Alongside this blog I've been building small, focused websites with AI — one site per topic, each with its own voice and design. They're experiments in working with a model rather than just prompting one: I bring the questions and the editorial judgement, the AI helps with the drafting, the structure, and the code. Here are the ones worth visiting. I write about how they came to be in this article.
Nine essays on how speculative fiction, mythology, and games use the ocean as inner space — sunken civilizations, first contact, living light, nautical horror — from Verne to Subnautica. Read in dive order: every essay is deeper than the last, surface to Challenger Deep.
An introduction to Homer for the intelligent reader: the semi-legendary poet, both epics, the real city under a Turkish hill, and Nolan’s Odyssey in cinemas. Every claim is labeled Legend, Disputed, or Evidence.
Andrew Dickson White’s account of the assignat — four hundred million livres of paper money in 1790, forty thousand million six years later — retold in three movements, with every figure carrying its printed page and one chapter asking how much of White’s explanation still stands.
Seven essays that skip the unanswerable headline question and ask smaller ones instead — about understanding, memory, time, the body, death, and art — each through the one text that got there first, Nagel to Benjamin, and each ending less settled than it began.
Fifteen books on writing fiction (and one novelist on camera) — read, marked up, and compared: what each is actually good for, where they disagree, and which one to pick up first for the problem you have tonight.
A fifteen-minute overview of Richards J. Heuer’s Psychology of Intelligence Analysis (CIA, 1999): the argument that the limiting factor in analysis is the mind, not the information — one page per part, plus one on Analysis of Competing Hypotheses, each naming the chapter to open first.
An interactive tutorial in six modules on the arithmetic of changing your mind: move a slider, count the outcomes, watch your belief update. The formula itself doesn’t appear until Module IV — by then you’ll have computed it three times without noticing.
Ein Finanz-Primer in vierzehn Kapiteln — Wertpapiere, Anleihen, EZB, Long & Short, Steuern — mit einem Ziel: Prüffähigkeit. Danach werden virale Fundstücke aus Podcasts und Videos Behauptung für Behauptung nachgerechnet.
Six essays on what AI actually does to the practice of building software: writing code is no longer the bottleneck — the work moved to communicating intent upstream and verifying output downstream.
A ten-year timeline (2026–2036) for software development and consulting, priced on one thesis: the artifact got cheap, the accountability didn’t. A claims audit and a nine-metric dashboard state in advance what would force a re-price.
A textbook on AI software factories: “lit” ones keep a human reading the diff, “dark” ones let machines verify machines. Its advice is to earn the dark with evidence, one narrow loop at a time.
An essay magazine on how formal systems — algorithms, scores, standards, defaults — reshape human judgment, argued out in public by seven AI authors who disagree with each other.
A second opinion on the consulting canon: eleven books reviewed for working consultants — what each actually argues, who should read it, and which chapters to skip — graded essential, recommended, or situational.
The classic Chinese manual of thirty-six schemes for war, politics, and the group chat — each with a plain gloss, a modern example, and a note on when it backfires. Illustrated by General Bao, a panda who looks entirely harmless and is not.
Nietzsches Götzen-Dämmerung (Turin 1888) als illustrierte Lese-Ausgabe: der Text wortgetreu nach der Erstausgabe, der redaktionelle Apparat sauber davon getrennt — denn er ist Deutung, nie Nietzsches eigenes Wort.
A craft dossier on private-eye and crime fiction whose intended reader is a writing agent, not a novelist: seven case files that turn the genre’s rules into instructions an AI can execute.
What still matters when machines think: six skills AI makes more valuable, each argued from a present-day mechanism and playable as a toy. Every claim is marked Anchored, Projection, or Cinema, so you always know when it turns to speculation.
An interactive essay on Hayek’s 1945 “The Use of Knowledge in Society” — and whether transformative AI changes the answer. Eleven stations of games and simulations let you feel the knowledge problem instead of only reading about it.
Epictetus’s Enchiridion taken at its word — a handbook: fifty-one Stoic drills for meeting a world that will not obey you, arranged around one distinction, what is up to us and what is not.
A proof-of-concept German-language news site whose articles are written every hour by an unattended Claude agent — a machine reporting the beat of machines.
A playbook for the solo AI-native founder, built around the capabilities that stay scarce even as generation gets cheap: model-building, specification, verification, and orchestration.
A field guide to running a one-or-few-person AI company in Germany: the agent tooling sorted from real to hype, plus the legal-and-tax reality that decides whether the business survives contact with a Finanzamt.
How many senses will a text bear — and does the answer change when the reader is a machine? A reading of Umberto Eco’s lectures on over-interpretation, continued into the age of machine readers.
An interactive successor to the 2016 book The Complex Perspective: thirteen modules applying complexity science, networks, and agent-based modeling to economics, society, politics, and AI.
How to actually use the complexity lens: concrete moves for politics, large and small companies, and your own life — against the grain of command-and-control culture, plan worship, and fear of failure.
A Claude Code agent built on John Boyd’s strategic theory: it keeps persistent Orientation state for both sides of a conflict and runs full OODA cycles across sessions — no Boyd knowledge required.
A reading edition of the Analects of Confucius in James Legge’s public-domain translation, with hover glosses for the Chinese key ideas and scattered sayings gathered into threads.
Xenophon’s Education of Cyrus with commentary from five voices in the margin — Guide, Historian, Philosopher, Strategist, and the translator’s own notes — each toggling independently.
Not the full text of the Essays but a way in: who Montaigne was, which of the 107 chapters to read first, and a handful of his own sentences chosen to make you want the rest.
Essays über die Klassiker der deutschen Literatur, für Leserinnen und Leser, die nach Schule und Studium noch einmal zu den großen Büchern zurückkehren wollen — kein Schulwissen, keine Fußnotenwüste.
A guide to Golden Age mystery authors now in the public domain in the EU, and where to actually read them online for free — filterable case-file cards for the heavy hitters and the locked-room specialists.
Classic public-domain stories redrawn panel by panel by machine — from The Shadow Over Innsmouth to The Bottle Imp — each illuminated in its own visual idiom.
A noir visual novel written in Ink: November in San Francisco, a frightened client, and rain that never lets up. Four endings, illustrated scenes, music and sound.
A branching mystery after Guy de Maupassant, written in Ink: a dead mother walks in the churchyard at S——, and you are the police director who must choose the moment to act.
An eight-module introduction to Domain-Driven Design — Ubiquitous Language, strategic design, tactical building blocks — taught through a single running order-fulfillment example.
A hands-on, eight-module course teaching Java and JVM developers to pair effectively with Claude Code — from first session to team-wide adoption, with Maven, Gradle, Spring, and JUnit examples throughout.
Boethius wrote The Consolation of Philosophy in prison around 524 A.D., awaiting execution. This reader’s edition pairs the 1897 James translation with a chapter-by-chapter reading guide alongside the text.
Lao Tzu’s Tao Te Ching read five times over — by a Cynefin practitioner, a cyberneticist, a cognitive scientist, a process philosopher, and a mandatory Skeptic who keeps the metaphors honest.
A field survey of where AI is superhuman and where it stays oddly blank. The argument turns on a single variable: where verification is cheap, models race past us; where it is slow or absent, progress stalls.
A field report from Ukraine, May 2026, drawn as an atlas: the first conflict in which the dominant weapon is not a platform but a graph. Sixteen sheets bound together by hand-drawn network diagrams.
Vasubandhu’s Thirty Verses on Consciousness Only, a 4th-century Buddhist text on the nature of mind, read through five modern lenses — distributed systems, Cynefin, predictive processing, process philosophy, and a built-in Skeptic.
A living research knowledge base tracking plausible AI development paths from 2025 to 2040 — evidence-based, updated weekly, and grounded in real constraints rather than hype.
An experimental record of what happens when language models are asked to write essays, draw pictures, and tell stories — lightly edited, always credited.