Archival Images of AI

Skills
AI
Strategy
Research
Cross-functional Leadership
Communication
notion image

Archival Images of AI

Strategy · Communication · Research · Cross-functional Leadership
Published 2024

Role

Project Lead, Author

Team

Researchers: Ploipailin Flynn; Nadia Piet; Dominika Cupkova
Imagemakers: Hanna Barakat, Dominika Cupkova, Nadia Piet, Zeina Saleem, Cristobal Ascencio, Eryk Salvaggio Advisors: Tania Duarte, We and AI; Rasa Bocyte,Netherlands Institute for Sound & Vision
Editor: Sofia Vieira
Proofreader: Heather Lane
Graphic Design: Kaashvi Kothari
Publisher: AIxDESIGN
 

Overview

Stock images of AI suck. They're filled with visual tropes – glowing brains, humanoid robots, walls of code – that reinforce myths about AI’s power and purpose.
So, we wondered – how can we make it easier to create better images of AI?
Archival Images of AI explores how existing images – specifically public domain images from digital heritage collections – can be remixed and reused to create new, better images of AI.

Approach

  1. EXPERT INTERVIEWS (8)
  • 3x cultural heritage professionals exploring contemporary use and access to cultural archives: Cees Martens, Isabel Beirigo, Monique Groot
 
INSIGHT 1
Digital heritage collections are a rich resource for image-makers – at least those who know about them. However, most image-makers don’t know about them, how to access, or use these collections to make their own images.
THINGS WE COULD MAKE
  • Helping people find images / Curated Collections: Collections of heritage images curated by AIxD community that would be helpful in making better images of AI
  • Helping people find images / Search Queries : Tips + tricks for searching collections to find images that might inspire imagemakers
 
INSIGHT 2
If people want to make a better image of AI, they usually coming to it with an idea of the image they need to make (a researcher or journalist who needs an image about their research topic). In order to make tools that’re helpful and practical for image-makers, this research team needs our own image briefs to experiment with.
THINGS WE COULD MAKE
  • Better Images of AI / The images we need: Image-making briefs for us to use in this project
 
INSIGHT 3
People may recognise a remixed image, but they aren’t familiar with the different ways in which an image can be remixed. We document remix techniques as shared by remix artists, why they employ these techniques, and offer examples of those techniques in use to create better images of AI.
THINGS WE COULD MAKE
  • Remix 101 – Documentation of remix techniques from remix artists and AIxD community members, each showcasing the technique in practice with better images of AI we'll make in this project
 
  1. RESEARCH-BY-MAKING
notion image
Over 3 months we made images of AI using public domain images found in digital heritage collections. Guided by the insights from our expert interviews, each image maker made dozens of images using a remix technique of their choosing, shared their creations in group Show + Tells, taught each other their tips and tricks, and continued on making. Read the play-by-play in our project diary here ↗︎
We recorded our challenges and insights along the way. These lessons shaped the outline of the AIoAI Playbook.
 
  1. PARTICIPATORY SENSE-MAKING SESSIONS [ ONLINE ]
  • July 2024 – How to find interesting things in the archives with Europeana
  • Dec 2024 – Playtest of Playbook [ PUBLIC ]
 
  1. PUBLISH
We collected all of our most helpful tips and tricks we learned along the way and published them in the Archival Images of AI Playbook. We published our most favorite images from the archives and the scraps we made from them. See the full list of outputs below:

Outcomes

  • A collection of our favorite public domain images images from these archives can be found here ↗︎
  • A collection of the scraps we found ourselves coming back to can be found here ↗︎
    • “scraps” are what we called the bits we “cut out” from the public domain images found in the archives