We can already see the huge impact that AIGC has brought to the design industry. With the gradual maturity of cultural and graphic tools such as Midjournal and Stable Diffusion, cultural and video tools like Sora have become popular, and “everyone can become a designer” and “AI still cannot understand the laws of the physical world” have become controversial hot topics in the design industry. In the design industry, architectural design undoubtedly has its unique characteristics: deep interaction with reality, complex collaboration chains, and long implementation cycles, which seem to delay the impact of AIGC compared to other fields and make it more difficult to achieve..
Efficiency improvement or replacement? Is there any visible industry trend in AI empowering architectural design in the future? On March 5th, InnoSalon: AI for Everything Phase 1- Generative AI+Architectural Design invited Dai Qiong, a first-class registered architect and deputy chief architect of the Design and Research Center of China Construction Technology Group South China Limited, Feng Dayuan, co-founder and data scientist of AIRI Lab, Huang Rong, a doctoral student in computational design at the Hong Kong University of Science and Technology (Guangzhou), and Liu Zidong, co-founder of Habitat Technology&CEO, Explore the application prospects and innovative possibilities of artificial intelligence in this industry with new and old friends from the industry..
AIGC, which focuses on the early stage of scheme creation and assists in architectural performance, has been widely applied in the industry.
Models that have undergone targeted training are/will gradually become more precise and controllable.
High quality datasets are crucial for AI to generate high-quality, stylized design solutions, and future data hegemony needs to be vigilant..
While humans evaluate and select AIGC achievements, AIGC also trains and influences design decisions and trends..
The development and application of artificial intelligence technology prompt architects to think about the origin of their professional roles.
In the foreseeable future, architects will continue to perform tasks that artificial intelligence cannot replace..
AI+Architecture Design: Practical Cases.
Looking back at the technological development of the architectural design industry, computational design methods have existed since the 20th century. The first large-scale labeled image dataset, ImageNet, was released in 2009. Generative adversarial networks emerged around 2014, and papers on diffusion models began to appear in 2018. Whether it is BIM technology that coordinates and integrates the entire design process, or various evaluation and simulation tools based on big data analysis, to the extent that cloud based cross regional and cross professional collaboration systems and continuously developing digital technologies and tools for architectural design are committed to making the entire architectural design process more intelligent and efficient. But since the release of ChatGPT, the current wave of artificial intelligence dominance has seen more applications based on architectural performance to help improve design efficiency and further bring about innovative projects that have an impact on the entire industry chain..
Dai Qiong introduced that as the “technology eldest son” of China Construction Group, China Construction Technology is promoting the full production chain development system of new construction industrialization – that is, high industrialization degree, high system integration, easy digital definition and modular intelligent manufacturing.
