Harnessing ‘human computation’ for 21st century humanitarianism

World experts in the emerging field of “human computation” gathered in Washington, DC last week to discuss the possibly unprecedented capabilities that can arise from innovative human-computer collaborations.

The study of artificial intelligence has traditionally asked, How can we build more intelligent machines?; human computation asks the more specific question, How can we leverage the strengths of humans and machines in combined systems that exceed the best each can do alone?”

One common approach to human computation reverses the familiar roles of humans and machines, with the computer assigning groups of people to address a problem and integrating their answers.

Last week’s Human Computation Roadmap Summit in the US capital explored the prospective impact of human computation to delineate research areas and activities that may lead to beneficial national and societal outcomes.  

‘Participatory dynamics’

Representatives from the White House Office of Science and Technology Policy and other federal agencies were present to lend insight on how to communicate workshop findings for the improvement of society through research into human computation.

A key aspect of the workshop was to model activity selection, workflow, group composition and interaction dynamics in existing human computation research – in other words, to employ the very methods being studied in service of their own advancement. 

The event began by engaging participants in a fast-paced game on climate risk management under changing conditions, enabling individuals and teams to experience the challenges of making decisions with limited capacity to process information. 

This activity not only highlighted the educational value of participatory learning and action – one aspect of human computation – but also illuminated the “participatory dynamics” involved, such as the cognitive bias and incentive mechanisms central to human computation more broadly.

River data

Exposing workshop participants to the concept of climate risk management motivated one group to pursue a human computational approach to Red Cross pilots of forecast-based flood preparedness in Africa. 

The working group’s approach will also build on the Climate Centre’s crowdsourcing game Upriver, and explore the potential of hybrid human-computer systems to analyze river data in humanitarian work, using new techniques like participatory sensing, gamification and citizen science.

Pilot projects such as this represent one of many possible classes of human computation solutions relevant to humanitarian work. 

Other examples include web mashups and collective action in support of disaster response; mass problem-solving to enable people to collaborate effectively toward solving wicked problems; organismic computing with possible applications for coordinating relief work; pervasive gaming and others.

It’s hoped the Washington, DC workshop will inspire new research into human computation and its potential to expand the scope and effectiveness of humanitarian response.

‘Infodoodle’: A visual capture of the opening session of the Human Computation Roadmap Summit in Washington DC. (Sunni Brown)