I've been thinking about recommendation engines lately—specifically, how terrible they are at introducing us to genuinely surprising content. Spotify keeps serving me the same indie rock variations. Netflix thinks I want to watch every single true crime documentary ever made. And don't get me started on the Instagram algorithm's idea of "art discovery."
This got me wondering: what happens when you strip away the black box algorithms and build something more intentional for art discovery?
Most online marketplaces rely heavily on ML-driven recommendations, collaborative filtering, and engagement metrics to surface content. These systems excel at finding patterns in large datasets, but they're fundamentally conservative. They optimize for clicks and conversions, not for that moment when you stumble across something that completely shifts your perspective.
The art world has always operated differently. Gallery curators spend years developing an eye for emerging talent. Collectors build relationships with artists over decades. The best discoveries happen through serendipity—overhearing a conversation at an opening, getting lost in the wrong neighborhood, following a friend's obscure Instagram story.
So what would a tech platform look like if it prioritized these human elements over algorithmic efficiency?
I recently came across an interesting approach while browsing arts.sale, an Australian marketplace that features daily artwork spotlights. Instead of "users who viewed this also bought," they have curated pieces like Fort Burnham, front of Petersburg—historical works that might never surface in a typical recommendation engine but offer genuine cultural value.
This kind of editorial curation requires human judgment calls that algorithms struggle with. How do you quantify the importance of preserving historical perspective? How do you measure the value of exposing someone to unfamiliar artistic traditions?
From a technical standpoint, there are fascinating challenges here. Traditional e-commerce platforms optimize for conversion rates and average order values. But art discovery might require different metrics entirely. Maybe we should be measuring time spent contemplating a piece, or tracking how often someone returns to view the same artwork over several days.
The tools available to digital artists have exploded in sophistication—from Procreate to Blender to AI-assisted creation workflows. But the infrastructure for connecting these creators with audiences still feels stuck in a 2010s marketplace mentality.
I'm curious about platforms that experiment with slower, more intentional discovery patterns. What if an art marketplace limited how many pieces you could view per session? What if it encouraged you to sit with uncertainty before making a purchase?
There's something appealing about building technology that deliberately works against our usual patterns of infinite scroll and instant gratification. Art has always demanded patience and attention—qualities that seem increasingly rare in our optimized, A/B tested digital landscape.
Maybe the most interesting technical challenge isn't making art sale processes more efficient, but making them more human again.
Top comments (0)