The Method · Long Read
Hemingway's Writing Style, as Faked by Five AI Models
Four classics, one station scene, twenty paragraphs of machine prose, and not a single real voice
TL;DR: The words were there. The authors weren't. ?? We all have a rough idea of how large language models work. One thing follows from how they're built: they are brilliant at choosing words; strictly speaking, that's all they do. Imitating the style of an author the model knows inside out, one it trained on extensively, is a perfect way to gauge what it can effectively do, what it can barely do, and what it simply cannot do at all. We asked five flagships — Gemini 3.1 Pro, GPT-5.5, Claude Fable 5, Claude Opus 4.8, and DeepSeek V4 — to write one paragraph each in the styles of four great writers, familiar to both models and readers alike: Hemingway, Edgar Poe, McCarthy, and Faulkner. The pr
Read more →Latest
All recent →NLP Is Back. Now Its Victims Are Neural Networks
Half a century ago, two Californians figured out how to control people through language. Today the same thing is being done to neural networks — and they call it a profession
Deep Research: Why It Is Slow, Expensive, Beautiful, and Dangerous
What makes you trust a flattery generator that trawls the internet for garbage
How to Transfer Your AI Memory Between ChatGPT, Claude, Gemini, and Grok (2026 Guide)
A quick guide to moving house: how to legally pull your own dossier out of ChatGPT, Claude, or Gemini and get a new model to recognize you instantly
The Wolf, Goat and Cabbage Problem: AI's Favorite Puzzle Is Solved — So We Built a Harder One
For two years a medieval river-crossing riddle humiliated ChatGPT. The top models finally passed it, so we ran a new common-sense trap past a wide assortment of LLMs. Only the most expensive models kept their footing
Can AI Write a Book for You? Yes, But You're Going to Regret It
An AI model can genuinely generate grammatically clean text at any length — but the prose will be terrible, and along with it you'll get plot amnesia, cardboard characters, impostor syndrome, and a whole host of other problems.
What Anthropic's Fable 5 Revealed: A Hidden Layer Decides What Your AI Will Give You
Anthropic launched its most capable public model, admitted it had been secretly degrading some answers, apologized — and three days later the government pulled the plug. Here’s what writers should take from these events.
- 01 ——
The Craft
4 pieces Writing as craft — voice, style, editing, long-form. - 02 ——
The Machine
8 pieces How AI works and where it breaks — memory, context, hallucinations. - 03 ——
The Method
7 pieces Working with AI — prompts, workflows, roles, Deep Research. - 04 ——
The Desk
3 pieces The writer's working environment — editor, sources, focus, tools. - 05 ——
The Stack
6 pieces Models, tools, pricing, market — Claude vs GPT, BYOK, industry shifts.
