I Built a Tool to Help People Find Their Art Style—Here's What Surprised Me
Quick note on today’s post: I’m sharing the BTS of building the Art Match Tool because it taught me a lot about bias, taste, and AI. Back to artists and culture next week.
We’ve been launching and testing our new Art Match Tool - a way to find your art style for your space based on your preferences and photos. The first ad we ran for it got a comment: “Too lazy to even make a real quiz. This is made by AI.”
Fair. The fear is real. The last thing I want is a tool that flattens taste into clichés or spits out slop. I spent my prior career work in tech, and when I thought about how AI could support LOTA’s growth, the question was simple: could it actually hold nuance and help more people find their way into art?
LOTA has always been anti-gatekeeping—widening the canon, amplifying emerging artists, and making taste feel possible for more people. The tool had to live up to that.
The problem I kept seeing
Client conversations looped the same lines: “I don’t know what I like.” “I just needed to fill the wall.” “I’ll never understand art.” It isn’t indifference—it’s fatigue. The pressure to “get it right” leaves people stuck between blank walls and impulse buys. Taste becomes practice only when we’re given room to try, miss, and refine (I wrote about this in my last post).
Meanwhile AI is everywhere—speeding up decks and workflows—but almost no one is asking how it might support taste instead of replacing it. That was the opening.
As I started building the app (ChatGPT prompting and Replit for front-end and integration) it became obvious there was a lot of bias to unwind. None of this will shock you; it mirrors how society talks about design and aesthetics, mostly through a Western lens.
For example, if you mentioned heritage or influences from the East or the Global South, AI drifted toward a “bohemian/worldly” trope—eclectic, charms, crystals—the “Eastern” shorthand as seen by a Eurocentric eye. As if culture and minimalism can’t coexist. At LOTA, we exist to break that binary. You can hold references to intricate temples and woven tapestries and still love quiet, minimal design.
Budget isn’t an aesthetic—it’s context.
Another bias: when people selected lower budgets, the model assumed a more “accessible” style. But a budget isn’t an aesthetic—it’s context. It shouldn’t determine whether someone prefers playful boldness or subdued elegance. In the latest version I pulled budget out of the style logic altogether and tightened the structure so the model can’t shove people into the very boxes we’re trying to break. AI didn’t invent those assumptions; it mirrors the ones society holds close.
High-end interior designer, Kelly Wearstler, recently wrote about this fear in “Can A.I. Have Taste?”—that AI strips the humanity out of creative work. I agree with her counterpoint: with the right structure, AI can give back time and clarity for the most human parts. Left alone, the outputs are flat; with guardrails, they can be surprisingly specific.
What actually worked
The breakthrough wasn’t a fancy algorithm. It was rules that force nuance. Culture isn’t a shortcut. Budget isn’t taste. Photos are context, not verdicts—the result now names the gap between where your room is and where you want it to go, and offers a bridge. And the language has to be specific: no filler. Call out the walnut credenza, the curve of the chair, the palette you actually chose.
We also added the Style Profile. Your top match shows up first, but you can see how you stacked across the other styles—where it was close, where it wasn’t. If it doesn’t feel right, you can tell us and choose the runner-up that fits better. The point isn’t to put you in a box; it’s to start a conversation and give you language.
Since launch, 1,725 people have tried the tool; 35% finished, and over half of finishers (311) uploaded real photos of their rooms—optional, messy, real life, which is where the specificity comes from.
What’s next
I’m going to let the dust settle on this version, but here’s what I’m exploring next:
-
See it in your room. Upload a wall; artworks appear at the right scale and lighting.
-
Aspirational mode. Share your current and dream spaces and get a bridge plan.
-
Collection journeys. Not just a match—a path for layering pieces over time.
To be clear, this has never been about tech for tech’s sake. LOTA started to widen the canon and support the artists shaping culture. The tool is a new doorway—language and confidence for people who’ve felt shut out, and a wider audience for the artists I love.
AI doesn’t have taste or vision. You do. The work is building systems that reflect that back without flattening it.
Still learning, still coding. If you care about how tech shapes taste, culture, and the art we live with, start here with our Art Match Tool.
P.S. If you’re tired of scrolling Pinterest for hours and still not knowing what art actually fits your space—or your soul—I can help. I’ve spent years discovering emerging artists and learning how to match people to work that genuinely resonates. Whether you need a framework to discover your style or hands-on guidance curating specific pieces, there’s a way we can work together. Learn More Here