Artificial intelligence (AI) is revolutionizing the study of ancient texts, enabling scholars to decipher and interpret manuscripts that were previously unreadable due to damage or degradation.By employing machine learning algorithms, researchers can reconstruct missing parts of texts, enhance faded writings, and even predict the original content of fragmented manuscripts.
One notable example is the use of AI to read the Herculaneum scrolls, ancient papyrus manuscripts that were carbonized during the eruption of Mount Vesuvius in 79 AD. Traditional methods risk further damaging these fragile artifacts, but AI-driven techniques, such as those developed by the Vesuvius Challenge, have shown promise in virtually unwrapping and deciphering the texts without physical intervention. By training neural networks on known letter patterns, researchers can identify and reconstruct the writings within these scrolls, potentially revealing new insights into Roman literature and history.
Enhancing and reconstructing faded inscriptions
AI has also been instrumental in enhancing faded inscriptions on ancient artifacts. For instance, the DeepScribe project utilizes machine learning to transcribe cuneiform tablets from Mesopotamia. These tablets, often thousands of years old, have suffered from erosion and damage over time. By training AI models on existing transcriptions, DeepScribe can predict and fill in missing or unclear characters, aiding scholars in understanding these early writings and the societies that produced them.
In addition to enhancing legibility, AI is being used to reconstruct fragmented texts. The Fragmentarium project, for example, employs machine learning to match fragments of medieval manuscripts scattered across various collections worldwide. By analyzing features such as handwriting, parchment color, and textual content, AI can suggest possible matches, allowing scholars to digitally piece together manuscripts that have been separated for centuries.
Recent advancements in AI have further expanded its applications in the field of ancient studies. In 2023, researchers at the University of Oxford developed an AI model capable of translating ancient Greek texts with high accuracy. By training on a vast corpus of texts, the model can assist scholars in translating and interpreting documents, expediting the study of ancient Greek literature and historical records.
Similarly, the Ithaca project, a collaboration between DeepMind and the University of Venice, has developed an AI system that can not only translate ancient Greek inscriptions but also date them and identify their geographical origins. This multifaceted approach provides historians with valuable context, aiding in the reconstruction of historical events and cultural exchanges in the ancient world.
Ethical considerations
While AI offers remarkable tools for unlocking ancient texts, it also presents challenges and ethical considerations. The accuracy of AI-generated reconstructions depends heavily on the quality and quantity of training data. In some cases, biases in the data can lead to incorrect interpretations, potentially misinforming historical understanding. Moreover, the use of AI in cultural heritage raises questions about the ownership and accessibility of digital reconstructions, particularly concerning artifacts held in museums far from their places of origin.
The integration of AI into the study of ancient texts is transforming the field of historical research. By enabling the decipherment of damaged manuscripts, enhancing faded inscriptions, and reconstructing fragmented texts, AI provides scholars with unprecedented access to the literary and historical records of ancient civilizations. As technology continues to advance, it is essential for researchers to address the accompanying challenges and ethical considerations, ensuring that the digital resurrection of ancient texts contributes positively to our understanding of human history.
References
Nature 637, 14-17 (2025) doi: https://doi.org/10.1038/d41586-024-04161-z
