GPT Astra: how neural networks will change the architect’s workflow
How neural networks are changing the workflow in architecture. Using the example of a competitive project for a technology park in Kronstadt, we show the entire process — from site analysis, searching for massing and facade solutions to creating IFC models in LOD100–300. We separately analyze unsuccessful iterations, the distribution of tasks between the architect and the neural network, as well as the actual costs in terms of man‑hours and machine time.
