Network Effects

"Many of the complex systems critical to human society can be understood as networks, including trans- portation, social interactions, trade, biological ecosystems, and communication, among others (Barab ́asi & P ́osfai, 2016; Jackson & Zenou, 2015; Newman & Newman, 2018). Networks consist of nodes (such as people, organisations, or resources) and connections (such as communication channels, infrastructural dependencies, or exchanges of goods and services). Network effects refer to consequences of the intricate relationships between the properties of individual connections and nodes, connectivity patterns, and the behaviours exhibited by the network as a whole (Siegenfeld & Bar-Yam, 2020)."(p. 23)

"The ongoing integration of AI capabilities into a wide range of existing networks, both virtual and phys- ical, is rapidly transforming the way our interconnected world operates. From business communication systems and financial trading networks to smart energy grids and logistical networks (Camacho et al., 2024; Ferreira et al., 2021; Mayorkas, 2024), entities or communication channels that were once controlled by humans are increasingly becoming AI-powered. This shift represents a systemic change in the way business, social, and technological networks operate, promising significantly improved efficiency and a greater diffusion of benefits from advanced AI, while also introducing novel risks."(p. 23)

"This underlying structure means that a networked system can suffer from a range of failure modes that individual, disconnected systems do not, such as the spread of malfunctions, phase transitions, and undesirable clustering or homogeneities (Cohen & Havlin, 2010). Importantly, a system’s behaviour within a network often differs from its behaviour when characterised independently.22 Non-AI examples of these phenomena include power grid blackouts (Buldyrev et al., 2010; Shakarian et al., 2013), flash crashes (Elliott et al., 2014; Paulin et al., 2019, see also Case Study 10), ecosystem collapse (Bascompte & Stouffer, 2009; Gao et al., 2016), or political unrest and conflict (Forsberg, 2008; Wood, 2008)."(p. 23)