A couple of printed books of un/learn ai laying on a table.
Someone holding the book un/learn ai open on a table.
Homepage of the digital publication un/learn ai.
Web view of an article inside un/learn ai.

un/learn ai 1

Date:

01.12.2024

Category:

Publication

Together with:

Paul Eßer
Lars Hembacher
Julia-Jasmin Bold
Maika Dieterich
Jean Böhm

un/learn ai - Approaching AI in Aesthetic Practices is the first of three publications documenting the findings and research of the research project KITeGG - Making AI tangible and comprehensible: Connecting technology and society through design. This first publication is divided into three sections, each focusing on a different aspect of the project.

  • Green: Project foundations, lab infrastructure and team introductions
  • Purple: Teaching approaches and student projects
  • Orange: Research findings and symposium contributions

The research project provides students with knowledge and an in-house infrastructure to work on and with AI algorithms, rather than just consuming AI services. This allows students to experiment and engage with AI in a creative, experimental and criti

Concept and Implementation

Staying true to the experimental approach, Paul Eßer and Lars Hembacher created the publication as a web-to-print publication using LLMs to create semantic links between articles by breaking the article into smaller chunks and linking them together. This feature is also reflected in the design decisions. The cross-linking allows the reader to read the publication in a non-linear way. In addition, AI-generated fonts from Paul's Type Table project were used on the cover and chapter dividers.

Based on the book design and web-to-print pipeline, I designed the web version of the publication, carefully translating the design decisions made into the web. Together with Maika Dieterich and Jean Böhm we implemented the web version of the publication.