Abstract

Generative AI, in particular large language models such as ChatGPT, have enabled broad application of AI in everyday life. While text-generating AI approaches are already widely used by the general public, the application of audio-visual generative AI is lagging behind. Audio-visual generative AI seems particularly suitable for knowledge sharing purposes, i.e., to provide serious content. Thus, this paper evaluates state-of-the-art audio-visual generative AI approaches, i.e., text-to-video, and their suitability for communicating serious content effectively. In a small proof-of-concept, several self-proclaimed AI-based video generation tools have been tested. Based on a sample video, we also assessed the uncanniness of AI-generated videos, to clarify their applicability for serious content, i.e., training and education, as we assume this might influence the perception of the content.

Download

Share

COinS