FOUNDATIONS · 9 MIN READ
What Is Generative Art?
Generative art is made by setting a system in motion with enough autonomy to help determine the finished work. The artist creates the rules, materials and boundaries; the system performs them. That system might be a computer program, a set of written instructions, a mechanical process or even a living organism.
The important idea is not that a machine replaces the artist. It is that the artist designs a space of possibilities. Individual works emerge from that space, sometimes including results the artist could not have predicted in detail.
IN THIS GUIDE
- The artist authors a system rather than manually composing every final output.
- Randomness is optional; rules, feedback and data can also generate variation.
- Generative art predates computers, although code greatly expanded its possibilities.
- Blockchain-based generative art can connect the act of collecting with the creation of a specific output.
01
The system is part of the artwork
In a conventional digital workflow, software is usually a tool used to execute a composition the artist has already decided upon. In generative art, the behavior of the software can itself be the medium. The artist decides which forms are possible, how they relate, what can vary and what must stay constant.
A simple system might place circles according to a few random values. A sophisticated one might simulate particles, growth, erosion, language, music or an artificial ecosystem. Complexity is not what makes the work generative. What matters is that the system participates meaningfully in determining the result.
02
Rules, chance and control
Generative artists constantly negotiate control. Too much control can make every output feel predetermined; too little can produce noise without artistic identity. The work often lies in shaping constraints until many different results still feel as though they belong to the same artistic world.
Randomness is one way to introduce variation, but it is not the definition of generative art. A deterministic cellular automaton, a drawing driven by weather data and a rule-based wall instruction can all be generative without using random numbers.
- Rules describe what the system may do.
- Parameters define values that can change.
- A seed can make a random-looking result repeatable.
- Constraints preserve the artist’s visual and conceptual intent.
- Curation determines which systems or outputs are ultimately presented.
03
Generative, algorithmic, computer and digital art
These terms overlap, but they are not synonyms. Digital art is the broadest technological category: it includes work drawn, photographed, modeled or edited with digital tools. Computer art is a historical and practical category for art made with computers. Algorithmic art is organized through explicit procedures or calculations. Generative art is defined by the system’s active role in determining the result.
Much contemporary generative art is algorithmic and computer-based, but generative practice is not limited to code. A written instruction performed by people, a mechanical drawing machine or a biological growth process can also be generative. The distinction is useful because it locates the artistic decision in the design and activation of a process.
04
A short history
Artists used generative thinking long before modern computers. Musical dice games, geometric tiling, weaving systems and instruction-based art all separate the design of a process from its execution. The word’s present art-historical meaning developed alongside systems, conceptual and computer art rather than appearing suddenly with NFTs.
In the 1960s, artists such as Vera Molnár, Georg Nees, Frieder Nake and Manfred Mohr used early computers and plotters to explore rules, variation and machine execution. Nees presented one of the first public exhibitions of computer-generated graphics in Stuttgart in 1965 and later used Generative Computergraphik as the title of his 1969 dissertation. Molnár began making combinatorial work before gaining computer access in 1968, demonstrating that the generative idea precedes any particular device.
Later tools made creative coding accessible to more artists. Languages and libraries such as Processing, p5.js, openFrameworks and JavaScript allowed systems to become interactive, animated and networked. Blockchain added another possibility: a collector could trigger a deterministic system at mint and receive one unique output from a larger work.
05
Generative art is not the same as generative AI
The terms are often confused. Generative AI usually refers to models trained on datasets that produce new media in response to prompts. Generative art is much broader and much older. A generative artwork may use AI, but it may instead use geometry, simulation, recursion, probability or hand-written rules.
In code-based generative art, the artist commonly writes or directs the algorithm and can explain the relationship between its rules and its outputs. That authorship model differs from prompting a general-purpose model, even though both processes can produce unexpected images.
06
Where 256ART fits
256ART focuses on code-based algorithmic and long-form generative art. An artist supplies an artwork script and a trait definition. At mint, the collection contract assigns or accepts a seed, derives the token’s traits and preserves the information needed to reconstruct the output.
For modern 256ART releases, the script, traits and browser-renderable live artwork are available from blockchain data. This makes the generative system—not merely a screenshot of one result—the durable object a collector owns.