Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Towards a Generative AI Design Dialogue

Aron E. Owen, Jonathan C. Roberts

TL;DR

A new design process is proposed where designers verbalise their thoughts during work, later converting these narratives into AI prompts, which helps AI generate accurate visuals and assists designers in refining their concepts, enhancing the overall design process.

Abstract

Traditional visualisation designers often start with sketches before implementation. With generative AI, these sketches can be turned into AI-generated visualisations using specific prompts. However, guiding AI to create compelling visuals can be challenging. We propose a new design process where designers verbalise their thoughts during work, later converting these narratives into AI prompts. This approach helps AI generate accurate visuals and assists designers in refining their concepts, enhancing the overall design process. Blending human creativity with AI capabilities enables rapid iteration, leading to higher quality and more innovative visualisations, making design more accessible and efficient.

Towards a Generative AI Design Dialogue

TL;DR

A new design process is proposed where designers verbalise their thoughts during work, later converting these narratives into AI prompts, which helps AI generate accurate visuals and assists designers in refining their concepts, enhancing the overall design process.

Abstract

Traditional visualisation designers often start with sketches before implementation. With generative AI, these sketches can be turned into AI-generated visualisations using specific prompts. However, guiding AI to create compelling visuals can be challenging. We propose a new design process where designers verbalise their thoughts during work, later converting these narratives into AI prompts. This approach helps AI generate accurate visuals and assists designers in refining their concepts, enhancing the overall design process. Blending human creativity with AI capabilities enables rapid iteration, leading to higher quality and more innovative visualisations, making design more accessible and efficient.
Paper Structure (4 sections, 1 figure)

This paper contains 4 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Design Dialogue Framework
  4. Discussion and Conclusion

Figures (1)

  • Figure 1: This image effectively demonstrates the DDF, an iterative visualisation methodology encompassing six steps: Define, Sketch, Describe, Engineer, Generate, and Evaluate. Embarking with defining the objective and sketching initial ideas, the process moves on to describing sketches in detail and engineering prompts for generative AI. The generated step delivers refined visualisations using bird population data and migration patterns in Pakistan. Ultimately, the evaluation step ensures the visualisation design goals, fostering iterative refinements and amplifying inspiration and creativity in the visualisation design process, are achieved.