Marco Alfaro is an Architect and Urban Designer with an Urban Design MArch degree from the Urban Morphogenesis Lab at the Bartlett School of Architecture UCL. Currently, He works as Urban Designer at Skidmore, Owings & Merrill (SOM). His work involves research and development of design methodologies, incorporating parametric tools, Artificial Intelligence and Biology into the design.
According to Marco Alfaro, creating symbiotic relationships between the elements involved in a project has been a priority, as well as data analysis, interpretation, and reconfiguration for him. He practices Architecture and Urban Design as the two disciplines are constantly evolving, and so is their relationship.
He is researching Artificial Intelligence, Hydroponics and exploring design within the Metaverse.
Using generative AI as a design tool has been an interesting journey for Marco. It always starts with a vision: Different scenarios present themselves in our daily lives, forcing us to question how we create architecture. Optimized structures in nature, the behavior of light, the characteristics of a material, and all the opportunities that each one consequently creates for design. Some concepts prove more successful than others, and ultimately, the beauty of using A.I. lies in its versatility and speed in the design process.
Since their inception, diffusion models have been more akin to an art than a science. As in architecture, an idea certainly morphs throughout its development, as expressing a concept in natural language does not match the precision of physically or digitally shaping a project.
According to Marco, when developing a concept, control is paramount. It’s crucial to comprehend how the AI processes information, ensuring the results remain consistent. This understanding allows the design process to be quite exploratory while maintaining a certain level of control over the output.
The process usually starts by testing a concept with a short prompt (or an image), which can then be altered with modifiers and parameters until, after several iterations, the desired result is achieved. In Midjourney, several techniques can be involved, such as remixing, image weight average prompting, multi-upscale compositing, etc.
Post-processing still plays an important role, as the generated images are prone to error and can always be improved.
The advent of these tools will undeniably influence the discipline. As architects become comfortable using AI tools like Midjourney, they’ll likely find their design process more fluid and dynamic said Marco. The speed and versatility of AI allow for rapid iteration and exploration of design concepts, leading to innovative and daring architectural designs that push the boundaries of what’s possible in the field. These tools also democratize the design process, making high-level design capabilities accessible to a wider range of people. Overall, integrating AI into architecture and design is an exciting development that promises to usher in a new era of creativity and innovation.
*The text was provided by Marco Alfaro and reviewed by PA Editorial Team.
If you’re curious about AI tools and their possibilities for architectural design, check out PAACADEMY’s AI workshops.