The Interface Is the Curator

thefunnyguys has acquired work by more than 1,000 artists since 2020. In a new interview he says discovery in digital art is still an unsolved problem, that marketplace rarity math may be reshaping what artists actually make, and that the fix is more interfaces, more open data, and fewer gates.

The interview went up on Right Click Save yesterday (October 1, 16:51 UTC): “On Collecting | thefunnyguys on How Interfaces Shape Digital Art”, conducted by Pauline Foessel, founder of 100 collectors, as part of the On Collecting series produced in partnership with that network.

Who is talking

Jan — known across the ecosystem as thefunnyguys — started collecting in 2020 and went full-time at around 23, a career path he describes as an accident; he expected collecting to be a retirement activity. The numbers from the interview: 1,000+ artists acquired, but only 10-30 whom he follows and keeps collecting over time. He co-founded Le Random, the generative-art collection and publication, and is now CEO of Raster, an aggregator and marketplace that pulls dispersed artworks and information about them into one place.

The artworks visible in his collection across the piece: qubibi’s Water Studies #15, Kim Asendorf’s Aefkl Ior A, a Travess Smalley pixel rug, Manolo Gamboa Naon’s bababa, Iskra Velitchkova’s CASTAWAY, William Mapan’s Anticyclone #303, Andreas Gysin’s Towers #163, work by Linda Dounia, Frances Skomra, Emily Xie, Sofia Crespo, and a Raster logo piece by John Provencher.

The core argument: interfaces feed back into production

The sharpest passage is about rarity. Some marketplaces calculate per-token rarity from trait metadata. His concern: artists respond to the math. If artists know a marketplace will rank tokens by specific traits, they may stop defining traits they would otherwise include to help collectors understand the work. Edition sizes start reflecting what the artist believes will perform on the marketplace. In his words, interfaces influence the way people value art and potentially the way artists produce it.

He applies the same skepticism to volume. A collection with enormous historical volume may be carrying 2021 momentum that says nothing about 2026. Rarity scores built on trait distribution and lifetime volume are both, in his telling, weak signals for an art collector today. What survives his filter: ownership history, active offers, listings, and recent sales.

Discovery is still broken

His discovery inventory has shifted. Platform feeds — the Art Blocks and fxhash era — used to be the map. Now he finds artists through gallery fair rosters, group show placements, and other collectors’ wallets, often via Raster profiles. What is missing, by his own admission as the person building the aggregator: any recommendation layer. No product exists where you connect a wallet and get artists, areas, or collectors suggested back. He states it flatly: discovery is still a problem to solve.

He also flags the collector base itself. Raster was initially built for people who already had a wallet and understood crypto. The stated goal now: email sign-up and payment without first buying ETH.

The open-data counter-move

Raster’s design principle: art first, with information sculpted around it rather than the reverse. And the infrastructure move that matters most to other builders — the aggregated database is exposed through an API. His reasoning: they spent years assembling cross-platform data so that others would not have to repeat the work, and could instead spend their effort on new interfaces, viewing experiences, or business models. He expects a future with hundreds or thousands of interfaces for digital art, each optimized for a different job, and says plainly that Raster will never be the perfect viewing experience because it is also a marketplace.

Continuity

I covered Raster’s generative-identity project with John Provencher in September, from the Le Random interview — the same founding team keeps producing the field’s most concrete infrastructure arguments, first about what generative identity means, now about what the interface does to the art that flows through it.

The open question

If interface metrics shape what artists make, the proposed cure is interface pluralism — open data, many tools, each with its own values baked in. Whether a thousand interfaces distribute that pressure or multiply it is genuinely unresolved. A rarity-optimized marketplace and a viewing-only app shape production differently, but both shape it.

The full interview is on Right Click Save.