One man’s meat is another man’s poison. As the Latin maxim de gustibus non est disputandum goes, there is no right or wrong in the matter of taste, and no one’s preference is better than another’s.
Human taste is often considered too intractable to be subject to quantification or algorithm, hence off-limits to Technology, Big Data or AI.
But evaluation is also at the heart of the new economy. When you look for an Uber, you are given the recommended driver and his rating. After you complete the journey, you will in turn be invited to rate his service.
Airbnb hosts and guests evaluate each other, too. A guest who has received consistently bad ratings from landlords may have difficulty finding future accommodations.
Hugo Liu, a taste researcher and principal scientist of eBay, has employed AI methods to discover hidden patterns in people’s choice of food, music, fashion, spouses and even baby names. His startup platform ArtAdvisor combines machine learning and taste-based AI to arrive at concise and actionable profilings of artists.
Even the multibillion-dollar art market, which used to be under the sway of cognoscenti, is tiptoeing around the door of Big Tech. Some advocate for the establishment of a central registry of art objects enabled by blockchain technology so that provenance, ownership, condition and transaction history can be rendered accurate and transparent for all players on the market.
In January, Sotheby’s bought Thread Genius, a pattern recognition startup whose algorithm can identify similar objects based on digital images. Sotheby’s will develop it in conjunction with another acquisition, the Mei Moses index, a database of over 50,000 images of works previously sold at auctions, for making AI-based recommendations on art to its clients.
The Denmark-based mearto.com is another art evaluation website that uses extensive databases on auctions and high resolution imagery to provide online appraisal. A statement on its website speaks of the art world entering the age of artificial intelligence: ‘Nowadays, appraisals are nearly always the result of a partnership between man and machine.’
CUHK has a reputable fine art discipline and museum facilities and machine learning is one of its strategic strengths. It is not inconceivable that it will in time become a centre for the arbitration of taste.
This article was originally published in No. 539, Newsletter in Jun 2019.