A single shortcut in 1965 or 1983 can calcify into global infrastructure when every workaround locks the flaw in place. This countdown of accidental software engineering designs explores Excel's phantom 1900 leap year, Windows's uncreatable DOS device filenames like CON and NUL, Make's invisible tab requirement, and Tony Hoare's billion-dollar null reference. It also examines how Excel's date auto...
How AI Agents Spot Bugs Where Systems Meet 13.11.2026 22:05
AI agents are excelling at finding "seams"—the overlooked friction points where individually sensible software components or decisions collide. Starting with a Home Assistant and Zigbee2MQTT outage where a hardcoded ten-restart watchdog limit and a retained MQTT message kept lights dark for hours, the discussion explores how Claude traced the root cause across separate codebases. It also examines...
Self-Hosted AI Browsers Still Lack Vault-Brokered Logins 13.11.2026 20:53
Running AI agents that log into websites creates a sharp split: the most mature vault-brokered credential tools, like 1Password Secure Agentic Autofill, are tied to cloud-only browsers like Browserbase, while true self-hosted options leave credential handling to you. This breakdown maps the commercial and open-source landscape—from Browserless and AWS Bedrock AgentCore Browser to Anchor Browser, S...
The Search for a Supervised AI Agent Browser 13.11.2026 19:31
AI browser agents like Claude in Chrome and Playwright's extension mode hijack your real tabs because your logged-in state is their whole value. The alternative is a supervised workspace like Citrolabs' ego lite, which isolates agents into separate spaces, or self-hosted tools like Auto Browser. Cloud platforms including Browserbase, Browser Use Cloud, and Cloudflare's Browser Run already offer li...
How a GeForce 8800 Side Note Became the Floor Under AI 13.11.2026 30:26
CUDA started as a side note on the 2006 GeForce 8800 gaming card, yet it became the software floor under PyTorch and modern AI. Tracing the platform from Ian Buck's Stanford PhD project Brook to NVIDIA's G80 launch, this deep dive explains what a GPU platform actually is in plain English. It breaks down heterogeneous memory, kernels, thread blocks, warps, and the PTX virtual instruction set, and e...
How Slurm and Kubernetes Split GPU Workloads 13.11.2026 32:24
Standing up a GPU cluster splits into two stacked problems: scheduling jobs onto hardware and keeping collective communication fast. Distributed training stalls if even one GPU is missing, yet Kubernetes's default scheduler places pods one at a time. This walkthrough compares Slurm's atomic batch queues with Kubernetes tools like the NVIDIA GPU Operator, Kueue, Volcano, KubeRay, and Kubernetes 1.3...
CUDA's Small Start and Its Open Ecosystem 13.11.2026 27:38
NVIDIA's CUDA platform began with a tiny team—Ian Buck, John Nickolls, and a handful of engineers—turning a Stanford research language called Brook into a decade-long money-losing bet. Today, roughly two to three thousand internal engineers maintain a closed core of drivers, cuBLAS, and GSP firmware, while millions of outside developers build on top. The real tension sits at the boundary: NVIDIA n...
How Streetcar Suburbs Hit the Transit Sweet Spot 13.11.2026 26:38
Streetcar suburbs were built around electric tram lines in the late nineteenth century, yet most still function decades after the rails were torn out. Their secret is a specific density band—roughly seven to fifteen homes per acre—that lands right on modern thresholds for frequent bus and light rail service. By pairing narrow lots and front porches with continuous corridors, they remain the rare h...
Plugging a Custom Voice Model into Wayland 13.11.2026 26:03
Most open-source voice projects assume you need speech recognition, leaving a developer who already trained their own models with nowhere to plug them in. Exploring tools like VoiceMode, Fono, VoiceUse, and the OpenAI Agents SDK reveals a gap between voice channels and actual Wayland desktop control. The conversation covers how to pair a fast command classifier with an LLM fallback, and why Waylan...
Why Home Voice Agents Still Need Home Assistant 13.11.2026 20:39
Home Assistant already speaks MCP in both directions, so demoting it to one tool behind a server-hosted voice agent is technically possible with frameworks like LiveKit Agents, Pipecat, and Speaches. The real bottleneck is the hardware on the ceiling: ESPHome satellites and the Voice Preview Edition expect Home Assistant to drive the pipeline. This look at the architecture compares full server-sid...
USB Far-Field Mics and the Echo Cancellation Catch 13.11.2026 21:07
A far-field microphone array is really a small computer with microphones attached, and the DSP chip behind them is the actual product. For a DIY Home Assistant and Whisper satellite, USB Audio Class boards like Seeed's ReSpeaker XVF3800, the older ReSpeaker v2.0, and the ReSpeaker Lite offer onboard beamforming and noise suppression, unlike generic AliExpress arrays or the PlayStation 3 Eye. The c...
The Wake-Word Race Inside Home Assistant Satellites 13.11.2026 27:44
Multi-room Home Assistant voice setups rely on cheap satellites—like the Voice Preview Edition or ESP32-S3-BOX-3—streaming audio to one central hub. Because wake-word arbitration is a pure network race with no volume comparison, the farthest satellite can win and silently drop your command. We unpack the real-world pitfalls of multi-room voice: why turning down sensitivity backfires, how power-sav...
Why Kitchen Smells Hijack ADHD Focus Like Noise 13.11.2026 25:02
Adults with ADHD often smell just as well as anyone else, yet brain imaging reveals their olfaction-associated regions handle those signals differently. This episode examines whether kitchen aromas like brewing coffee can hijack focus the same way background noise does. From the contested clinical status of sensory processing disorder and sensory over-responsivity to the changing-state effect in s...
Inside Laptop Microphones and Why They Hurt Dictation 13.11.2026 20:07
Almost every mainstream laptop—from the ThinkPad E15 to the Dell XPS 13 and Surface Laptop 7—ships with just two microphones, leaving only the MacBook Pro 14 with three. This deep dive explores whether you can spec a better, dictation-grade beamforming mic on a Framework laptop, why premium MEMS capsules cost just a couple of euros, and how call-focused audio processing like Voice Focus actually t...
When Fine-Tuning Over-Generalizes Instead of Forgetting 13.11.2026 26:10
Fine-tuning a speech model on IP addresses can make ordinary phrases like "ten one" transcribe as "10.1." Rather than catastrophic forgetting, this quieter failure is over-generalization, where a model applies a new surface pattern everywhere. The discussion unpacks why labels act as a teaching surface, how to catch formatting bleed using contrast sets and B-WER/U-WER splits, and what replay-share...
Five Projects to Build a Language Model From Scratch 13.11.2026 27:49
Training a language model from scratch used to require massive budgets, but a thousand-fold drop in compute costs has changed the math. A de facto five-step ladder—centered on Andrej Karpathy's micrograd, Zero to Hero lectures, minbpe tokenizer, microgpt, and nanochat—takes learners from a 100-line autograd engine to a working GPT. Along the way, building a byte-pair tokenizer explains why models...
Donating Bandwidth Without Becoming an Exit Node 13.11.2026 21:30
Donating bandwidth to help people bypass internet censorship doesn't mean turning your home connection into a risky exit node. Tools like Tor's Snowflake, Psiphon's Conduit, Lantern's Unbounded, MoaV, and the WEPN pod all act as entry relays, passing encrypted traffic to external servers. We break down how each tool works, why there's no native router package, and the real-world tradeoffs between...
Serving Parakeet Speech Models on Cheap CPUs 12.11.2026 21:54
Deploying custom speech models does not require keeping an expensive GPU running around the clock. Exploring NVIDIA's Parakeet TDT 0.6B architecture reveals why quantized INT8 weights run exceptionally well on everyday CPUs, how serverless cold starts shrink from minutes to two seconds using GPU memory snapshots, and where fine-tuning accents risk running into quantization errors. The trade-offs b...
Similar podcasts
Replaio is not a podcast publisher; show names, artwork and audio belong to their authors and are distributed through public RSS feeds.