The Work of Design in the Age of Artificial Reproduction
Posted
Melissa Baley - A.I. recreation of "The Sttop", a fictional exhibition branding and marketing material from a previous course at UBalt.
Lohaine Gonçlaves - A.I recreation of Brazilian natural foods store logo from personal projects.
Jacob Hicks - A.I. recreation of "Safe Space", an LGBTQ-focused installation from a previous course at UBalt.
Joe Magar - A.I. recreation of typographic exploration inspired by illuminated manuscripts, tower of babel imagery, and bayes theorum, from a previous course at UBalt.
| Educator/s: | Jacob DeGeal |
|---|---|
| Institution: | University of Baltimore |
| Level: | Advanced, Graduate, Junior, Senior, Undergraduate |
| Duration: | 1 week |
| Category: | Design Methods, Graphic Design, Product Design, Typography, Visual Communication |
| Filed Under: | AI, BA Program, Brainstorming, Collaboration, Color Theory, Community College, Composition, Design Futures (Designer of 2025), Digital, Ethics, Experimental, Four-year Program, Human-Computer Interaction, Illustration, Interdisciplinary, Iteration, Multidisciplinary, Online Learning, Photography, Process, Technology, Two-year Program, Writing |
| Bookmark Project | |
Project Brief
Using ChatGPT and DALL-E, recreate one of your past designs.
- Choose a single designed artifact that you created in the past.
- Using your free DALL-E credits, try to recreate your artifact as close as possible to the original.
- You will need to refine your query several times to fix issues or communicate clearly to the A.I. interface.
- Download your final version, and document your final written prompt.
Learning Objectives
- Design solutions using fabrication and software as a means for creative generative and intervention.
- Contemplate the ethics, benefits, and threats of machine learning as a tool for generative design.
- Reflect on and deconstruct your own work and process by translating into written prompts.
- Analyze contemporary and emerging technologies through the lens of historical technological developments.
Deliverables
- one generated AI image that reproduces as close as possible an orginal design artifact, .jpg format
- the original design artifact, .jpg, .png or .pdf format
- the final written prompt in DALL-E, screenshot or .txt
Readings/Resources
- Big Data, Big Design 'Chapter 1: Peek Inside the Black Box' - Helen Armstrong
- 'The Work of Art in the Age of Mechanical Reproduction' - Walter Benjamin
- 'Maillardet\'s Automaton' - The Franklin Institute
- 'The San Francisco Ballet Is Catching Heat for Promoting Its ‘Nutcracker’ Performances With A.I.-Generated Art' - Artnet News
Reflections
This project yielded a spectrum of results. In some cases, the final artifacts were wildly different, in others, DALL-E was able to recreate the imagery with surprising accuracy. Through reflection and discussion, this led to insights into imagery and descriptions that are abundantly sampled by DALL-E, and which are more human/cultural concepts and styles that are harder to explain in words. Another interesting constraint to this project is that students were restricted in the number of iterations they could provide to DALL-E, since it operated on a limited credit-based system (10 free credits per month). This constraint is similar to the physical limitations a designer might encounter in analog methods —such as film cameras—and forced the students into a decision-making process unique to digital production. Some students provided feedback that they would have preferred to continue this topic with more original new pieces, rather than only recreations of past work. However, some students did continue using ChaptGPT and DALL-E as brainstorming tools in future projects.
