A new project called CrowdSolve debuting on Indiegogo today asks the question: What happens if you use the “wisdom of the crowd” to solve crime? It’s a marked contrast from the heroic image of the solitary detective, a la Batman or Sherlock Holmes, but one that might resonate more with our recent collective fascination with Serial, the ongoing podcast spun out of This American Life documenting the murder of a Baltimore woman and the conviction of her accused killer.
Serial’s story arc includes a murder, the subsequent investigation and a conviction, but it has still resulted in a veritable legion of amateur detectives who collectively pore over and obsess about the details of the case, gathering on Reddit and other platforms and working together to potentially uncover new evidence and propose alternate theories of what really happened. CrowdSolve’s founder Colin Helibut wonders whether the same intense focus could be leveraged to help the roughly 50,000 to 120,000 wrongfully convicted people in prison in the U.S. right now (using study estimates of 2.5 percent to 5 percent in jail currently wrongfully convicted, with U.S. prison population of 2.4 million).
CrowdSolve’s goal with Indiegogo is to raise $50,000 to help pay the government and administrative fees associated with accessing public case files and other documents, and to build visualization tools that will allow the community to better make sense of their data. The community will include moderation, and an up/downvoting system similar to Reddit to help make sure things stay on track and the best information rises to the top, and participants will build cases and theories that they can then hand off to others, or work on collaboratively and have vetted by the community.
Of course, everything about this idea should be setting off alarm bells in your head right now – the crowd’s “wisdom” is not always best defined as such, after all. Reddit has demonstrated the weakness of the concept of crowd-sourced investigation by mis-identifying suspects in the past, including the Boston Marathon bomber. Helibut tells TechCrunch that the idea behind CrowdSolve is actually to counter some of Reddit’s inherent weakness when it comes to crowdsourced crime fighting.
“Reddit is a wonderful and hugely important community, but at its core, it’s purpose is general conversation and news aggregation,” he said. “Their basic setup and infrastructure is not well-suited to data-organization, citation, annotation, etc. On the other hand, CrowdSolve is being built from the ground-up to foster the exploration and exposition of criminal investigations in a transparent, tightly monitored way.”
Helibut also answers concerns about the weakness of the model by pointing to the sacrifices made in doing nothing: The Innocence Project and others devoted to reversing wrongful convictions have limited resources, he points out, and there are precious few alternatives that can scale to properly address the problem.
It’s true that if CrowdSolve can actually help reverse wrongful convictions, it will have a social impact that’s hard to overvalue, but there’s a very good reason investigators often refrain from sharing case details with the public early on – groups can make mistakes just as easily as individuals, and the ramifications of making erroneous assumptions in the public sphere is often far more devastating than making them behind closed doors. Even with so-called “cold cases,” amateur investigations have the potential to upend lives, including those of innocent people close to the case.
Data, analytical tools and moderation might be the key to regulating a process that could easily spiral dangerously out of control, but there’s little evidence yet to suggest using the distributed processing approach to solving crime will have the same impact to social good that crowdsourced gene sequencing has had to science: the former involves a lot more reliance on the fragile and changeable beast that is human judgement, and that’s a powder keg ready just waiting for a match.