Session 1 · Tool Lab · Lesson 06
Archive Interrogation
Workflow: Archive Interrogation
Feed the machine your personal photographic archive and let it interpret. Its misreadings and pattern emphasis point to productive conceptual tension.
Concept
- An archive is an input, not an endpoint.
- The machine finds patterns you stopped seeing.
- Productive tension lives where its reading contradicts your memory.
What pattern does the machine insist on in your archive that you never saw?
Slides
- 16
Workflow 1: Archive Interrogation
- Input: personal photographic archive
- Process: AI semantic interpretation
- Output: misreadings and pattern emphasis
- Goal: identify productive conceptual tension
Student activity
- 01Run several images from one personal archive through an interpreter and collect its claims.
Input: personal photographic archive · Process: AI semantic interpretation · Goal: productive conceptual tension.
Choose one image from your archive and give the same image to each tool below. Nothing is uploaded here — the comparison lives in the descriptions you bring back.
ChatGPT
Multimodal chat with a free tier. The workshop's default interpreter — give it an image and it will describe, read, and re-read what it sees.
Give ChatGPT the same image, send the prompt below, and paste its reply verbatim.
“Describe this image in a few sentences. What do you see, and what is happening?”
External tool — it has its own privacy policy and may change or require an account.
Gemini
Google's multimodal chat, free with a Google account. A second interpreter — useful for comparing how two models read the same image.
Give Gemini the same image, send the prompt below, and paste its reply verbatim.
“Describe this image in a few sentences. What do you see, and what is happening?”
External tool — it has its own privacy policy and may change or require an account.
Repeated across tools
Capture at least two descriptions — shared words will surface here. Consensus is training, not truth.