New research is now starting to explain why so many biological networks, including the human brain, exhibit a hierarchical structure, which will help improve attempts to create artificial intelligence. This is demonstrated by showing that the evolution of hierarchy, or a simple system of ranking, in biological networks may arise due to the costs associated with network connections. Like large businesses, many biological networks are organized hierarchically, such as gene, protein, neural and metabolic networks. This means that they have seaparate units that can be divided and made even smaller.
Yet why is it that so many biological networks evolve to be hierarchical? The results of this study suggest that it evolves because hierarchically wired networks have fewer connections; biological networks are expensive, so there’s an evolutionary pressure to reduce the number of connections. As well as shedding light on the emergence of hierarchy, these findings could also accelerate future research into evolving more complex, intelligent computational brains in AI and robotics. The researchers, from the University of Wyoming and INRIA, stimulated the evolution of computational brain models both with and without a cost for network connections, and discovered that hierarchical structures emerge much more frequently when a cost for connections is present.
For over a decade, the authors say, they have been working to understand why it is that networks evolve to have the properties of modularity, hierarchy and regularity, and with these results they’ve since uncovered evolutionary drivers for each of these properties. These findings not only explain why biological networks are hierarchical, but also why many human-made systems such as the road and Internet are as well.
The next step, the researchers say, is to harness and combine this knowledge in an effort to evolve large-scale, structurally organized networks to create better artificial intelligence and increase their understanding of the evolution of animal intelligence.
If you’d like to learn more, you can click here!