Sebastian Nehrdich, Associate Specialist at University of California Berkeley’s Electrical Engineering & Computer Sciences (EECS), unveiled ways to improve translation fluency and accuracy with artificial intelligence (AI) in a “Workshop on AI Prompting” on April 23 ,2024 for the Fo Guang Dictionary of Buddhism (FGDB) Translation Project. Hosted by the FGS Institute of Humanistic Buddhism (FGSIHB), nearly 40 translators and data collectors from the USA, Singapore, Hong Kong, mainland China and Canada participated online.

The training session emphasized AI command usage for efficient translation. Nehrdich, an AI specialist, introduced simple prompts that would make a huge difference in translation efficiency and quality through popular AI tools. Using Claude as an example, he demonstrated that with simple prompts, translators can utilize AI for sentence analyses and word-for-word explanations. This would enhance translators' comprehension of ambiguous sentences. Nehrdich also explained how AI algorithms, closely linked to databases, can be prompted to explain and translate complex Buddhist text sentences accurately. In response, Venerable Miao Guang suggested retranslating AI translations into the source language to ensure semantic integrity.

In addition to AI mechanisms and translation efficiency, Nehrdich’s introduction  on the functions of Dharmamitra Translator (MITRA), a free innovative Buddhist AI translation tool, also stressed the importance of using multiple AI tools to enhance accuracy and translators’ understanding of the source text. As he translated the same sentence with both Claude and MITRA, he highlighted MITRA’s feature of translating ancient Buddhist texts from and into languages including Sanskrit, Chinese, and Tibetan. With this function, MITRA becomes a valuable tool for translators, helping them understand and translate difficult sentences as coherent wholes. Referencing on AI tools’ analyses and translations, translators would be equipped to deliver highly accurate translations.

At the end of the workshop, Nehrdich used car manufacturing as a metaphor to highlight the collaborative relationship between AI and human translators, indicating that even the best performing cars (AI tools) require skilled drivers (human). He looked forward to future workshops on AI tools with the FGDB project to further advance translation practices.