Generative art uses a procedure to produce visual, sonic, or other creative outputs. The artist shapes the rules, parameters, and possible variations rather than specifying every final detail individually. Computer code is common, but generative art is broader than AI-generated imagery and does not inherently require a blockchain.
From an algorithm to an edition
In a generative NFT project, a script may combine an artwork's seed with rules for shapes, colors, or movement. Different seeds can create distinct outputs within a recognizable artistic system. Some projects generate an output when a token is minted; others select or create the work beforehand.
A deterministic script can reproduce an output from the same inputs and dependencies. That does not mean all generative works are static: animation, interactive inputs, or configurable parameters can change what is displayed.
What the token preserves
An NFT can record a work's identifier and ownership history, while the script, seed, libraries, and media may be stored onchain or elsewhere. Reproducibility depends on retaining the resources required to run or display the work.
Owning the token does not automatically transfer copyright. Artistic authorship, licensing, technical permanence, and market scarcity are separate questions. A large number of algorithmically different outputs alone establishes neither artistic quality nor lasting demand.