Illustrated and beautifully written in the first half of the 15th century, the Voynich manuscript defies any attempt at making sense of it. To this day it has mystified everyone who has taken a serious look at it.

At first glance the drawings give the impression of a book about plants and herbs, astrology, cosmology, perhaps occult or secret women knowledge. All accompanied by paragraphs of text in an unknown script – presumably descriptions and explanations of the drawings. Yet beyond this pretense of plausibility nothing seems to admit a conventional interpretation: The concrete drawings don’t quite match anything in the real world, the abstract drawings don’t quite depict anything familiar either, the text does not quite behave like any known natural language, the script has a logic of its own, everything about it feels dream-like.

In the talk I will present the process and results of my recent immersion into the manuscript. This includes a tour of the mysteries of the book, recent academic progress and ideas of interpretation, my personal approach to engaging with it, the tools rapidly developed with Claude Code that emerged along the way (now public, currently at https://ling.squoze.net/ola/), and of course: some novel ideas and observations made a long the way.

This experience has shown how much custom AI-built tooling can boost productivity for individual and personal research. Though it is not primarily a talk about AI, I hope to demonstrate how it can make the tumble down a rabbit hole much more enjoyable.