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...
Vizitează neapărat site-ul podcastului și susține-i creatorul: myweirdprompts.com
Autor
Daniel Rosehill
Categorie
Site-ul podcastului
Cel mai nou episod
20 nov. 2026
Unde asculți?
Podcasturi în aplicație Replaio Radio În curândPodcasturile ajung în curând în aplicație. Instaleaz-o acum și fii primul care descoperă o abordare complet nouă a podcasturilor
Episoade
Why API Email Still Relies on SMTP 31.10.2026 20:36
Developer email services like Resend seem to replace decades-old protocols with a simple HTTPS POST, but the transport layer remains entirely intact underneath. Modern API mail platforms handle queues, connection pools, and MTA operations like bounce classification and retries behind the curtain. While open-source projects now build personal inboxes on top of these APIs, DNS setup, port blocks, an...
Why Model Verbal Tics End Up Driving Script Benchmarks 31.10.2026 19:16
After relying on GPT-6.1 Sol for several weeks, production shifted to Claude Sonnet 5.5 to eliminate predictable dialogue habits. Repetitive quirks like headmasterly quiz questions, emotional forecasting, and recurring pet phrases degrade conversational realism over high-volume runs. A look at the open-source ScriptBench tool reveals why benchmarking dialogue requires prioritizing organic characte...
Verifying Generated Arabic Text as a Non-Reader 31.10.2026 18:53
Image models frequently output elegant, calligraphic Arabic lettering that is disconnected, backwards, or entirely nonsensical. Because generative pipelines lack font engines and contextual shaping passes, polished visual styling often masks broken ligatures, dot faults, and false joins. Learn the structural mechanics behind why Arabic breaks differently than Latin script in image generation, alon...
How Forensic Tools Search Android for Spyware 31.10.2026 25:07
Android forensics for detecting mercenary spyware like Pegasus has long relied on incidental logs that quickly overwrite themselves. Examining Amnesty International's Mobile Verification Toolkit and AndroidQF reveals why direct shell extraction was replaced with a custom pushed collector binary, why backups reveal almost nothing on modern Android, and how the introduction of Android Intrusion Logg...
Why Editing Vector Databases Silently Re-Embeds Text 31.10.2026 18:33
Vector databases do not reverse numbers into words when an interface displays editable text. Plain sentences sit as metadata payloads or separate records matched by foreign keys. Changing text triggers a hidden re-embedding pipeline, risking stale searches where deleted paragraphs continue answering queries. An examination of APIs from Ragie and Qdrant illustrates how deterministic chunk hashing p...
Decaying Old Memories in Vector Search 31.10.2026 25:13
Vector stores like Qdrant and Milvus now ship native decay functions to let newer records beat old ones without database pruning. This look into time-decayed vector search breaks down the core mechanics of linear, exponential, and Gaussian decay curves, how target, scale, and midpoint parameters interact, and why subtle timezone drifts silently ruin ranking. It also examines the architectural deba...
Why Swipe Decks Make Better Training Data 31.10.2026 18:40
Calling a card-stack interface a Tinder clone misses its real mechanics and measurement power. Standard design systems like Material or Apple HIG completely omit the swipe deck, leaving open-source libraries sparse and unmaintained. Yet paired with active learning engines like Label Studio, this interface transforms data labeling by triaging ambiguous edge cases. Adding explicit buttons for skippi...
What Really Runs When You Compact an Agent's Context 31.10.2026 21:30
Agent harnesses rely heavily on context compaction, but dedicated models built specifically for the task barely exist. Commercial tools like Claude Code and OpenAI use main models with built-in prompts or opaque server-side endpoints. Meanwhile, purpose-built open-source fine-tunes, token pruners like LLMLingua-2, and exotic latent compressors face practical trade-offs. Examining context editing,...
Where Harmonizable Processes Actually Work 31.10.2026 19:33
Standard signal processing assumes statistics hold still, placing spectral energy strictly along a diagonal. When frequency components correlate instead, harmonizable processes take over under everyday engineering labels like cyclostationarity and spectral correlation. This branch of mathematics explains how multi-path Doppler radar spots tied-together echoes, why edge sensors can diagnose failing...
Why Weighting Software Scores Fails to Pick Winners 31.10.2026 21:00
Multi-criteria decision analysis promises a clean, mathematical way to rank software options, but the actual weights assigned matter far less than the hidden assumptions. Normalization scales alter rankings before weights are even applied, compensatory models let high scores override dealbreakers, and adding a duplicate candidate can reverse the final standings. Operations research concepts like t...
Why LLM Output Tokens Cost More Than Input 31.10.2026 23:25
Provider rate cards charge three to five times more for output tokens than input tokens, driven by the physics of memory bandwidth bottlenecks during decode. Comparing models using industry-standard three-to-one blends misrepresents actual costs, especially as agentic harnesses and retrieval-augmented generation skew workloads anywhere from twenty-two-to-one to over a hundred-to-one. Analysis expl...
Choosing Class Breaks for Skewed Price Data 31.10.2026 18:39
Splitting continuous API prices into discrete tiers requires choosing between competing statistical philosophies. Cartography solved this problem decades ago through choropleth class interval selection, showing how variance-minimising methods like Jenks natural breaks contrast with quantiles, density cuts, and head-tail breaks. Preprocessing choices like logarithmic transforms, zero-price offsets,...
Choosing AI Models on Cost and Performance 31.10.2026 18:54
Picking an AI model requires balancing benchmark scores against runtime costs, but popular formulas often produce completely different winners from the exact same data. A simple cost-to-performance ratio inevitably favors cheap models with unusable scores unless filtered by a quality floor. Exploring methods from Pareto frontiers and knee-finding to weighted sums and TOPSIS reveals why selecting t...
Turning Anonymous Diarization Clusters into Names 31.10.2026 23:39
Turning anonymous speaker clusters into verified names requires more than matching audio clips. The pipeline relies on fixed-length speaker embeddings, pooling layers, and angular margin losses like ArcFace to represent vocal identity across varying speech lengths. While short introductions seem convenient, limited phonetic variety degrades accuracy. The mechanics of x-vectors, ECAPA-TDNN architec...
Choosing the Right Voice Activity Detector 31.10.2026 32:08
Voice activity detectors look interchangeable until deployment constraints reveal their hidden costs. Comparing Silero, TEN VAD, WebRTC, and Cobra shows that download size is only a fraction of the story: runtimes, licensing restrictions, and transition delays drastically alter performance. From WebRTC's lightweight Gaussian mixture models to Cobra's embedded runtime and Silero's multi-domain neur...
Porting Voice Models to Core ML and MLX 31.10.2026 30:46
Porting speech and audio models from Linux PyTorch environments to Apple platforms involves distinct engineering trade-offs. Core ML requires graph capture and converter compatibility to leverage the Neural Engine, CPU, and GPU, packaging specifications and weights into structured bundles. Meanwhile, MLX requires direct reimplementation of the forward pass in array code, bypassing the Neural Engin...
Running Speaker Identification Locally on Phones 31.10.2026 22:34
Running speaker identification directly on mobile devices is entirely practical without relying on cloud servers. While research environments favor heavy PyTorch pipelines like pyannote.audio, edge runtimes such as sherpa-onnx execute compact ONNX models directly on Android and iOS. This breakdown explores the practical engineering trade-offs of offline voice matching, comparing embedding models l...
How Audio Models Tell Voices Apart 31.10.2026 28:23
Voice recognition systems often struggle with speaker diarization: tracking who spoke when across overlapping speech, whispers, and brief interruptions. This exploration unpacks the mechanics behind pyannote and end-to-end neural diarization (EEND), examining how voice activity detection, speaker embeddings, powerset segmentation, and Bayesian hidden Markov models like VBx separate anonymous speak...
Why Financial Centres Outlive the Forces That Built Them 30.10.2026 25:12
Financial capitals rarely coincide with political seats, surviving shifts in transport, technology, and regulation. Looking at Charles Kindleberger's theories on transport breaks, this episode traces how trade routes and institutional ecosystems anchor commerce in places like Zurich and New York. It explores how Zurich leveraged the Gotthard railway tunnel over historic rivals, how wartime credit...
Why Modern AI Rests on Statistics 30.10.2026 23:16
Modern artificial intelligence often frames its breakthroughs in computing power, yet its fundamental mechanisms rely entirely on classical statistics. Adding petabytes of scraped text cannot correct for systematic sampling bias, and monitoring models requires understanding distribution tails rather than relying on average latencies. From Ronald Fisher's agricultural experiments and the foundation...
What Theoretical Physicists Actually Do All Day 30.10.2026 29:42
Official labor counts list under twenty thousand physicists in the United States, yet eleven million jobs run on physical principles under titles like product designer or engineer. Inside academia, the theoretical physicist's reality clashes with the solitary genius myth. Examining interviews with faculty and researchers at CERN and national labs reveals a career defined by fractured calendars, en...
MWP 4 Kids: Why the Moon Changes Shape 30.10.2026 9:43
The moon seems to change from a full glowing circle to a slim crescent sliver, but it never actually loses its shape. Using a bright lamp and a simple round orange, a calm bedtime demonstration reveals how sunlight always illuminates exactly half of the lunar sphere. As the moon completes its twenty-nine-day orbit around Earth, varying angles show different amounts of that illuminated face, creati...
Where Email Attachments Actually Live and Go 30.10.2026 23:12
Email attachments seem like simple files handed between computers, but MIME transforms them into text packed directly inside the message body. Storing all that data poses an enormous engineering problem: attachments accounted for 85 percent of storage at Mail. Ru before deduplication. Mail systems dissect complete messages on disk, balance reference counters, hash encoded payloads, and handle tran...
Why Private IP Addresses Look the Way They Do 30.10.2026 22:45
Popular network charts often label private IP ranges by organization size, claiming 10.0.0.0 is for enterprise and 192.168.0.0 is for homes. In reality, RFC 1918 simply preserved historical class boundaries, and any network can use any private block. What actually matters is subnet sizing, route specificity, and the routing collisions that occur when hotel Wi-Fi and home networks share identical d...
Why Modern Audio Datasets Ship as Parquet Files 30.10.2026 31:00
Parquet was designed for untidy tabular records, but dataset repositories increasingly use it to bundle audio recordings alongside transcripts and timestamps. The file structure splits data into row groups, column chunks, and pages, allowing readers across different programming languages to selectively decompress columns without streaming full datasets. From footer-based metadata and dictionary fa...
Podcasturi asemănătoare
Replaio nu este editorul podcasturilor; numele emisiunilor, coperțile și materialul audio aparțin autorilor lor și sunt distribuite prin fluxuri RSS publice