BATTERY CITY
We did not draw this city. We trained it. Then we made every piece of it ourselves.
The models gave us a forecast: satellite maps projected 30 years forward. Everything after that was craft. The forecast became 3D massing, generated in Grasshopper. The machine's own imagery became the skin, mapped onto those forms until they read as real. Then animation, then film. Nothing borrowed, nothing stock; each tool was cast for the part it plays best.
Then 2020 told everyone to shelter in place. So the city went to the jury instead: an AR and VR app that put each reviewer inside it. It was designed as an answer to a lockdown. It turned out to be a phenomenal way to present work to anyone, anywhere. The app first lived inside a Dalí.
Forecast a future, give it form, and hand people the keys.
research
en masse demystification of data and inner workings of ai tools, instrumentalizing architecture: bringing legibility & quantification through spatial perception. we will be using the current form of computer vision systems and the databases they create as our arsenal to understand how machines see our world, and critically utilize machine learning to decipher how they interpret what they see.
advanced predictive map development
supervised ML models, generative design
through the application of pre-trained ML models, predictive color coded satellite maps were generated. These were generated in less than 1% of the industry standard time.
cgi
3d modelling, texturing, mapping, rigging, animation
immersive storytelling
ar
