Limna and the Shifting Epistemologies of Art-Market Value

www.evolvingcities.org Limna and the Shifting Epistemologies of Art-Market Value Dr. Leander Gussmann, Art x Science School for Transformation (JKU, Linz, Austria). Context and Research Question The Limna app for smartphones, marketed as a “pocket art advisor,” marks a shift from connoisseurial, expert-driven valuation toward data-centric validation. Traditional pricing rested on opaque, socially negotiated “private codes”. Limna instead draws on Artfacts’ knowledge graph of roughly 800,000 artists and more than 1 million exhibitions to produce primary-market price bands. This research asks: How does this hybrid epistemology, which claims objectivity while being trained on entrenched hierarchies, redefine value in today’s art market? I examine whether algorithmic valuations transform epistemic authority or repackage existing power and prestige systems. I Say: Attention as Currency Tensions: Black Box Vs. Reality Algorithmic valuation tools demand critical data literacy. Users must interrogate proxy variables (e.g., exhibition counts as quality proxies) and recognise that rankings derive from curated social networks. These systems are performative—price bands can shape market expectations and outcomes. Future work should move beyond interface analysis to ethnographic fieldwork: interviews with developers, collectors and galleries could reveal whether these “probability engines” empower new buyers or entrench “attention elites.” Comparative studies across platforms and micro-cases of individual artists would illuminate how algorithmic tools redistribute epistemic authority in art markets. Method: The “Sociological Scanner” I reconstruct Limna’s datapoint pipeline through public documentation and interface analysis. Artfacts manually curates and verifies exhibition data before assigning “attention points”. Limna then inputs these scores, plus artwork dimensions and sales history, into a machine-learning model. This process is a form of commensuration, turning heterogeneous events into comparable numbers. Empirical work shows that social metadata, comprised of institutional ties and prior sales, explains about 73 % of price variance, while visual features account for only 5.5 %. Accordingly, Limna behaves less like an assessor of intrinsic quality and more like a scanner of attention capital. Interface Appraisal Figure 5: Original image at 100%, enlarged 200% and 400%. They Say: The Promise of Accuracy Limna’s marketing promises transparency and reduced risk for newcomers. Its FAQ states that under 20 % of price estimates diverge from gallery quotes. Promotional material cites coverage of hundreds of thousands of artists, over a million exhibitions, and sixteen thousand galleries. Founders frame the app as clarifying the “basic mechanisms” of art pricing and democratising market knowledge, implying that visualised career trajectories can challenge the opacity of private pricing. Interface Discover Contact Information Leander Gussmann Art x Science – School for Transformation JKU Linz Austria Email: [email protected] A gap exists between platform rhetoric and performance. During Art Basel 2021, journalists reported that Limna valued a Conny Maier painting at €4 500, whereas the gallery price was more than three times that. Such errors probably happen less frequently now. Such discrepancies, nevertheless, highlight that models trained on historical data have limits; they cannot capture hype or sudden market shifts. The system’s emphasis on established institutions risks reproducing geographical and gender biases baked into historical data. Even though Limna categorises artists (e.g., "fresh", “emerging,” “established”), its underlying hierarchy still favours canonical venues. Data Literacy & Future Research Call to action Drawing on Franck’s economy of attention, I contend that Limna quantifies curatorial investment. Museums, biennials and galleries lend reputation; their exhibitions generate data points that feed an “artist factor” and produce price bands. The app thus functions as a judgment device (Karpik, 2010) and participates in metric governance: rather than discovering intrinsic value, it normalises specific forms of visibility. Artists are incentivised to maximise measurable institutional recognition—through momentum and cultural recognition scores—at the expense of less quantifiable experimentation, thereby reorienting careers toward ranking and network centrality. Treat platforms like Limna not as neutral price engines but as performative judgment devices: question their proxies, ask who and what their datasets leave out, and insist on transparent metrics and accountable use whenever algorithmic valuations inform collecting, curating, or policy.