How big is the internet?
Powers of Ten ^10
…recreated using the vastness of internet imagery as frames
A collaboration between Kelli Anderson & Adam Pickard
- SIDE A: Kelli Anderson's version uses google search result images as frames
- SIDE B: Adam Pickard's version uses DALL•E to recreate the animation
- Includes a zine (printed by Kelli in Brooklyn) explaining each of our respective processes
- Includes "magic" powers of 10 countdown packaging
- 180 highly-flippable 4/4 pages, printed in Spain
$30
Both sides of this flipbook recreate the Eameses’ iconic 1977 film, Powers of Ten, using only found-imagery from the internet.
Kelli Anderson took photos from Google Images' 'visually similar' search function and used them as frames:
Adam Pickard used prompts in DALL•E to recreate the film using their enormous dataset:
Process Notes
Kelli Anderson (SIDE A)
In 2015, I created this small remake of Powers of 10 by stitching still images together by hand, frame-by-frame.
All of the animation frames are found images, “two people on a picnic blanket”, “hand on chest”, “earth from space”, etc. plucked from their original contexts (where they were not congregating to form the Powers of 10.)
I used the inpainting feature of DALL·E (a text-to-image AI tool) to see if it could remake Eames Powers of Ten. My process was to give it a text prompt, then shrink the image it output, leaving the surrounding area transparent. I would then upload that to DALL-E and fill the transparent area with the next prompt. I repeated this step 57 times (for the 57 prompts listed here) and then combined all of the steps in After Effects. I left the prompts in the final video to show the language that DALL•E interpretted for each of the 57 steps.