Creating Images with Generative AI: An Imaginative Aid

Increasingly, both public and professional creators are being assisted by generative artificial intelligence (AI) systems in the production of images. Concerns have been expressed about the potential for these technologies to decrease human creative agency or aesthetic diversity. To examine whether these concerns are warranted, the psychological and philosophical literature on creativity and imagination is examined. Drawing on this, two senses are distinguished in which generative AI systems can be used as imaginative aids: (1) to aid a user in visualizing an idea; and (2) to aid a user in cognitive play. The latter, unlike the former, is central to creativity and it is also rarer in the use of these systems. The case is made that to facilitate this kind of use more widely and ameliorate the aforementioned worries, the development of these systems ought to focus on not only technical improvements, such as greater control over elements like training data, but also attitudinal changes, so that users do not suffer from illusions of creatorship that may inhibit the development of their aesthetic aims and autonomy.