Crowd power makes sense of the unfathomable
When it comes to solving global crimes, some of today’s most successful detectives are increasingly ordinary citizens, working in concert to uncloak clues buried deep within the internet.
Their collective efforts identified the individuals responsible for shooting down Malaysia Airlines Flight 17 (MH17) over Ukraine in 2014. The MH17 perpetrators will now be brought to trial (in absentia) next March, such was the thoroughness of the investigation.
Citizen sleuths also identified the agents behind the poisoning of a former Russian spy and his daughter in the United Kingdom in 2018. In these two groundbreaking examples, the coordinating organisation was Bellingcat, an investigative journalism group based in the UK.
But many more citizen detectives will now be able to join in such activities, through the University of Melbourne’s new Hunt Laboratory for Intelligence Research, named after Atlee Hunt who established Australia’s first national intelligence capability in 1901.
The Hunt Laboratory builds on the University’s SWARM Project, incorporating its novel approach to intelligence analysis based on a combination of crowdsourcing and structured analytical techniques.
The SWARM research already has government interest. The project was initially developed with funding from the Intelligence Advanced Research Projects Activity (IARPA), a division of the Office of the Director of National Intelligence in the US. However, the ‘Hunt Lab’ will broaden its remit to other kinds of intelligence-related research, including citizen intelligence activities.
Hunt Laboratory leader, Dr Tim van Gelder from the School of Biosciences, says the idea of many minds at work might sound like an obvious approach; however, group dynamics are complex, especially in hierarchical institutions.
What we’re exploring is an alternative to the way intelligence problems are traditionally handled, he says.
Since WWII , the classic approach has been for an analyst to research an issue, draft a report that is circulated to colleagues for feedback, and submit a final report to decision-makers. Recent decades have brought changes in many underlying technologies, but not to this model of creating the final report.
Dr van Gelder says the consequence of this is that most intelligence reports are built around the perspective of the analyst primarily responsible for drafting the report. But in today’s extremely fluid technological, social and political environments, this makes intelligence more challenging.
We have the spies-on-park-bench clichés, but intelligence is, at its core, ‘sense-making’. It is helping decision-makers make sense of a situation or circumstance, and this is becoming more important than ever, he says.
The world is becoming more complex, more dangerous, more unpredictable and more ambiguous.
In this context intelligence agencies need a more effective and more reliable approach to obtaining the best intelligence and to making the best decisions. Dr van Gelder says this generally requires multiple perspectives and a range of skills acting in concert.
The SWARM team’s approach has been to construct a cloud platform to facilitate group thinking and analysis, and evidence-based reasoning to produce a single, coherent, intelligence-type report. The platform – built by Professor Richard Sinnott’s team in the Melbourne School of Engineering’s eResearch group – has some similarity to Google Docs, but goes further in supporting the development and evaluation of alternative analyses.
Bringing many minds together also introduces a diversity of skills and perspectives. This can greatly reduce the human biases that undermine good decision-making, particularly biases imposed by rank, gender, culture or personality type.
The SWARM technology includes an anonymity option to avoid decisions being skewed by domineering personalities or rank.
The smartest person in a room on a particular subject might not be the most senior person, Dr van Gelder says.
The platform allows people’s inputs and analyses to be considered on merit, not on who the author is.
He says SWARM more generically can be considered a ‘collaboration environment’ for producing reports.
Much of what white-collar workers do in both government and corporate sectors is basically report writing. We now offer a collaborative platform for producing better-reasoned reports.
A cloud-based version of the platform will also allow the Hunt Laboratory to mount public ‘intelligence challenges’ attracting teams that will compete, for a prize, to produce the best analyses. The challenges will involve genuine intelligence problems, which require the research and analysis of open-source information that may be beyond the resources of a single person or organisation to undertake.
These could include any context in which a group must make a collaborative decision and support that decision with a report that outlines the findings, evidence and reasoning behind them. Citizen intelligence, citizen journalism and even government policymaking are all possibilities.
A challenge might, for example, pose a strategic question such as: ‘What are China’s plans for expanding bases beyond the South China Sea?’ Bringing the focus of a crowd to bear on such a question, with the diversity of skills, knowledge and perspectives inherent in that, is likely to produce quite a different report to one arrived at solely through a military lens.
You have to remember that intelligence agencies are often small and thinly spread. They might only have one person specialising in a particular subject, so citizen intelligence can become a very valuable resource.
People will sign up to the platform as teams or individuals and be notified when a challenge has been presented.
Dr van Gelder says any organisation needing intelligence on a particular topic can put forward a challenge for the Hunt Laboratory to mount. It will be a fee-for-service arrangement, with the fee being part income for the laboratory and part reward for the group or individual producing the best report.
Open-source intelligence is an internationally growing phenomenon. Dr van Gelder says the Hunt Laboratory’s conversations with US and Australian intelligence agencies are shaped around how the laboratory can constructively foster citizen intelligence as a resource. This includes trying to anticipate, and prevent or mitigate, problems which may arise when involving the crowd in sensitive intelligence work.
The Hunt Laboratory currently represents the largest group of researchers in Australia focused on improving the quality of intelligence analysis.
Organisations with intelligence functions themselves generally don’t have research arms, Dr van Gelder says.
The idea of the Hunt Laboratory is to provide a trusted research partner and this includes having key researchers obtain the appropriate security clearances. We have runs on the board and credibility through our work with IARPA and now we are well placed to grow beyond the SWARM project to be the primary provider of intelligence research capability in Australia.
Dr van Gelder says the Hunt Laboratory team is confident the public can make a valuable contribution to security.
When we ran the SWARM technology through comprehensive testing last year it was teams from the general public that produced some of the best analytical work. It showed clearly ‘the power of the crowd’ and how this could augment the capability of intelligence agencies to make sense of our fast-changing region and world.
The Hunt Laboratory is a multi-faculty group based in the University of Melbourne’s Centre for Environmental and Economic Research.
Recruitment manager: Sujai Thomman
Project director: Professor Mark Burgman, Imperial College, London
Technical team director: Professor Richard Sinnott