The emergence of artificial intelligence (AI) in graphic design and the adoption of AI tools such as Midjourney and Dall-E 3 have revolutionized the field, allowing complex designs to be created quickly. Given these challenges, the effective and timely use of AI-generated graphics is critical to the industry.
It was not the first time that AI played a role in the art field. AI was employed in graphic design to automate repetitive tasks such as image manipulation, typography, and layout (Mustafa, 2023, p. 244) in the last decade (e.g. using AI-based layout tools that can analyze data such as user preferences and browsing history to suggest layouts that are most likely to appeal to users). The advent of deep learning and neural networks has enabled AI to analyze and interpret vast quantities of data in a manner that can be analyzed and interpreted, and design work according to the requisite specifications – for example, their ability to visualize textual descriptions and combine different concepts into unique images. AI-based design tools have also made it possible for designers to generate designs based on data and user preferences (Kim & Ahn, 2005). This, by a large amount, improves the personalization and relevance of the designs (Liu & Elgammal, 2019).
There are multiple images generating AI in the market, among them, DALL – E 2 (build-in application of ChatGPT 4 and 4o) and Midjourney are the two that catch the most attention. DALL-E 2 and Midjourney rely on generative models that have been trained on a vast repository of text and images (Barozai, 2024). DALL – E 2 employs a variety of techniques, including converters (for text processing) and diffusion models (for image generation. Midjourney utilizes a similar approach to DALL – E 2, combining transformers and diffusion models).
Moreover, Midjourney specializes in transforming textual descriptions into visually compelling visuals that enhance storytelling (Byrne, 2023). This tool offers new methods for engaging audiences by creating images that accurately reflect the visual context and emotional depth of the text. While DALL-E 3 expands the scope of creative possibilities by enabling the combination of disparate elements and concepts into novel images. (Marcus et al.)
However, despite the convenience AI-generated art brings, there is a potential for AI-generated designs to become excessively similar and generic, resulting in the homogenization of design. This, in turn, may lead to a reduction in uniqueness and individuality (Meng, 2012). Another difficulty AI scientists are trying to solve is the consistency of the images. The term “consistency” here stands for the ability to keep the same element for a series of illustrations. For example, if a CG artist is trying to draw a set of illustrations of one specific character in multiple different scenes, the consistency of the character’s setup is hard to keep for AI. The character in the first picture may have long black hair, whereas the same person in the second picture might still have long black hair but with a different eye color.
Moreover, Since AI-generated images lie in the grey zone of copyrights, many artists protest against the use of AI in the industry, for the database relies on capturing various art pieces from various artists without their permission.
Although it is still controversial whether AI art could be commercialized, still, by no doubt that AI changes the industry. I believe the essence of art does not completely lie in its technique, but in the thoughts and ideals the artists gained throughout the process. The use of the tool mentioned above opens a window for those who are not able to express their artistic thoughts due to a lack of painting and drawing techniques. Despite the future of AI art still lies in uncertain, people won’t stop the process of exploring and expressing art via different mediums. It is always the human thoughts that shine.
Reference
Meng, X. (2012). The impact of artificial intelligence on graphic design: A survey of designers’ attitudes. Design Research Quarterly, 4(2), 11-22.
Mustafa, B. (2023, March). The Impact of Artificial Intelligence on the Graphic Design Industry. RES MILITARIS, 13(3), 243–255. https://resmilitaris.net/index.php/resmilitaris/article/view/3333
Kim, H. J., & Ahn, S. H. (2005). Applying artificial intelligence to graphic design education. Journal of Digital Design, 5(1), 57-64.
Barozai, D, K. (2024, April 23). DALL-E vs MidJourney: Making the Right Choice for AI Image Generation. Folio3. https://www.folio3.ai/blog/dall-e-vs-midjourney/
Liu, B., & Elgammal, A. (2019). Creative AI: A review of AI systems that generate creative outputs. ACM Computing Surveys,52(5), 1-37.
Byrne, U. (2023). A parochial comment on Midjourney. International Journal of Architectural Computing, 21(2), 374-379. doi: 10.1177/14780771231170271.