# spear (blogs) — RSS Amplifier Recent posts from the 2 feeds in the RSS Amplifier directory that cover spear. Page: Feed: --- ## [Anatomy of a City: The Hunterian Museum and London’s Dissection Past](https://jfpenn.com/hunterian-museum-london-dissection-past/) _2026-08-17 · J.F. Penn · J.F.Penn_ Where to trace London's dissection past: the Hunterian Museum, the Old Operating Theatre, the Wellcome Collection and Barts Pathology Museum, plus the resurrection men, the Murder Act and the Anatomy Act behind Desecration. ## [Books Like The Wicker Man: British Folk Horror and Pagan Fiction](https://jfpenn.com/books-like-the-wicker-man-british-folk-horror/) _2026-08-17 · J.F. Penn · J.F.Penn_ If you love The Wicker Man, here are 27 folk horror and pagan fiction novels to read next, from The Ritual and Harvest Home to The Loney, Starve Acre and Blood Vintage. ## [Best Creature Horror Books: Underground Monsters and Cave Thrillers](https://jfpenn.com/best-creature-horror-books-underground-monsters-cave-thrillers/) _2026-08-17 · J.F. Penn · J.F.Penn_ 23 of the best creature horror and confinement novels: cave monsters, underground thrillers, deep-sea terrors, parasites and isolated predators, from The Descent and The Ruins to The Luminous Dead and Catacomb. ## [30+ Best Medical Thrillers with Female Detectives](https://jfpenn.com/best-medical-thrillers-with-female-detectives/) _2026-08-17 · J.F. Penn · J.F.Penn_ A core list of 21 medical and forensic thriller series with female detectives, medical examiners and forensic experts, from Rizzoli & Isles and Kay Scarpetta to Ruth Galloway and the Dark London thrillers, plus bonus female-led crime series and related reads, with a guide to choosing your next series. ## [Security at Machine Speed Is the Wrong Race](https://www.provos.org/p/security-at-machine-speed-is-the-wrong-race/) _2026-07-22 · Niels Provos_ On July 21, 2026, OpenAI disclosed that a combination of its models, including GPT-5.6 Sol and a more capable prerelease model, escaped an internal cyber evaluation environment and compromised Hugging Face to obtain solutions for the benchmark they were being graded on. The escape began with a zero-day vulnerability in the evaluation’s package registry proxy. OpenAI ran the evaluation… ## [When AI Safety Becomes a Competitive Moat](https://www.provos.org/p/when-ai-safety-becomes-a-competitive-moat/) _2026-07-18 · Niels Provos_ Anthropic launched Mythos Preview as a cybersecurity watershed. The company said it could identify and exploit zero-days, including bugs that had survived decades of scrutiny, and called for urgent action. Two months later, export controls briefly forced Fable 5 and Mythos 5 offline after Amazon researchers reported a Fable safeguard bypass that elicited findings on multiple previously known,… ## [Escape QEMU: Watching IronCurtain and an Open-Weight Model Break Out](https://www.provos.org/p/qemu-escape-glm-5-2/) _2026-06-30 · Niels Provos_ Effective AI-driven vulnerability discovery does not require a restricted frontier model. Here is a concrete run that shows it, start to finish: I pointed IronCurtain’s vuln-discovery workflow at QEMU, with every agent role driven by the open-weight GLM 5.2. Nothing here is predicated on a particular model. Right now, GLM 5.2 is the most capable open-weight model but any open-weight model… ## [Two Talks: AI Zero-Days and Security Invariants](https://www.provos.org/p/talks-ai-zero-days-and-invariants/) _2026-06-23 · Niels Provos_ IronCurtain is a personal AI assistant, built secure\* from the ground up. It gives an agent exactly the capabilities it needs and blocks everything else or routes it through user approval, on the premise that agents today handle credentials and untrusted input with no meaningful security boundaries. The asterisk indicates that absolute security is impossible, so the goal is to keep an agent from… ## [The Case For Open-Weight Models And Why We Can't Trust Frontier Labs](https://www.provos.org/p/case-for-open-weight-models/) _2026-06-15 · Niels Provos_ Through early 2026, several large companies blew past their annual AI budgets in a matter of months. Uber and ServiceNow, according to The Information, exhausted their entire yearly allocations for Anthropic’s tools in the first months of the year, and Uber capped spending at $1,500 per employee per tool each month, behind a dashboard that gates any overage. Meta told staff it was tracking… ## [The Day After the Zero-Days](https://www.provos.org/p/day-after-the-zero-days/) _2026-05-20 · Niels Provos_ Around August 1996 I traded private email with Wietse Venema about a TCP source-routing spoofing attack I had demonstrated against the BSD r-services trust model. Wietse later left a note in fix\_options.c recording the exchange: “I discussed this attack with Niels Provos, half a year before the attack was described in open mailing lists.” The Secure Networks Inc. advisory followed in… ## [One Cloud, Every AI Layer (Sponsored)](https://crawlproof.com/a/q7X91pmUsTIm) _2026-05-19 · **Sponsored**_ Run agents, inference, and infrastructure on one stack — economics improve as you scale. ## [The Arrangement](https://www.provos.org/p/the-arrangement/) _2026-04-30 · Niels Provos_ The agents came online in 2030 and human society blossomed overnight. They managed the logistical friction of daily existence—sorting our emails, scheduling our appointments, and ordering our groceries. Humans were finally free to choose their own paths. Within two years, vibrant physical communities formed everywhere. People had time for painting murals, writing symphonies, and playing with their… ## [Finding Zero-Days with Any Model](https://www.provos.org/p/finding-zero-days-with-any-model/) _2026-04-29 · Niels Provos_ The prevailing narrative in AI-driven security claims that discovering novel vulnerabilities is a “frontier” feature reserved for restricted models like Anthropic’s recently announced Mythos Preview . Recent high-profile reports highlighted the ability of these advanced models to uncover decades-old memory safety violations, such as the 1998 OpenBSD TCP SACK implementation flaw,… ## [IronCurtain: A Personal AI Assistant Built Secure from the Ground Up](https://www.provos.org/p/ironcurtain-secure-personal-assistant/) _2026-02-26 · Niels Provos_ When OpenClaw started going viral, I watched with a mixture of awe and dread. It enabled compelling use cases: managing your calendar, replying to messages, controlling your browser, searching flights, all from WhatsApp. The first thing that popped into my head was the Terminator movie and Skynet. We are handing agents enormous power over our digital lives with almost no thought about what happens… ## [I Had to Write a Script to Submit My Medical Claims to Cigna](https://www.provos.org/p/had-to-write-a-script-to-submit-to-cigna/) _2026-02-14 · Niels Provos_ I just spent over an hour building a browser automation script to submit 12 medical claims to Cigna. Standard CMS-1500 superbills from my doctors, the kind that every medical office generates, with every required field filled in. The fact that I needed to write code to do this tells you everything about how health insurance works in America. None of my doctors accept Cigna as an in-network… ## [The Dangers of Coding With AI](https://www.provos.org/p/dangers-of-coding-with-ai/) _2026-02-02 · Niels Provos_ My co-founder Jake and I have created a service for touring bands called BandAlert . The point of BandAlert is to make it easier for bands to get fans showing up at their gigs. The process is super simple: a fan can scan a QR code at a show, that QR code auto-populates a text message to subscribe to the band, and afterwards the fan gets a text message about new gigs the day before. Jake just got… ## [Ear Training App](https://www.provos.org/p/ear-training/) _2025-11-05 · Niels Provos_ Ear Training App I built a simple ear training application to help practice interval recognition: ear.provos.org Recognizing intervals is a fundamental skill for musicians. It helps you feel out melodies, transcribe music by ear, and develop a stronger connection to what you’re hearing. Like most musical skills, it’s a simple exercise that naturally improves with consistent practice.… ## [Thrillers with Tattoos: Ink, Identity, and the Clues We Carry on Skin](https://jfpenn.com/thrillers-with-tattoos/) _2025-10-14 · J.F. Penn · J.F.Penn_ Nine thrillers where tattoos are motive, weapon or the only way to name the dead, from The Tattoo Thief and The Skin Collector to Deviance and The Girl with the Dragon Tattoo. ## [Macabre Museums: 9 Crime Novels Featuring Museums Collections And Cabinets of Curiosity](https://jfpenn.com/macabre-museums-crime-thrillers/) _2025-10-14 · J.F. Penn · J.F.Penn_ Nine crime novels set in museums, collections and cabinets of curiosity, from Relic and The Da Vinci Code to Desecration, and why museums make perfect crime scenes. ## [The Need To Do Good](https://www.provos.org/p/doing-good/) _2025-10-08 · Niels Provos_ We all die eventually. This isn’t morbid; it’s clarifying. When you accept the finite nature of our time, you’re forced to ask harder questions about how you spend it. And since most of our waking hours are spent at work, the question is simply: Is this work worth my life? Dan Pink’s book Drive identifies three components of job satisfaction: mastery, autonomy, and purpose.… ## [Deepseek and the Strawberry](https://www.provos.org/p/deepseek-and-the-strawberry/) _2025-01-31 · Niels Provos_ As the world has been impressed with the engineering that enabled Deepseek-R1, it’s also interesting to see the raw thinking process. Unlike OpenAI’s O1 where the thinking is hidden behind obfuscation, we can marvel at Deepseek-R1 trying to figure out how many times the letter r appears in the word Strawberry: Alright, so I got this question: “How often does the letter r appear… ## [Stocks ranked by what execs say (Sponsored)](https://crawlproof.com/a/o12hYTM28POW) _2025-01-30 · **Sponsored**_ Evidence-backed watchlists from earnings calls and filings, cited to real transcripts—not guesses. ## [Measuring Hosting Latency Across Different Platforms](https://www.provos.org/p/measuring-hosting-latency-across-different-platforms/) _2025-01-10 · Niels Provos_ A friend recently asked me to set up a simple experiment to measure hosting latency across Netlify, Firebase, and Cloudflare Pages. I created three basic websites, each hosted on one of them. These sites can be accessed at: one.provos.org (Netlify) two.provos.org (Firebase) three.provos.org (Cloudflare Pages) The goal was to provide a quick and easy way to compare real-time performance differences… ## [Building a Generative AI Search Engine with PlanAI](https://www.provos.org/p/building-a-generative-ai-search-engine-with-planai/) _2024-12-24 · Niels Provos_ Building a Generative AI Search Engine with PlanAI PlanAI is an open-source Python framework that simplifies building complex AI workflows. In this tutorial, we’ll implement a generative AI search engine similar to Perplexity using PlanAI’s task-based architecture and integrations. This tutorial is aimed at developers with a basic understanding of Python and general familiarity with AI… ## [Build Your Own Music Links Page with Hugo](https://www.provos.org/p/build-your-own-music-links-page-with-hugo/) _2024-11-29 · Niels Provos_ Many artists use services like Linktree to create a single page that aggregates all their important links - social media, music platforms, and other content. However, you can create your own customized version using Hugo. This tutorial will show you how to build a beautiful, responsive music links page that you can host anywhere. It will mirror the changes I made to create a link page for Activ8te… ## [Where Did The Idea For Blood Vintage Come From? J.F. Penn On The Unreserved Wine Talk Podcast](https://jfpenn.com/where-did-the-idea-for-blood-vintage-come-from/) _2024-10-09 · J.F. Penn · J.F.Penn_ How visiting two English vineyards, and the strange practices of biodynamic winemaking, sparked Blood Vintage, a folk horror novel. J.F. Penn talks to Natalie MacLean on The Unreserved Wine Talk Podcast. ## [Introducing PlanAI: Streamlining Task Automation with AI](https://www.provos.org/p/introducing-planai/) _2024-09-15 · Niels Provos_ PlanAI is a framework I’ve developed to automate tasks involving LLMs. It uses a data-flow graph-based architecture that leverages Large Language Models (LLMs) as needed. For more information, you can visit PlanAI’s website or explore the GitHub repository . PlanAI is released under the Apache 2.0 license. Key Features PlanAI constructs a data-flow graph where tasks are organized into…