Archinect's Lexicon focuses on newly invented (or adopted) vocabulary within the architectural community. For this installment, we're featuring a term central to current conversations on how architects can use artificial intelligence in the design process.

"Generative design" is a technology-driven design process that employs algorithms and/or artificial intelligence (AI) to develop optimized design solutions. It represents an iterative design approach inspired by natural evolutionary processes, leveraging computing power to explore a large number of possible permutations of a solution, quickly and in real-time.

In the context of computer-aided design (CAD), a designer will input into the generative design software specific parameters such as material types, manufacturing methods, budget constraints, and performance criteria. The software then explores all possible permutations of a solution, testing and learning from each iteration what works and what doesn't.

Generative design can be used to design everything from products to buildings to systems. It has the potential to create incredibly efficient, sustainable, and innovative designs that may not have been conceived by a human mind, as it can explore many more design alternatives than a human could in the same amount of time.

This approach has gained prominence in fields like industrial design, architecture, and engineering because it not only increases efficiency but also allows for the development of designs that would otherwise be impossible or highly impractical to conceive by humans.

There are numerous examples of how architects have used generative design in their work. Here are a few:

1. The Sagrada Familia, Barcelona: While Antoni Gaudi didn't have access to today's generative design software when he began designing the Sagrada Familia in 1882, his methods were remarkably similar to the generative design process. He created physical models of his designs and then used strings and weights to study the forces in the structures, a process which mirrors the structural load analyses performed by today's generative design algorithms.
2. The Broad Museum, Los Angeles: The "veil-and-vault" concept for this museum was designed with generative design algorithms. The "vault," or storage area, and the "veil," a porous exterior structure that lets in diffused light, were designed with the help of algorithms to achieve the desired aesthetic and functional effects.
3. Beijing National Stadium (Bird's Nest) for 2008 Olympics: The tangled "nest" structure of the stadium was designed with generative algorithms that allowed the architects to create a random, natural aesthetic while still ensuring that the structure was sound.
4. The Autostadt Roof and Service Pavilion by GRAFT Architects: This structure was designed with algorithms that helped create a complex, curved structure that would have been extremely difficult to design without the help of generative design.
5. The Gherkin (30 St Mary Axe), London: This distinctive building designed by Norman Foster used generative design principles to optimize the building's shape for energy efficiency and structural performance.

This article is part of the Archinect In-Depth: Artificial Intelligence series.