The Role of Computational Social Science in Addressing Societal Challenges

The next and final keynote speaker at COMNEWS 2023 is Noshir Contractor; his focus is on the potentials inherent in computational social science. Communication research has become central to any academic discourse around the world over the past decades, but this also means that we must take on the grand societal challenges of the present day.

The first such challenge is the acceleration of technological change, and Noshir here points to the early question-answering system IBM Watson, which showed its abilities by winning the TV show Jeopardy – its creator has said that building the team that built Watson was more difficult than building Watson itself. This includes assembling the team, managing their egos, and developing a collaborative ethos. The second challenge is health: appropriate healthcare remains highly unevenly distributed around the world, and even available solution are not necessarily taken up unless people’s social networks can be enrolled in spreading the message. The third challenge is the drive towards a mission to Mars: several public and private space agencies are working towards a mission to Mars, yet significant challenges in managing crew cohesion, teamwork, and morale remain. And the final challenge is the use of AI at work: how do human/AI teams capitalise on the differing abilities they bring to teams, and how can people be encouraged to work well with AI?

We are now ready to leverage computational social science in addressing these problems – because its methods can help us in designing and testing potential solutions. For instance, Noshir’s team has developed a dating-style platform for assembling collaborator teams, and by analysing the data this generated identified the various factors that influenced team formation – showing that familiarity with and recommendations of potential team members played an important part in the selection of collaborators. Such systems could also be tweaked to encourage more team diversity – but notably not by showing would-be team builders that adding certain participants would improve team diversity.

In health, the framing of messages in favour of health solutions is generally seen as important, but social influence plays an even more important role; one of Noshir’s projects used network analysis to examine the recommendation networks for contraceptive technologies in Kenia. In research to prepare for a mission to Mars, simulations of long-term space missions involving human test subjects are now common amongst many space agencies, and these can feed into agent-based behavioural modelling – but such modelling alone is not enough to address the problem, unless appropriate interventions to resolve relationship problems between crew members can also be found.

And what about the impact of AI on work? What are the structural signatures of high-performing human/AI teams? A current project explored this in a number of collaborative taskwork and teamwork settings – and it turns out that the AI is more appreciated if it provides taskwork contributions (doing things that contribute to the work) than teamwork contributions (facilitating the collaboration process itself).

This can also close the feedback look between computational social science and grand societal challenges: much as the research can address these challenges, the challenges also encourage the development of new concepts and methodologies.