writer
Joseph Rainer
From the journal — 4 pieces by this author
- August2026
- 16minAugust 19The Machine
Does AI Actually Read Your Documents? We Resurrected a Dead Twain Character to Find Out
We renamed Jim, brought Buck Grangerford back from the dead, and planted a single fake sentence in the middle of 112,000 words. Then we asked Fable 5, GPT-5.5, GLM 5.2, and 6 other models what they’d read. Most of them hadn’t
By Joseph Rainer
- July2026
- 7minJuly 31The Desk
Markdown: the language you use to talk to ChatGPT and Claude
A quick guide to plain text formatting with no hidden styles: why hash signs and asterisks were right under your nose for twenty years, and now suddenly you need them to talk to AI models
By Joseph Rainer
- 7minJuly 30The Desk
When a Stranger's Text Can Give an AI New Orders
Prompt injection lets an outsider hide commands in a page or a document and make an AI follow them instead of its user. Prompt injection is already in the wild, spanning from courtrooms to state-run espionage, and there is no clean fix — only ways to limit the exposure.
By Joseph Rainer
- 11minJuly 3The Method
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
By Joseph Rainer
