What if you came across a book written in a language that nobody knows how to read? There is such a book at Yale University that has baffled historians for years. It is called the Voynich Manuscript. It is believed the 240-page book was written in 1420, by persons who remain unknown. The writings are accompanied by hand-drawn illustrations that are nothing short of fantastical. At first, it mainly attracted humanities scholars. In 1921, William Newbold, a philosopher at the University of Pennsylvania who had an interest in cryptography, claimed that a 13th-century friar wrote it as a scientific treatise. Newbold believed each arcane letter was actually a collection of minuscule symbols readable under proper magnification, which would have meant the microscope was invented centuries before we thought. After Newbold’s death, John Manly, an American literature professor and fellow codebreaker, disproved Newbold’s theory, showing that his methods were arbitrary and scientifically unreliable. Since its discovery in 1912, the 15th century Voynich Manuscript has been a mystery and a cult phenomenon. Full of handwriting in an unknown language or code, the book is heavily illustrated with weird pictures of alien plants, naked women, strange objects, and zodiac symbols. Now, history researcher and television writer Nicholas Gibbs appears to have cracked the code, discovering that the book is actually a guide to women’s health that’s mostly plagiarized from other guides of the era. So, “AI decodes mysterious 600-year-old manuscript”? Not so much. The media certainly deserves a significant amount of blame for the exaggeration. The experts just say too many steps were missed to start making any claims about the manuscript itself. And in many ways, it makes sense that attempts to crack the Voynich manuscript using “artificial intelligence” would be covered so breathlessly. A New Yorker article on the history of the manuscript describes it as “the perfect canvas on which to project our worries about the difficult and the frightening and the arcane,” and the same could be said about AI. In the contemporary media landscape, this diverse and complex group of technologies is often used as a stand-in for fears about automation and unknowable (and uncontrollable) machine intelligence. Pitting AI against the Voynich manuscript is like watching Godzilla fight Mothra: the spectacle is so fun that we don’t care or think too hard about the details. Still, for experts, the fact that the manuscript remains impenetrable might be a relief. After all, if you’ve spent years and years of your life trying to decode a mysterious document, it would probably be a bit of a blow if some bloodless machine cracks it overnight. Artificial intelligence algorithms, for example, often require large datasets for training and testing before they can be widely applied, and analysis of the Voynich manuscript can help physicists and other scientists test whether new number-crunching methods can identify meaningful patterns in vast amounts of abstract data. The 2013 Brazilian physics paper used the Voynich manuscript to illustrate how statistical physics methods can be adapted to find hidden linguistic patterns and concluded that the text didn’t seem randomly generated. And Kondrak and Hauer’s machine-learning paper focused primarily on describing the language-analysis algorithms they used to detect Hebrew as the underlying language. Even if neither theory has been accepted as a Voynich solution, they may still prove effective in other arenas.