An artificial intelligence created by New York-based risk detection firm Dataminr was able to predict where the next spike in coronavirus cases will be in the UK and US by analyzing social media posts, The Next Web reports.
According to the company’s website, “growth in clusters of eyewitness, on-the-ground, first-hand public social media posts on COVID-19” allowed their algorithm to detect “hotspots 7-15 days prior to exponential growth in COVID-19 official case count.”
These social media posts include “posts ranging from people indicating they tested positive, people indicating they are experiencing symptoms, people indicating they have been exposed but not tested, first-hand accounts of confirmed cases” and so on.
Dataminr also predicted future spikes in 14 different US states. Seven days later, all 14 states were hit hard by the coronavirus, TNW reports.
“These kinds of posts appear to be a leading indicator of where case count may be spiking in the future,” Dataminr CEO Ted Bailey told Business Insider.
The analysis of social media for predicting outbreaks could allow governments to better prepare for the next big crisis.
But any predictions by these kinds of systems have to be taken with a massive grain of salt. No artificial intelligence in the world can verify with 100 percent accuracy the validity of the claims social media users are making about tests being positive, or relatives showing symptoms.