19.06.24, 16:00 Uhr
Online • bidt – Bayerisches Forschungsinstitut für Digitale Transformation, Department Informatik der CIT School TU München
For many decades, the use of automation in the media was restricted to the production of the vehicles—such as the newspapers—that carried the content, rather than the content itself. That started to change in the 21st Century and has accelerated since 2022 with generative AI models, such as ChatGPT. This lecture will define media content automation and briefly describe its history—from early imaginings in the 1960s to early applications in the 2010s. Three drivers of its deployment will be proposed—demand, digitisation, and data—and justified. Contemporary examples will be described within a framework that distinguishes between the older data- and rules-based systems and the more recent generative AI models. Finally, a summary of research on such systems in the news context will be provided, focussing on audiences’ evaluations of their output.