Abstract
Professors encounter AI being used in daily academic practice with university students studying creative media, digital animation, and game art design. When evaluating the viability of those AI services, users must be able to explain AI’s recommendations in a clear and understandable manner. Insisting upon well-defined attributes of transparency helps users understand why certain decisions were made, fostering trust in the system. This paper offers specific methods for validating transparency during the on-going use of Ais when creating prototypes and storyboards. Heavy usage of AI services will erode the reliability of the cloud service. This study discusses reliability as measured by such factors such as precision, latency, throughput, and recall rates. In our practice-based studio work, we have adapted several Ais to changing user needs and preferences over time. This paper documents the temporal effects of user data through samples gathered in both lab settings and production workflows.
Presenters
Bernard GoldbachLecturer, Department of Digital Arts and Media, Technological University of the Shannon, Tipperary, Ireland
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
2024 Special Focus—Images and Imaginaries of Artificial Intelligence
KEYWORDS
Transparency, Reliability, Counterfactualisation, Generative Pre-trained Transformer
Digital Media
This presenter hasn’t added media.
Request media and follow this presentation.