The world is being quietly rearranged by people who write very long documents.


The title they went with Toward a Universal Color Naming System: A Clustering-Based Approach using Multisource Data Noisy translates that to

Researchers build the first standardized color naming system across industries


A team collected 19,555 color names from 20 different sources and used clustering to identify 280 distinct perceptual color categories that humans actually use, replacing thousands of overlapping shades. This means design tools, clothing retailers, and search engines can now label colors the same way instead of inventing their own names for what is essentially the same shade.
Right now, the same color gets called 'coral' in one design tool, 'salmon' in another, and 'peach' in a third. This creates real friction: fashion retailers can't easily search their own inventory by color, generative AI systems produce inconsistent results, and anyone building a visual search tool has to solve the naming problem from scratch. A shared standard means these systems can talk to each other. It also means the person using the tool gets consistent results instead of discovering that 'mauve' and 'heather' are different names for the same RGB value.
Watch whether major design platforms (Figma, Adobe) or fashion/e-commerce companies (Shopify, ASOS) actually adopt this 280-color standard in their tools within the next 18 months.

If you insist
Read the original →