Hipster effect: when all non-conformists look the same



    An interesting research paper was published by researcher Jonathan Touboul of the Laboratory of Mathematical Neurobiology at the College de France. The topic of the scientific article, in fact, is disclosed in its title: “Hipster effect: when all non-conformists look the same” (pdf) .

    “In such diverse fields as statistical physics, spin glasses, neurobiology, social sciences, economics, and finance, there is a situation everywhere when interacting objects decide to act in accordance with the majority (mainstream) or opposite to it (hipsters),” the author writes. “However, stubborn attempts to stand out often lead to the fact that all hipsters consistently make the same decisions, that is, they look the same.”

    Jonathan Tubul explains this paradox using methods of statistical physics, namely the spin glass model .

    According to the definition from Wikipedia, spin glasses are a material that is considered as a state of a magnetic system with a random distribution of spin-spin interactions. There is no long-range order in the system, and the disorder in the system is frozen, that is, it does not change with time. The energy of the exchange interaction oscillates, changing sign, depending on the distance between the atoms.

    Tubul considers conformists and nonconformists in the system as units with a positive or negative sign {-1; 1}. According to the author, an important condition in the correct model is a delay in the transfer of information between nodes of a social network.

    In this model, the personi randomly decides to become a hipster with probability q or remain a conformist. Each person is determined by his current state s i (t) , and the state of the network is described by the vector s (t) ϵ {-1; 1} n , which varies randomly with probability m i (t) from the point of view of an individual person i . His point of view is influenced by how he sees the system and the history of the development of the system. In particular, people i assigns a fixed weight J ij ≥ 0 person j and observing its state after delay Ƭ ij. Thus, at time t, person i observes the following trend:



    After choosing a configuration, the system develops in accordance with a random Markov process.

    Simulation of this process shows that, as the system develops, there are long time intervals when representatives of the hipster population synchronize with each other.

    Also popular now: