AI Model Values Unknown Street Artist Over Picasso, Revealing Market Biases

A study spearheaded by Magnus Resch, an art economics lecturer at Yale, alongside a data scientist and an AI specialist from Silicon Valley, created a multimodal AI model to explore if artistic value could be assessed without considering market context. This model, which analyzed millions of images, including renowned masterpieces and high auction results, aimed to forecast auction prices based on visual traits and metadata. Notably, it appraised a Picasso at less than $1,000, while attributing a seven-figure value to a piece by an unidentified street artist captured in New York. The model yielded market-relevant outcomes only when artist identities and gallery ties were included, indicating that reputation trumps visual merit. Resch contends that AI should enhance transparency in the art market, allowing collectors to find works that resonate with their tastes, independent of traditional barriers. He is working on an AI-driven app to promote this transparency.