Research

Revisit Flatwriter

Semi-open System for
Architecture Design

University of Kansas
2024

In collaboration with:
Chien-Shuo Pai, MArch
Assistant Professor
Graduate Institute of Architecture
National Yang Ming Chiao Tung University (Taiwan)

The Shape Machine represents the most promising solution to the anticipated bottleneck in AI over the next decade. Unlike traditional AI systems that rely on symbolic representations—numbers, text, and abstract symbols—Shape Machine directly manipulates shapes, effectively bypassing the need to translate visual elements into symbols. This capability allows Shape Machine to function as a true visual programming system, which is crucial for overcoming the limitations of symbolic representations. By working directly with shapes, Shape Machine enables AI to interact with and generate complex visual designs more accurately and intuitively. This direct manipulation of visual elements streamlines the design process, eliminating the loss of detail and fidelity often associated with symbolic translation. As a result, Shape Machine enhances the precision and creativity of AI-driven design, making it a game-changer in fields such as architecture and beyond. Its ability to handle visual data natively ensures that AI can better mirror human design processes, leading to more sophisticated and innovative outcomes. In essence, Shape Machine is set to revolutionize AI applications by addressing current limitations and paving the way for more advanced and effective visual computing systems.