Emergence in artificial life

Even when concepts similar to emergence have been used since antiquity, we lack an agreed definition. However, emergence has been identified as one of the main features of complex systems. Most would agree on the statement ``life is complex’’. Thus, understanding emergence and complexity should benefit the study of living systems. It can be said that life emerges from the interactions of complex molecules. But how useful is this to understand living systems? Artificial life (ALife) has been developed in recent decades to study life using a synthetic approach: build it to understand it. ALife systems are not so complex, be them soft (simulations), hard (robots), or wet (protocells). Then, we can aim at first understanding emergence in ALife, for then using this knowledge in biology. I argue that to understand emergence and life, it becomes useful to use information as a framework. In a general sense, I define emergence as information that is not present at one scale but is present at another scale. This perspective avoids problems of studying emergence from a materialist framework, and can be also useful in the study of self-organization and complexity.

💡 Research Summary

The paper tackles the long‑standing problem that “emergence” lacks a universally accepted definition, despite its central role in complex systems theory and its obvious relevance to biology, where life is widely regarded as a complex phenomenon. The author proposes an information‑theoretic perspective: emergence is defined as “information that is absent at one scale but present at another.” This definition sidesteps the pitfalls of purely materialist or reductionist accounts and naturally accommodates both spatial/organizational (synchronic) and temporal (diachronic) scales.

Two major categories of emergence are distinguished. “Weak emergence” requires only computational irreducibility: the system’s macroscopic behavior cannot be predicted without simulating all intermediate steps, even though the underlying rules are known. “Strong emergence” adds the notion of downward causation, whereby higher‑level structures constrain or generate novel dynamics at lower levels. The paper illustrates strong emergence with biological examples (cellular organization limiting molecular possibilities) and argues that similar mechanisms can be observed in artificial life (ALife).

ALife is presented as a synthetic laboratory for studying emergence, divided into three domains:

1. Soft ALife – computational models such as Conway’s Game of Life, cellular automata, and boid flocking simulations. Simple local interaction rules give rise to complex patterns, self‑replicating structures, and even universal computation. While most examples are upward (low‑to‑high) emergence, some studies report downward information flow, supporting the strong emergence view.

2. Hard ALife – embodied robots interacting with physical environments. Emergent properties appear both at the individual robot level (coordinated locomotion, sensorimotor loops) and at the collective level (task allocation, swarm intelligence). The physical embodiment makes these systems closer to biological organisms, and the emergence of group‑level capabilities that individual robots lack exemplifies both weak and strong emergence.

3. Wet ALife – chemical protocells, active droplets, and xenobots that bridge chemistry and engineering. These systems demonstrate self‑assembly, metabolism‑like processes, and motility. Crucially, the higher‑order organization (membranes, reaction networks) can restrict the behavior of constituent molecules, providing concrete instances of downward causation.

To quantify emergence, the author adapts Shannon entropy into a normalized “emergence index” E = –K ∑ p_i log p_i, where p_i are the probabilities of states at a given scale and K normalizes E to the interval