| Time: | March 24, 2026 |
|---|---|
| Download as iCal: |
|
In the IntCDC Constructive Conversations, the Cluster of Excellence “Integrative Computational Design and Construction for transformative Architecture” hosts renowned national and international scientists to speak on current topics and challenges in the field of IntCDC.
We are delighted to welcome Prof. Athanassios Economou from Georgia Institute of Technology to our upcoming Constructive Conversations event on Tuesday, 24 March 2026 from 1:00 – 2:00 p.m.
Athanassios Economou will give a lecture on Shape Machine and Generative AI.
Figure caption: An automated production of Durand's Plate 22 from his Précis of the lectures in Architecture in Shape Machine. (Application of porticoes one interaxis wide with rooms three and five interaxes wide)
Figure caption: An automated description of kerfing woodworking in Shape Machine. (left) Shape rules (1) to (5) generate the underlying grid and shape rules (6) to (8) replace the grid with the kerfing meandering motif; (right) Boustrophedon production of the kerfing pattern in 42 steps.
Abstract
Text-to-image, image-to-image, and more recently, text-to-vector Machine Learning (ML) models have been increasingly pervading contemporary architectural practice and pedagogy. And yet, the Find and Replace operations, fundamental for any ML model —and essential in Word, Excel, Photoshop and any computational system and digital technology that relies on atomic symbols (alphabets, numbers, pixels, voxels)–– have yet to be implemented successfully in Computer-Aided Design (CAD). How would we ever expect to take advantage of the current unprecedented data driven technologies when CAD systems are not properly set up to do so? It is claimed here that there is another way to look at this conundrum and it is within reach: Shape Machine for Rhino 8, a new software developed at the Shape Computation Lab at Georgia Institute of Technology, features evanescent symbols and vector-based shape-Find and shape-Rewrite operations along with a logical processing framework to allow a completely new way to interact with CAD models and AI data-driven technologies. Several applications across architecture and computer science are discussed to showcase the potential impact of this new technology and its relations to the current state of AI in architecture.
More information: shapemachine.design.gatech.edu
