3DALL-E: Integrating Text-to-Image AI in 3D Design Workflows

Abstract: Text-to-image AI systems are capable of generating novel images for inspiration, but their applications for 3D design workflows and how designers can build 3D models using AI-provided inspiration is less understood. To investigate this, we integrated DALL-E, GPT-3, and CLIP within a CAD software in 3DALL-E, a plugin that allows users to construct text and image prompts based on what they are modelling. In a study with 13 designers, we found that designers saw great potential to incorporate 3DALL-E into their workflows and to use text-to-image AI for reference images, renders, materials, and design considerations. Additionally, we elaborate on prompting patterns and provide measures of prompt complexity observed across participants. We conclude on a discussion of how 3DALL-E can merge with existing generative design workflows and propose prompt bibliographies as a form of human-AI design history.