Translation Dies in the Tongue
Finalist (Graduate Student)
Artist(s):
· Carla Erausquin Bayona
School Information:
Syracuse University
Graduate Student
Artwork Information:
Medium: 3d print
Category: AI Artwork and Sculpture
Description:

Translation Dies in the Tongue is a sculptural project that uses AI image-to-3D systems as speculative translation devices. The work begins with handmade collages and drawings inspired by Moche ceramics, which are then processed through generative AI models that reinterpret the images into unstable three-dimensional forms. These digital outputs are further refined, transformed, and translated into sculptural objects.

Rather than treating AI as a neutral tool for reproduction, the project focuses on mistranslation, distortion, and visual hallucination as generative conditions. Each iteration moves further away from an identifiable origin, producing new forms that simultaneously preserve and erode cultural references. The resulting sculptures exist between artifact, reconstruction, and synthetic memory.

The project relates to “Speculative Futures” by examining how machine perception may reshape cultural memory and material history in increasingly AI-mediated futures. It questions what happens when systems trained on contemporary datasets attempt to reconstruct fragmented historical imagery, and how meaning shifts as images move between dimensions, technologies, and contexts. Through this process, the work proposes a future where translation itself becomes unstable, generative, and materially embodied.

Technical Information:

The project combines handmade collage, digital modeling, and AI-based image-to-3D generation workflows. Original drawings and collaged compositions are processed through generative AI systems, including image-to-3D translation models, to produce speculative sculptural forms. The resulting meshes are refined using Blender through processes such as remeshing, texture mapping, and digital sculpting. Some forms are prepared for full-color 3D printing, while others remain as digital renderings and animations. The workflow intentionally preserves artifacts, distortions, and reconstruction errors generated by the AI systems as part of the conceptual framework of the work.

Additional Images:
© © © ©