Tech companies are now turning to centuries-old books to train artificial intelligence systems, with Harvard University releasing nearly one million books dating back to the 15th century for AI researchers. This move aims to provide a data source that is less controversial than copyrighted material and offers a wealth of cultural and historical information missing from modern online content. The dataset, Institutional Books 1.0, contains over 394 million scanned pages, offering a diverse linguistic and historical perspective for AI developers. While this initiative presents opportunities for AI advancement, it also raises concerns about outdated or harmful content and the responsible use of AI technology.
Source: https://www.wsbtv.com/news/ai-chatbots-need/A6IBX3GDNFHDDOYN6K6U3DPQ5M