This is a guest post, written by David W. McDonald, Chair and Associate Professor in the Human Centered Design and Engineering Department at the University of Washington.
Designing a good application is hard. Designing a good computer system is harder. Designing a good system that accounts for the vagaries of people, their motivations, and their flaws is even harder. Yet that is the challenge that designers of social computing systems must solve. The difficulties of designing social computing systems derive from both the complexity of the software and hardware configurations, and the fact that some participants in a social computing system will not behave with positive goals or intent. That is, sometimes people who participate in a social computing system try to take advantage of the system and the other participants for their own benefit. The way that social computing was used to increase looting during the 2011 London Riots or the ways in which ISIS has gamed Twitter are concrete demonstrations that social computing can be used for less than noble goals. With growing participation rates in social computing systems, system designers must work to prevent or mitigate potential risks that these systems pose for the participants and to society.
Last June I attended a visioning workshop on Human Computation sponsored by the Computing Community Consortium (CCC). At the workshop I participated in a small group activity that explored risks in social computing systems. Our group explored potential risks by taking an “anti-design” approach. What I mean by “anti-design” is that our group inverted commonly accepted working assumptions to understand how complex social computing systems could go wrong and be used against desirable societally positive goals. We described our design exploration and identification of a number of critical research areas in a recent Interactions article titled Antisocial computing: exploring design risks in social computing systems.
Exploring design risks in social computing systems is a challenge. Social computing systems represent a complex interleaving of human behaviors, human interactions, algorithms and collections of hardware and devices resulting in a design space that is necessarily uneven and irregular, making some design choices difficult to fully comprehend. Further, when we start to design social computing systems, we often start with an idealized view of who the participants are and how they will behave. Often we begin with unconditional positive regard for the future participants. Designers often do not consider how people ‘play’ with systems and that what is ‘fun’ to one set of individuals is destructive or abusive from another person’s perspective. We do not assume that the participants are out to abuse another individual, create mayhem and panic or that they are out to ruin users’ experience of the system. Ultimately, because we often focus on positive outcomes with positive assumptions about user behaviors, designers only explore a small portion of the total design space leaving potential risks unclear.
The potential societal benefits from social computing are immense. More effective crisis response, safer neighborhoods, more satisfying work and family relationships, entertainment and leisure, responsive education and workforce training, cost effective health and wellness are desirable and achievable through well designed social computing platforms. Yet, the potential economic and societal value of well-designed and effective social computing systems is not assured. Despite the best of our intentions, the wicked design space of a complex social computing platform presents risks to participants and society that designers must begin to explore, understand and make design decisions that mitigate or prevent those risks.
Full Article: D. W. McDonald, D. H. Ackley, R. Bryant, M. Gedney, H. Hirsh, and L. Shanley. 2014. Antisocial computing: exploring design risks in social computing systems. Interactions 21, 6 (October 2014), 72-75. DOI=10.1145/2675705 http://doi.acm.org/10.1145/2675705
