The 5 Laws of LLM-Assisted Design
These laws establish guidelines for incorporating LLMs (Large Language Models - AI systems that can understand design concepts and assist in creative processes) into the creative design process while upholding principles of artistic integrity, user-centered design, and ethical considerations. They promote AI as an innovative design partner while preserving the essential human elements of creativity, empathy, and design thinking.
Designers are free to choose any large language model suited to their creative process. This flexibility allows leveraging the unique strengths of different LLMs, fostering innovation, and adapting to the diverse needs of design challenges.
All design concepts or outputs generated with LLM assistance must be thoroughly understood and contextualized by the designer. Designers must validate and adapt these outputs to ensure relevance, creativity, and alignment with project goals. Blindly adopting generated outputs without understanding their implications is strictly prohibited.
The refinement of design outputs must involve human oversight, blending AI-generated ideas with the designer's expertise and creativity. While LLMs can propose innovative concepts, the ultimate design decisions, adjustments, and artistic coherence rest with the human designer.
Designers and teams must continuously evaluate and refine their workflows involving LLMs. Feedback loops, case studies, and sharing best practices should drive the improvement of how LLMs are integrated into the design process, ensuring their effective and responsible use.
LLM-assisted designs must adhere to ethical principles and promote inclusivity. Designers must critically evaluate outputs to avoid perpetuating biases or stereotypes and ensure that designs generated with LLM assistance align with accessibility and sustainability goals.
