Generative AI is increasingly being applied to various domains, which has subsequently raised concerns on how generative models might affect the productivity of artists. In this study, it is argued that generative models can improve the productivity and inspiration of artists and perhaps assist in successfully overcoming difficulties common in the artistic domain, such as writer’s block or creative blocks. Generative models such as Stable Diffusion can be used to generate images that could serve as inspiration while conversational systems such as ChatGPT can be used to generate feedback or enhance inspiration by generating text prompts for multi-modal diffusion models. More importantly, this paper introduces a newly created surrealism dataset which was curated specifically for this research. A Denoising Diffusion Implicit Model was trained on the dataset to assess its quality and effectiveness. The lack of specific datasets that cover certain areas of art styles, such as surrealism, is discussed, including how the newly curated dataset might be beneficial for artists and AI enthusiasts as a possible benchmark for generative models in the domain of surrealism art. Finally, we hope that the contribution of such a dataset will have a positive impact, especially in sourcing inspiration for artistic generative modelling.