Towards Complex Artificial Life
This work addresses the challenge of creating evolvable artificial life systems for researchers in artificial life and evolutionary computation, but it appears incremental as it builds on existing concepts like modularity and concurrency without demonstrating broad new capabilities.
The paper tackled the problem of constructing an artificial organism capable of evolving into more complex forms by using an object-oriented combinator chemistry, resulting in a moving, self-replicating, spatially distributed assembly called a 'roving pile' that supports modularity and concurrency.
An object-oriented combinator chemistry was used to construct an artificial organism with a system architecture possessing characteristics necessary for organisms to evolve into more complex forms. This architecture supports modularity by providing a mechanism for the construction of executable modules called $methods$ that can be duplicated and specialized to increase complexity. At the same time, its support for concurrency provides the flexibility in execution order necessary for redundancy, degeneracy and parallelism to mitigate increased replication costs. The organism is a moving, self-replicating, spatially distributed assembly of elemental combinators called a $roving \: pile.$ The pile hosts an asynchronous message passing computation implemented by parallel subprocesses encoded by genes distributed through out the pile like the plasmids of a bacterial cell.