Forscher ließen KI auf das Voynich Manuskript los – das Resultat überrascht

The Voynich Manuscript is the most studied unsolved document in history—but what have thwarted cryptographers, linguists, and computer scientists for six centuries? In 2019, an AI algorithm achieved in 84 hours what 32 research projects failed to do: a language identification with statistically significant results. After that, silence fell. This video reveals what lies within that silence. The parchment of the Voynich Manuscript dates from 1404 to 1438—yet 33 of the depicted plants demonstrably originated in Mexico, which was not officially recognized until 1493. Greg Kondrak, a computational linguist at the University of Alberta, identified Hebrew as the strongest language correlation in 2019. Zipf's Law—the statistical fingerprinting method for genuine human language—is more consistently present in the manuscript than in the complete works of Shakespeare. Not a single known forgery in history meets this condition. In 2024, an anonymous team without institutional affiliation analyzed the Voynich Manuscript using over 1,000 language models and published the raw data freely. No publicly funded follow-up project has been approved since 2019—in any country, at any university, and by any national research council. Who is behind the silence, and what is written on the last page—folio f116v—which to this day lacks a complete official transcription? The historical mystery surrounding the Voynich Manuscript touches on religious history, colonial history, and the question of how institutions decide which knowledge receives funding—and which does not. Subscribe to the channel and turn on notifications so you don't miss Part 2.