Machine Hallucinations: Nature Dreams and Quantum Memories emerge out of Refik Anadol Studio’s long-term project of compiling the largest nature-themed dataset for AI-based generative artworks with the aim of attracting attention to the preservation of nature.

Machine Hallucinations is an ongoing exploration of data aesthetics based on collective visual memories of space, nature, and urban environments. Anadol and his team collect data from digital archives and publicly available resources, then process millions of photographic memories with machine learning classification models. The sorted image datasets are then clustered into thematic categories to better understand the semantic context of the data universe. This expanding data universe not only represents the interpolation of data as synthesis, but also becomes a latent cosmos in which hallucinative potential is the main channel of artistic creativity. As a thoroughly curated multi-channel experience, Machine Hallucinations offers a new form of sensational autonomy via cybernetic serendipity.

Quantum Memories emerges from Refik Anadol Studio’s long-term investigation merging machine learning, the aesthetics of probability, and Google AI Quantum Supremacy experiments. It is a data sculpture that draws upon more than two hundred million publicly available nature-related images, which are used to train a GAN algorithm. Refik Anadol Studio then developed a custom software in collaboration with Google AI Quantum Research Team that utilized new ways of using circuit depth data and generated Quantum Noise. Quantum Memories can be considered both a radical visualization of an alternate dimension of the natural world, and of our digitized memories of nature.

The resulting AI Data Paintings incorporate pigments, shapes, and patterns that we associate with nature, but only exist in the mind of a machine as “hallucinations.” Each image in the series displays a cluster of chosen hallucinations, and Anadol makes selections from countless serendipitous allusions to real images occurring in the machine-mind. When idle and unsupervised, the AI generates abstracted visuals of these collective memories, constructing new artworks through unique lines drawn by algorithmic connections.