I learn and create the laws of the universe and primordial processes from everyday life.

Late one night, as I cooked a pot of noodles, I had an idea: Could heating and evaporating water mirror the birth and death of stars? Water has been found in interstellar clouds, comets, and planets' atmospheres, and life in the galaxies is often considered to be related to water. Stars form after the gravitational collapse of molecular clouds, which converts potential energy into thermal energy. Similarly, the phase change of water molecules from liquid to gas involves transforming thermal energy, which could be connected to the large-scale energy conversions and morphological changes observed in stellar processes. Just as an individual pixel contributes to an entire image, molecular changes contribute to forming planets, stars, and other cosmic structures. There may indeed be a set of related principles governing small-scale transformations and large-scale cosmic events.

While the recent surge in AI image processing has captured significant public attention, what intrigues me most is the underlying thought process of neural network algorithms, their research goals, and the sources of the image data they use. Drawing inspiration from biological cell communication, researchers have developed neural cellular automata, a leap from classical cellular automata. Unlike their binary predecessors, neural cellular automata can generate images using floating-point numbers between zero and one, allowing for more nuanced representations. These systems can be trained to regenerate an image after it has been damaged or altered. More than just a tool for image processing, neural cellular automata hold potential applications in synthetic biology, offering insights into processes of growth, repair, and regeneration.

"Stochastic Camera (version 0.3) - the boiling terra" intends to apply these computer science studies on tissue texture to produce landscape images. Depicting a planet capable of self-growth and self-repair, akin to living tissue. By recording the process of water boiling with an infrared thermal imager, applying neural cellular automata, and 3D rendering to the resulting flat images, the work generates hypsometric maps and stellar spectra, which evoke the idea of planets and stars as dynamic, evolving entities. The boiling water, captured in its ever-changing thermal patterns, becomes a metaphor for the continuous transformation of matter in the universe, from the boiling water to the birth and death of stars.

‍