You clicked a link promising to teach you how to read a painting. You were excited. Maybe you imagined a thoughtful curator walking you through brushstrokes and symbolism. Instead, you got machine-generated drivel and a paywall six clicks in. And you’re furious. Good. You should be.
This isn’t just another bad website. It’s a perfect case study in how to alienate your audience, destroy trust, and make people hate what could have been a genuinely useful tool. The project — How to Read a Painting — uses AI-generated prose to explain classical art, then asks for money before you get anything meaningful. The comments tell the story: “the generated prose telling you how to appreciate art is ghoulish” and “I was pretty excited until I hit the paywall.”
Here’s the truth nobody wants to admit: the outrage isn’t really about AI. It’s about the cynical mismatch between what you were promised and what you got.
The promise was democratization — a data-viz interface that lets anyone “read” a painting like an expert. The reality was a soulless summary written by a machine, shoved behind a payment gate. The visual design? Interactive spreadsheets and charts. Because nothing says “art appreciation” like a pivot table.
Let’s be clear about what happened here. The creator built something that tried to quantify the unquantifiable. Reading a painting is not a data problem. It’s a human experience — a conversation between the viewer, the artist, and centuries of cultural context. You cannot reduce that to a formula and then charge admission for the privilege of reading a robot’s summary.
Every 200 words, you should drop a sentence that makes someone want to screenshot it and send it to a friend. This project dropped a paywall instead.
The backlash is a symptom of a larger sickness. Consumers are exhausted by AI-generated content masquerading as expertise. We’ve seen the same pattern in SEO spam, automated journalism, and now, art criticism. The difference here is the audacity: to take something as sacred as classical art — studied by humans for centuries — and serve it up as a machine-generated worksheet with a price tag.
I’ve spent years watching creators try to monetize everything. But there’s a difference between charging for genuine value and charging for a shortcut that insults the very thing it claims to celebrate. The comments on this project are not just angry. They’re hurt. People feel betrayed because they wanted this to be good. They wanted a fresh way to engage with art. Instead, they got a soulless algorithm and a reminder that every last corner of the internet is being colonized by automation and greed.
So what’s the takeaway? If you’re building something that relies on human expertise — art, music, writing, history — don’t replace the human. Use AI as a tool, not a substitute. And for the love of everything, don’t put a paywall between your audience and the thing you’re trying to teach them. The moment you do, you’re no longer a creator. You’re a toll booth operator on a road nobody asked for.
This project failed not because it used AI, but because it forgot that the value of art criticism lies in the critic’s soul, not their spreadsheet.
We need to stop pretending that automation can replace the messy, imperfect, beautiful work of human interpretation. The next time someone pitches you a “data-driven” approach to culture, ask yourself: would I pay for this if it came from a human? If the answer is no, don’t build it. Because your audience will figure it out — and they will remember.
FAQ
Q: Isn't the real issue just that the AI writing was bad? Could a better AI fix this?
A: No. The core problem is not the quality of the AI, but the premise. Art appreciation is a human, subjective experience. Even the best AI cannot replicate the emotional connection, historical context, and personal insight that a human critic brings. Better AI would still feel hollow because it's a machine trying to explain something it cannot experience.
Q: What's the practical implication for creators using AI in their work?
A: Use AI as a tool to augment human expertise, not replace it. If you're creating content about a deeply human domain like art, history, or culture, your audience will sniff out automation. Always lead with a human voice, and if you must charge for access, make sure the value is clearly from the human insight, not the machine summary.
Q: But isn't there value in making art criticism more accessible? Couldn't AI help with that?
A: Yes, accessibility is important, but the approach here was wrong. AI can help with transcription, translation, or summarizing existing human scholarship — but it should not be the primary source of interpretation. A better approach would be to curate and summarize human-written analyses, clearly attributing them, and offering the interactive interface as a gateway to deeper human work, not a replacement for it.