Human-Machine Social Systems

Bridging sociology, social psychology, human-computer interaction, and web science, HUMANET combines experimental and computational research methods to study a pressing new social challenge: the ever-increasing interdependence between humans and autonomous machines such as robots, bots, and algorithms.

Starting from the idea that humans and machines form a complex adaptive social system, HUMANET investigates how human-machine, machine-machine, and human-human interactions influence and affect each other and how they add up to different collective outcomes.

Conceptually, we distinguish between three levels of analysis: interactions, networks, and communities, and study direct, indirect, and strategic influence, network composition and structure, and techno-organizational constraints and sociocultural context.

In terms of methodology, we combine controlled virtual lab experiments, agent-based models, digital-trace data, online field interventions, and heterogeneity and case comparison to conduct and integrate comprehensive analyses at the three levels.

Empirically, we focus on the problem of social influence and contagion and provide causal and observational evidence from a range of online human-bot communities, including an online collaboration community, a discussion site, and a crowdfunding platform.

HUMANET aims to expand our empirical knowledge of existing human-machine social systems, generate testable theories about human-machine interactions and networks, advance methods for modelling and conducting experiments with artificial agents, and bring public attention to the increasing algorithmization of our daily lives.

Its ultimate goal is to build the foundations of a new cumulative empirical sociology of humans and machines.

A better understanding of human-machine social systems is becoming crucial for reducing misinformation, preventing financial crashes, improving road safety, adapting to labor market disruptions, and more.