SD PLATFORM BEIJING 2023
AI Design Lab :
Exploring the Intesection of AI, Coding, and Architecture
with YuanJie, Beijing, China
30 Jul – 6 Aug 2023
Tutor:
Soomeen Hahm
Hanjun Kim
AI Design Lab :
Exploring the Intesection of AI, Coding, and Architecture
AI drawing, known as text-to-image generation, has emerged as a captivating phenomenon within the realm of artificial intelligence, capturing widespread attention and stimulating discourse across various domains. Its potential extends beyond conventional artistic practices, transcending boundaries and enabling the creation of visually captivating and distinctive imagery. This technology has also fostered collaborative opportunities between humans and AI, blurring the conventional lines of authorship and creative agency.
As an architectural designers, participants have the opportunity to harness the power of AI drawing to enhance the creative process and explore new design horizons. ChatGPT, an advanced AI language model, suggests exciting ways to leverage this technology: from visualizing designs to iterative exploration, concept development, collaborative processes, design optimization, and even igniting inspiration and creativity.
In light of these advancements, this intensive 2-day course is designed to introduce participants to the vast potential of artificial intelligence in architectural design. Throughout the course, attendees will delve into the world of AI-generated prompts and gain proficiency in leveraging multiple AI models. It is essential to approach this technology as a tool that enriches and amplifies the design process, rather than relying solely on its outputs. By embracing the potential of AI while balancing it with expertise, intuition, and critical thinking, participants will be empowered to redefine the boundaries of architectural design and harness AI as a transformative force in their creative endeavors.
Objectives
– Provide an overview of AI models and their relevance in architectural design.
– Introduce GPT models and demonstrate their application in generating design prompts.
– Enable participants to create their own design models using Grasshopper and advanced C# scripting techniques.
– Explore the capabilities and implementation of controlNet, Stable Diffusion, and flocking algorithms in architectural design
– Foster collaborative design projects and encourage participants to explore real-world applications of AI and advanced algorithms.
– Apply the learned techniques to hands-on design exercises and explore the potential of AI-driven design.
Software
– GPT + GH(C#) + ControlNet + Stable Diffusion
– Depth Map (2.5D) + GH
– GH(C#) + ControlNet + Stable Diffusion
Methodology
– Create a height map in Midjourney/Stable Diffusion
– Read height map in Grasshopper -> Generate terrain
– Flocking C# in GH using GPT
– Rendering/Change view using ControlNet/Stable Diffusion
Participants
Liu Jiayi, Wang Haotian, Zheng Zhixuan, Yang Yang, Li Jiaying
Ouyang Airong, Ye Xinyu, Zheng Yuqiao, Liu Xingyu, Wang Xinyu, Shi Xinqi, Zhu Yanrong