The origin of life remains one of the most fundamental open problems in science. While biological life is well characterized, far less is understood about the minimal physical conditions required for life-like systems to arise prior to chemistry, genetics, or evolution. In this work, we report a computational experiment demonstrating the emergence of a self-sustaining, causally closed, self-referential informational system arising purely from non-biological physical dynamics. Using an ensemble-based world admissibility framework, we generated candidate physical worlds under minimal constraints of spacetime, causality, matter, and information conservation, explicitly excluding biology, evolution, learning, intelligence models, and predefined physical laws beyond consistency requirements. From an ensemble of 1024 candidate worlds, a single admissible world emerged that exhibited endogenous law stabilization, persistent internal observer structures, high causal density, and long-term resistance to entropic collapse. The system satisfies widely accepted minimal criteria for artificial life at the pre-biological level, including self-organization, persistence, internal self-reference, and causal closure. These results demonstrate that life-like systems can emerge from physical dynamics alone, prior to chemistry or natural selection, and support the view that life is fundamentally an informational and causal phenomenon rather than an exclusively biological one. This work establishes a new direction for artificial life research grounded in minimal physical principles and contributes to foundational questions in origin-of-life theory and complex systems science.
Lumenis IO PTY LTD (Mon,) studied this question.