Abstract:

Human technology is an open-ended system that increases in diversity and complexity over time. Explanations of open-endedness normally appeal to the idea of cumulative cultural evolution (CCE) and the ability to accumulate increasingly effective solutions. Less attention is paid to the search spaces in which cultural evolution unfolds and how changes to these spaces induce open-ended growth. Adopting a macrolevel modelling approach, where cultural evolutionary dynamics are simulated as two interacting processes of optimization (where populations generate and filter for increasingly effective technologies) and discovery (where populations change their existing search space in an effort to seek out new configurations), we investigate the situations in which open-ended CCE does and does not emerge. A key finding is that populations routinely end up in optimization traps: here, the long-term dynamics gravitate towards an equilibrium state, failing to exhibit open-ended growth. Escaping these optimization traps requires populations to reach a balance between optimization and discovery, whilst accelerated growth in complexity only occurs when populations engage in directed expansion of the search space. Such findings suggest that discovery dynamics, and the ability of populations to change the underlying search space, are a fundamental limiting or enabling constraint on the open-ended evolution of technology.

Speaker's social media:

@replicatedtypo