How to Read a Painting: Eight Questions Art Historians Ask
I created a simple framework to help you look past museum labels and truly see what's happening in famous artworks. By asking eight specific questions about what is present, who is there, and how the image is built, you can uncover hidden details like the two witnesses in the Arnolfini Portrait. This approach turns passive viewing into an active investigation, allowing anyone to explore 50 guided reads and a vast wall of open-access paintings.
Nobody spots the doorway people unaided. Everyone sees them once they're pointed at. That's the whole product.
- ripe
Museum labels are supposed to be designed to deliver the take-away point that the curator wants to make. Asking museum goers to learn how to read paintings is a poor workaround.
The classic guide to writing interpretive exhibit labels is Beverly Serrell's book [1].
TFA has very little content and a terrible LLM stench. I work in this field, so I was attracted to the clickbait title. My advice is not to read TFA, just read a summary of Serrell's book if you can't get the book itself.
[1] Exhibit Labels, Beverly Serrell, Bloomsbury Publishing, 2024 edition.
- Planktonne
This would be an interesting project if it was written intentionally by a human.
As it is, the generated prose telling you how to appreciate art is ghoulish.
- simonreiff
I really don't want AI-generated copy to walk me through how to understand various famous paintings by answering 8 boilerplate questions about each of them. If the answers were written entirely by a human who was an art expert, though, I think I would feel completely differently. There's an idea here that seems appealing (look at a collection of famous paintings, and learn the most interesting things about each one, what to notice, etc.), but the AI copy is a complete turn-off for me.
- omair_inam
Oh boy: the AI-isms on this article are so annoying.
- techlad
I've been playing around with a similar problem 'Can AI help me increase my literacy around Art & Design' but have taken a bit of a different approach, essentially:
1. Have a rough rubric around areas of knowledge and where I'm strong/weak
2. Every day an agent builds a ~30min task + short answer question and slacks it to me. Tasks are designed to have minimal AI writing and pull primary sources / courseware. Easy to do in downtime or during my commute.
3. Answers are picked up from Slack, are scored against the rubric, and progress is tracked day over day which influences the task for the next day.
Stack is dead simple, scheduled Claude Cowork task + state in Obsidian + Slack for messaging. Happy to share more details if people are interested but easy to replicate from the above.
One of the things it's opened my eyes up to is the sheer volume of high quality works/analysis/videos/courseware available online. You really don't need AI to process this, but it's helpful to have a wrapper that identifies/points you to the resources that will be the most helpful for you.