Tipping Points and Diffusion in Networks
https://www.economist.com/news/2009/04/20/tipping-point
https://link.springer.com/article/10.1007/s10640-016-0074-7
The articles above are about “tipping points” in the world. However, these tipping points discussed are not exactly the same tipping points that we discussed in class about firms pricing their products and people’s expectations of its popularity. Instead, the articles focus on tipping points as they exist in nature and how they pertain to networks. For example, the first article discusses how in epidemiology (study of the incidence, distribution, and possible control of diseases and other factors relating to health), tipping points refer to how small changes in a structured system can significantly alter the network and have major impacts. One example the article gives is the influenza epidemic after World War I. Interestingly, nomenclature in epidemiology is commonly used in business organizations today, such as “viral marketing,” in which spreading a product/service through a strongly connected group is meant to have a similar effect as an epidemic spreading through a population. The other article also mentions how tipping points are a concept in environmental science as well, where passing a threshold in an ecological system can drastically and irreversibly damage it.
I thought this was very interesting and relevant to both our discussion of tipping points and the diffusion of networks. In fact, I thought that the examples that this article brought up reflected how the theory involved in creating our model of a market for a good actually applied to a structural network. For example, we raised the question in lecture of how to spread an idea through a network represented by a graph. We identified some nodes in a given graph as being more important, or at least more necessary, for a complete cascade of an idea/product across the network. Thus, we could probably relate these special nodes as a representation of a “tipping point”: starting with a graph with some nodes choosing “A” and other clusters choosing “B,” the spread of “A” depends on somehow switching certain specific nodes (due to their positions in the graph) to choose “A,” since if they switch the rest of their neighbors will end up following. However, without them, “A” will cease to diffuse through the network.