Haytham Nawar | Generative Pictographic Language

Описание: Dr Haytham Nawar is an Associate Professor of design and the former Chair of the Department of the Arts at the American University in Cairo. Nawar is also the founding director of Cairotronica, Cairo International Electronic and New Media Arts Festival, Egypt.
​
[email protected]
​

Generative Pictographic Language

Writing systems serve as instruments of language and culture, representing the diverse civilizations throughout history. While some scripts fade away over time, new ones emerge through cultural interactions and the influence of existing scripts. By blending different scripts, we envision a future where old scripts are synthesized. This vision is realized through a machine that utilizes its neutral connections to create a new script stroke by stroke. By recognizing the similarities between input scripts, the machine generates a pictographic language that reflects the unconscious resemblances across these writing systems. Through multiple iterations, the final result becomes a testament to the hidden links among various civilizations.

The objective of the AI-generated script is to introduce a pictographic language that offers insights into the visual expression of ancient cultures. It proposes a modern pictographic script, with its traits directly drawn from the data processed by the machine. By establishing consistent patterns in shapes and strokes, the machine leverages its neural connections to create a new pictographic script on a digital screen, stroke by stroke. Through the recognition and grouping of similar patterns, the machine identifies the shared characteristics of these scripts, such as stroke style, figure complexity, and proportions. This enables the machine to generate a unique pictographic language that reflects the homogeneity of each writing system when combined. Writing systems have played a crucial role in shaping civilizations, and the final result serves as an exploration of the often overlooked connections between diverse cultures across many civilizations. The project involves training various models to generate a script based on the similarities found in the input data. Running the machine on GPUs for several days at a time and iterating the process allows for the refinement of a visually appealing result.

This project delves into the captivating history of human communication, examining ancient scripts and writing systems. From ancient carvings to complex ideograms, writing has been vital for language documentation and cultural preservation. Leveraging machine learning, the project creates a generative pictographic language by drawing inspiration from a diverse database of vector-based writing systems. By exploring the visual language of ancient civilizations, the project aims to uncover universal iconography and highlight the intrinsic link between writing and culture. The project's outcomes include 3D objects, surface engravings reminiscent of ancient tablets, an animated film, and a printed book. These mediums showcase the collaboration between the artist and the machine, opening up new possibilities for creative partnerships between humans and machines.