Plots, love letters and remedies: The medieval secrets being revealed by AI (www.bbc.com)

Artificial intelligence has taken a significant leap in uncovering historical secrets by cracking complex medieval ciphers, with a breakthrough achieved on the Borg cipher, a 408-page manuscript encoded with obscure symbols. This manuscript, hidden in the Vatican library for over 400 years, contains secret remedies linked to medical practices of the era. Using machine learning, researchers were able to decode the text, revealing bizarre treatments that offer insights into past beliefs about health. The effort to decode this and similar ciphers not only enhances our understanding of historical practices but also opens doors to previously inaccessible knowledge that may reshape our narratives about historical figures and events. The implications of this research are profound for the AI and machine learning community. As experts like Beáta Megyesi from Stockholm University and Michelle Waldispühl from the University of Oslo lead initiatives to develop AI tools capable of transcribing and deciphering ancient codes, they envision a future where AI can analyze images of manuscripts directly, bypassing cumbersome transcription processes entirely. Their work aims to create robust AI systems that integrate transcription, decryption, and pattern recognition to efficiently crack ciphers, a challenge compounded by the diverse languages and scripts of historical texts. This innovative approach has the potential not just for deciphering existing codes but could also be transformative for understanding ancient, undeciphered languages, thereby expanding our grasp of human history.