The mechanics sound playful, and they are. Behind them sits a more serious editorial idea about attention. In an age of endless scrolling, Basteri wants people to stop long enough to notice composition, colour and small details in works they may already think they know. AI is part of the process, but he describes it as a tool inside a directed artistic practice rather than a substitute for judgment. The original paintings remain the cultural starting point; René and Julie are the narrative devices that pull a contemporary audience back into them.

René, Julie and the daily game
René is easy to spot once you know him: small, green, unmistakably bespectacled. Julie is the harder prize. She is meant to be found, not announced, which turns a passive scroll into an active search. According to Basteri’s account of the project, the fastest participant to locate her selects the following day’s painting from two options. The format creates a light social contract with the audience. Looking carefully becomes the entry ticket, and choosing tomorrow’s image becomes a small act of curation shared with everyone else who follows along.
That loop also explains why the project sits awkwardly — and interestingly — between categories. It is not a museum didactic panel, not a pure meme account, and not an automated image feed. It is closer to a serial game that happens to be set inside art history. Unjica has covered other projects that try to make looking itself the medium, from immersive art experiences to experiments where art and science share the same stage. VANFROG.ART belongs in that broader conversation, with a narrower wager: that a frog and a duckling can make people reopen paintings they had filed away as familiar.

Classroom, classical painting and directed AI
Basteri’s day job matters to the tone of the work. As a support teacher in Tuscany, he uses AI activities with students to underline a point that also structures VANFROG.ART: artificial intelligence depends on human instructions, critical thinking and responsible use. In his framing, the model does not replace curiosity; it is subordinated to it. The same language appears when he talks about the artworks. AI assists a process he directs. The result should still be readable as an authored intervention into a known image, not as an anonymous remix without a point of view.
Selected pieces make that authorship visible in different keys. In a work inspired by Edvard Munch’s Melancholy (1891), René answers solitude with an embrace, turning the scene into a conversation about sadness and comfort while Julie remains hidden. In a version after Pierre-Auguste Renoir’s A Girl with a Watering Can (1876), the frog enters a garden encounter that feels almost conversational. A piece after Henri Rousseau’s The Sleeping Gypsy (1897) lets René drift into a nocturnal landscape. Another, after Wassily Kandinsky’s Composition VI (1913), nearly dissolves him into abstract rhythm, as if the search had moved from figure to pure form.


Readers curious about how artists extend a painting beyond a unique canvas may also recognise a practical cousin in print and reproduction culture, including guides such as Unjica’s note on selling art as prints online. VANFROG.ART takes a different route — daily interaction rather than edition sales — but it shares the same problem of keeping a still image alive in a networked present.
From phone screen to book and public screens
The project did not stay only on Instagram. An illustrated bilingual book, titled in English Where Is the Shy Duckling Hiding? Find It in the Paintings, translates the search into a printed sequence of looking and discovery. The cover presents the game as something you can hold, not only swipe. That move matters editorially: it suggests Basteri is testing whether the same attentive habit can survive offline, where there is no algorithm refreshing the feed.





VANFROG.ART remains early. It began only in April 2026, and its public path is still opening. For Unjica’s readers, the interest is not that a frog entered the museum postcard. It is that Basteri’s mix of teaching, storytelling and carefully bounded AI treats classical painting as something that can still demand time — one search for Julie at a time.



