generativeart

Generative art refers to artworks whose production is mediated by autonomous systems, such as computer algorithms, mathematical rules, or artificial intelligence. In a generative process, the artist defines a system or set of rules, constraints, and parameters, while the system generates one or more outputs that may be fixed or continually evolving. The artist's role is to design the process, select inputs, and determine how outputs are realized and presented.

The field emerged in the mid-20th century with computer and algorithmic art experiments by pioneers such as Vera Molnar, Georg Nees, Frieder Nake, and Michael Noll. In the 1960s and 1970s artists like Harold Cohen with AARON and later software artists expanded the approach, exploring generative rules, randomness, and systems that produce unique or editioned works.

Common techniques include procedural generation, stochastic processes, cellular automata, fractal algorithms, L-systems, and more recently machine learning models such as neural networks, GANs, and diffusion models. Tools include programming environments such as Processing, p5.js, and Python libraries; interactive installations are common.

Generative art challenges traditional authorship and originality, because a work's appearance emerges from a system rather than a single act of creation. Many generative pieces are produced as editions or sequences; some are evolving installations or live performances. The rise of AI-based models has expanded capabilities but also raised questions about copyright and creativity.

It remains a recognized practice in digital and new media art, used in galleries, museums, and online, and intersects with design, architecture, and game development.