On March 25th, 2020, the most powerful computer on the planet exceeded a speed of 1.5 exaFLOPS, meaning a speed of 1,500,000,000,000,000,000 operations per second. This computer is not some massive supercomputer in a top secret lab, but a network of computers all over the world owned by ordinary people. The effort coordinated by the Folding@Home organization sent out simulations to be solved by anyone who volunteered their computer’s processing power. At its peak, over 700,000 people were contributing to this project, making it one of the biggest collaborative projects to solve the coronavirus problem. To understand this phenomenon, it’s important to understand what crowdsourcing is and how it is used.
With the increasingly connected nature of the internet, more issues are being approached collaboratively, and the coronavirus is no exception. Crowdsourcing is “the practice of obtaining needed services, ideas, or content by soliciting contributions from a large group of people…”. Historically, crowdsourcing has been applied to low stakes problems or issues which are not time sensitive. This would in turn lead to the solutions from the crowd being less practical. Often these problems would be local and non-critical which created a stigma of crowdsourcing being an ineffective way to solve problems.
This all changed when it came to the Coronavirus pandemic. With many people stuck at home when the virus data was made public, people capitalized on it in unique ways. One of the earliest projects was the data released by Johns Hopkins University tracking the spread of Coronavirus infections. Maps and graphs made from this data visually depicted how far and fast the infection could spread. Johns Hopkins opened up the data they used to make their tools to the public, and the internet capitalized on this and made other resources the average person can use to inform themselves. This more implicit way to crowdsource is in contrast to the approach taken by other organizations like MIT. MIT created an open-ended competition for different solutions and asked people to be creative.
To follow up, Google and Apple collaborated to create a coronavirus contact tracing API that apps can use to notify users if they were in close contact with infected individuals. This API was not enabled by default, but rather apps had to build this technology into the software and users had to opt in to install the apps and all the attached consequences. Privacy concerns arose among the public, who considered that Google does not have a track record of respecting users’ privacy if not specifically requested to do so. If everyone participated in this project, then it would be a lot easier to track and mitigate the spread of the virus, but if no one did, then the work put into the API would be essentially useless. This raises the question of if it is worth giving up personal information to a large corporation to secure the health of the community. As more data is collected on the success of these apps, a better conclusion can be drawn on how effective this crowdsourcing attempt was.
A very successful crowdsourced project recently was the Folding@Home simulation. Folding@Home creates models of the virus and investigates how the proteins in the virus fold and interact with other molecules the virus comes in contact with. Simulations of the folds are created and sent off to anyone who has the Folding@Home app installed on their computer. So many people installed the app and accepted simulations the servers were overloaded and could not provide any more simulations to people getting started. One contributing factor to the success is the relatively low cost of entry. The app can be configured to only receive and work on data when the device is not in use, and requires no additional work on the part of the user to contribute. Another factor that boosted Folding@Home’s popularity is the fact that this project was also somewhat gamified — with users being able to join teams and keep track of contributions on a public scoreboard Unfortunately, the success of this project might be difficult to replicate in the future as it requires minimal effort for anyone contributing, but still has to be an effective use of all that computing power.
Finally, a massive crowdsourced effort was to create masks and face shields for the local community to compensate for the breakdown in transportation at the beginning of the pandemic. Due to the lack of easy access to PPE, people across the nation stepped up and knitted cloth masks and 3D printed face masks to donate them to essential and healthcare workers. This crowdsourced solution was not prompted by anyone, but the movement was able to get widespread enough through word of mouth and social media that it was almost enough to make up for the lack of support through the official channels. This form of crowdsourcing relies on volunteers willing to do something to make a change on their own despite no group directly pushing for a change. It seems to be the hardest to replicate, and might only be possible in the most dire situations.
All of these examples show how with the right conditions the community can come together to solve problems faster and more effectively as a collective than as separate individuals, even if some were more successful than others. In the past, crowdsourcing was seen as a lazy way to solve a minor problem whereas now it seems like a viable way to create solutions for the global community. Was this just a fluke of the right conditions at the right time, or does this mark a shift towards a future where crowdsourcing becomes the norm rather than the exception? How can a project best optimize itself for crowdsourcing? These questions are what might take us into the future but right now, collaboration is saving lives in ways individuals simply cannot.
Edited by Lina Itenberg