Generative Art
Generative art is artwork produced by an algorithm that combines rules, randomness and pre-made assets to create many unique outputs. In crypto, it usually refers to NFT collections, often profile picture (PFP) collections, where each token is a distinct generated image.
A creator designs layers of traits, such as background, body, clothing and accessories, and assigns each trait a probability. A script then combines these layers to produce many different images. In some projects, code runs at mint time instead and creates an output from a random seed. Each result is linked to a token through its metadata, which records the image location and traits, while a smart contract handles minting and ownership.
Generative collections are among the most common kinds of NFT. Many holders use their token as a profile picture on social media, which is where the PFP name comes from. Because trait frequencies are known, collectors compare how rare each item is, and marketplaces often display rarity information alongside a collection's floor price.
CryptoPunks is an early, widely cited example. It is a set of pixel-art characters generated by software, with traits such as hats, glasses and pipes appearing in different proportions across the collection.
A common misconception is that rarity sets value. A rare trait does not ensure demand, and prices can fall sharply or disappear. Buying an NFT also does not automatically grant copyright to the artwork. If the image is hosted off-chain on a server that goes offline, the artwork can stop loading even though the token itself still exists on the blockchain.