Architecture typically distinguishes between completed structures and conceptual designs. The physical limitations of real buildings contrast sharply with the boundless creative possibilities in architectural imagination, often called “paper architecture.” Design technology has historically supported this creativity, and recent developments in artificial intelligence have further spurred such theoretical exploration.
The “Transductions: Artificial Intelligence in Architectural Experimentation” exhibition at Pratt Institute in Brooklyn showcased works from over 30 architects investigating how AI can open new avenues in architecture. The exhibition, curated by architects Jason Vigneri-Beane, Olivia Vien, Stephen Slaughter, and Hart Marlow, aimed to capture architecture’s early experimentation with AI rather than exhibit projects ready for immediate construction. It reflected dynamic interactions among various architectural mediums and techniques, from imagery and animation to mixed-reality and fabrication.
Since the 1963 inception of Sketchpad, a pioneering architectural software, technology has continually pushed architectural boundaries. Programs like Revit, SketchUp, and BIM have since revolutionized architectural design, energy tracking, and sustainable construction, among other advancements.
The architects in “Transductions” view AI as an innovative tool, not a threat to their profession. Vigneri-Beane acknowledges AI’s challenges but finds value in its creative rhetoric. While AI doesn’t automatically produce meaningful results, it enhances architectural vocabulary and visual acuity, demanding significant effort to yield worthwhile outcomes.
Vien regards AI as a potent instrument complementing an architect’s toolkit, enriching the historical media used to generate and represent ideas. Despite fears about AI, architects like Andrew Kudless and Martin Summers recognize its unique capabilities in conceptual exploration. They explore AI’s potential to generate iterations or high-resolution sketches, blending realistic and abstract representations.
Ultimately, these practitioners see AI technologies as partners, enriching rather than replacing traditional architectural methods.