"Illustration of the subconscious mind": Reinterpreting Dream Material as Artistic Creative Workflows Supported by Generative AI
Sijia Liu, Yanru Qian, Fan Zhang, R. Allen, Ray Lc
Creativity & Cognition July 13, 2026 DOI: 10.1145/3803784.3807563 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Qualitative study Peer reviewed |
|---|---|
| Sample size | 6 |
| Population | Six artists participating in a four-week online residency |
| Intervention | Generative AI |
| Duration | Four-week online residency with weekly sessions |
| Key findings | Artists adapted their use of generative AI across creative stages: during early preparation they embraced unexpected outputs as inspiration, while during production they demanded accuracy and coherence. The authors propose that as artistic creation moves from abstract to concrete, GenAI shifts from inspirational reinterpretation to generative support. |
Abstract
Dreams are sources of inspiration for artists tapping into their subconscious states, but the uncertainty of fragmented elements and difficulty of recall make it challenging to harness dream-based inspiration for creative work. Generative AI shares a similar ambiguity with dreams, producing unexpected outputs that offer reinterpretation in the creative process. It suggests GenAI may help engage with such materials. To explore how artists work with dream content using GenAI , we held a four-week online residency with six artists including weekly sessions to process dream materials, followed by artistic creation for exhibition. Artists adapted their use of GenAI across stages: during early preparation, unexpected outputs were embraced as inspiration; during production, accuracy and coherence from GenAI were demanded. As artistic creation moves from abstract to concrete, GenAI shifts from inspirational reinterpretation to generative support. This work highlights differences in preparation and creative production when tapping into dreams for creativity.