top of page

AI-human Collaboration:
How Algorithm Understand English Ballads

Timeline

Week 1-6: background research+data analysis

Week 7-10: design

Week 11-13: report +exhibition

Role

data analytist,

lead designer,

with 1 visual designer,

1 programmer, and 1 writer

Methodology

Design with data,

Research through Design

Project Overview

With a digital imaged-based archive of English ballads, the team worked on the data analysis and designed a meaningful digital experience of it. Specifically, we leveraged the dataset to explore the potential of human-AI collaboration. We created a Style Gan algorism that could auto-analyze the context of the text-based ballads and co-create painting with human audiences.

Screen Shot 2023-01-10 at 14.24.36.png
ballads.webp

The artifact is exhibited at the Innovation Center, University of Edinburgh

Screen Shot 2023-01-09 at 20.07.04.png

AI-human co-created artpiece #01

8ec738_c978dc2d00a046a28025877f60df21dc_mv2.jpeg

AI-human co-created artpieces

Reflection

We have constructed a transmedia visual artwork based on the human-AI collaboration to present English ballad data to the audience brand-newly. From the words written by the lyricist to their visualization and through the work with deep learning algorithms, ballad illusion provides a new platform for people outside the culture with an experimental visual presentation. The audience's understanding of the changes in the colors and elements in the artwork stimulated empathy for the stories and emotions behind the ballads. It developed a deep interest in the cultural background. This novel narrative provides a pioneering reference for the data-based cultural heritage.

 

In the human-AI collaboration, we found that the right to interpret our visual works is unique to the designers and tends to be a cultural democracy constructed by multiple parties, including algorithms and audiences with different cultural backgrounds. This nonlinear creative process also liberated the designers, allowing us to get much enlightenment about AI art and transmedia practice.

Transmedia has brought improvements in the accessibility and attractiveness of the metadata with Fuzziness in the storytelling. The Fuzziness, challenging respect for cultural connotations carried by data, and appealing explanations are not contradictory. The transmedia retains, exhibits, and continues to enhance outsiders' understanding of the emotion expressed by the ballads. It uses vivid and affectionate ways to resonate with the audience and facilitate participatory culture through democratization. The decisive outcome from AI-generated artwork, mysterious and appealing dynamic effects, and the changing content seem to annotate the ballads and arouse the audience's empathy outside the culture. It makes people wonder to what extent empathy can transcend the original cultural background of the audience.

 

As some AI artists claimed, the relationship between AI and human-being in art has changed from "AI-generated" to an "AI-enhanced" form, so the co-creation between humans and AI has emerged. Based on our creation, we have a new understanding of the role of ai in artistic creation and cultural narrative. Instead of a tool for artificial intention, AI is treated as a seasoned artist under our supervision with subjective consciousness and art expertise during cooperation and creation. This benefit stems from the collaboration with AI and communicability, which is quite enlightening and not entirely controllable.

 

Compared with traditional artwork completed independently by humans, the co-creation of AI and humans has produced a subversive change in the creative process with more exploratory value to the outcome. Furthermore, compared with some previous AI-generated works regarding AI as a tool, which is criticized as "not to be identified as art," this kind of in-depth collaboration even gives a new intention of artistic creation. It brings the designer and the audience together into an unknown world.

bottom of page