Abstract
The human creativity is now being supported, or challenged, by an easy access to generative tools utilising Artificial Intelligence (AI). Within the last few years, Generative Adversarial Network-based image generation tools have become available for the public. In this paper, we explore how one such tool, called Midjourney, could be used for generating moodboards. Our work focuses on exploring moodboard image generation via text-based prompts. We take note of the active online discussions, where examples and ethical concerns have strongly emerged during the late 2022. As a result, the AI produced imagery that the authors could not have produced themselves in any reasonable amount of time, if at all. The GAN-generated images are of high quality, and visually useable for design purposes. However, there are ethical concerns on the sources of the images used for training the AI systems, and copyright concerns on the outcomes.
Keywords
GAN, Generative design, Moodboard
DOI
https://doi.org/10.21606/nordes.2023.114
Citation
Mikkonen, J.(2023) Advent of GAN: How does a generative AI create a moodboard?, in Holmlid, S., Rodrigues, V., Westin, C., Krogh, P. G., Mäkelä, M., Svanaes, D., Wikberg-Nilsson, Å (eds.), Nordes 2023: This Space Intentionally Left Blank, 12-14 June, Linköping University, Norrköping, Sweden. https://doi.org/10.21606/nordes.2023.114
Conference Track
exploratorypapers
Advent of GAN: How does a generative AI create a moodboard?
The human creativity is now being supported, or challenged, by an easy access to generative tools utilising Artificial Intelligence (AI). Within the last few years, Generative Adversarial Network-based image generation tools have become available for the public. In this paper, we explore how one such tool, called Midjourney, could be used for generating moodboards. Our work focuses on exploring moodboard image generation via text-based prompts. We take note of the active online discussions, where examples and ethical concerns have strongly emerged during the late 2022. As a result, the AI produced imagery that the authors could not have produced themselves in any reasonable amount of time, if at all. The GAN-generated images are of high quality, and visually useable for design purposes. However, there are ethical concerns on the sources of the images used for training the AI systems, and copyright concerns on the outcomes.