© Ulises (@ulises.studio)
Artificial intelligence is becoming an undeniable presence in our daily lives. It teaches, generates content, and disrupts the fragile boundaries—both visual and imaginative—that once governed our interactions on social media. On platforms like Instagram, we witness a flood of imagery where every kind of speculative exercise is freely shared, recalibrating our understanding of the relationship between architecture and image. Amid this transformation, entire professions find themselves on uncertain ground, as AI begins to challenge areas once defined by human expertise.
Yet beneath this apparent abundance lies the opaque core of closed-source AI: an algorithmic black box that systematically conceals the origins of the data it consumes. As a result, its outputs are inevitably prone to factual distortions, anachronisms, and subtle or overt biases. This same machinery can hollow out the significance behind the languages and stylistic signatures of canonical architects—manifest, for instance, in AI-generated visions speculating how famed designers, living or dead, might have reimagined the Eiffel Tower. We shared one such image to observe and better understand how people—especially architects—respond to AI's current possibilities and limitations, and the ways it mimics architectural intent. The response was quite fascinating, revealing a mix of curiosity, concern, and critical reflection.
Typological datasets are rooted in expansive, collective archives—pools of visual and spatial knowledge assembled from countless contributions. They produce derivative images of what already exists, never fully capturing the deeper layers behind an architectural project. After all, architecture is more than just an image. Algorithmic synthesis, therefore, yields results without clear authorship, flattening the depth and intention carefully developed over time within a design language.
