Annual Jasia Reichardt Lecture 2026

William Latham traces his practice through the idea of alternative evolution, connecting early drawings, computational sculpture, scientific visualisation and artificial intelligence. Introduced by Bronac Ferran, he describes how evolutionary trees transform geometric primitives through rules, initially using his imagination to calculate the results. His collaboration with IBM translates these procedures into software, developing organic grammars and Mutator, where an artist repeatedly selects and breeds variants. He characterises this role as gardening within a vast space of possible forms. Latham follows the work into rave graphics, computer games, virtual reality and molecular visualisation, explaining how collaborations and changing hardware reshape its possibilities. Fantasy viruses and mirrored compositions extend his interest in biological structures without representing organisms. Recent experiments with Google DeepMind connect Mutator to Gemini, which scores successive generations against prompts such as chickens or fly heads, replacing the human selector. He distinguishes this targeted evolution from proposed systems that change direction autonomously. Questions examine abstraction, physical sculpture, corporate ethics and artistic replaceability. Latham acknowledges that current AI selection evaluates a single viewpoint, not complete three-dimensional resemblance. He also considers whether the artwork resides in individual images or in the evolving software, and how creative control passes between people and machines.

AI Generated Summary