Human simulation is the application of techniques from computational simulation and data science to achieve two complementary goals. The theoretical goal is to enhance research in the humanities and human-science domains with reference to the nexus of human minds, human cultures, and physical environments. The practical goal is to solve urgent social and public-health problems using data-driven decision-support tools and artificial environments for policy exploration.
The Human Simulation Group, which operates this site, is an international collective of scholars and researchers representing several organizations and numerous research projects. There are not many venues where philosophers, ethicists, and historians work alongside social scientists, public-health experts, policy professionals, legal scholars, computer engineers, and strategic communications specialists. This radical multidisciplinarity is part of what makes the Human Simulation Group distinctive.
What can Human Simulation do? Read on to find out.
Homo sapiens is an extraordinarily complicated animal. The study of our social behavior, our minds, our cultures, and our evolution is challenging for even the very best experimentalists and theorists. We have a lot of university disciplines to draw on, and all of them are useful. But complexity forces us to take perspectival slices through knotty problems, and it is easy to forget that our disciplinary perspectives only ever tell part of the story. The practical side effect of this is a persistent inability to solve peculiarly complex problems. Climate change, structural racism, persistent poverty, needless disease, human trafficking, extremist violence, corrupt politics, immoral technology: we might fix these problems if we knew how, but the underlying complexity makes the fixes hard to find and saps our determination to find the solutions we desperately need.
Fortunately, increasingly technologized societies have also given birth to powerful new methods that are better suited to handling complexity than methods of the past. Data science and computational simulation are a particularly potent combination, and these are the distinctive tools of human simulation.
From a theoretical perspective, powered by these new methods, human simulation demands expertise from every university discipline, whatever might be relevant to a given problem. Universities all over the world are striving to overcome balkanization of their departments, which is steadily rendering them increasingly irrelevant (and expensive) cultural luxuries. That’s not an easy path. Human simulation smooths the way, making radical multidisciplinarity the norm instead of an exception. Pointless turf wars and reductionist thinking can become disasters of the past, supplanted by a more cooperative approach to hypothesis-generation, study-design, theory-building, evidence-assembly, and interpretation-construction.
From a practical point of view, the union of data science and computational methods in human simulation brings complex and urgent human social problems within reach. Can we do something about suicide rates that keep climbing? Everything we’ve tried so far has failed to arrest the increase. But new methods built for handling complexity could make a difference. What about the exploitation of children in the commercial sex industry? This multi-billion dollar market, driven by an invisible and seemingly insatiable demand, has been launched to a new level by an unintended side effect of social media that connects a supply to that lurking demand. It is proving incredibly difficult to reverse the expansion of that market because the complexity of problem defeats traditional intervention methods. We need new ways of thinking and human simulation delivers them. The opioid, obesity, and COVID-19 epidemics are extremely complex and persistent. Public-health professionals need new kinds of decision-support tools that can take account of that complexity.
Human simulation has other virtues, too. Many leading-edge theories in the human sciences and public health target processes that occur on massive temporal and spatial scales, or posit mechanisms that depend on the complex interaction of thousands of independent causal processes. These kinds of complexity defeat traditional empirical methods not only because of intrinsic complexity but also because relevant data is scarce. The union of data science and computer simulation offers a way through such challenges by using simulated data from validated simulations to plug gaps in existing datasets. The result is artificial experimental platforms, within which we can ethically and safely conduct virtual experiments, testing “what-if” scenarios, evaluating policy alternatives, and even testing theoretical syntheses. Human simulation is a way of learning about ourselves, our evolution, and our future in artificial environments that nonetheless offer complex – and often surprising – responses to questions posed in the language of code.
HumanSim.org provides a forum for researchers, scholars, and industry experts to keep each other and the public up-to-date about their ongoing projects and findings in the computational human sciences, computational humanities, and computational public health. You’ll find project descriptions, code for computer models, links to decision-support dashboards, and information about the international network of researchers who are embracing the marriage of data science and computational simulation. HumanSim.org is a tool for harnessing the growing power of the global research community whose passion is applying computational tools to the study of human affairs.