Artificial Intelligence vs COVID-19


The Coronavirus has shaken the world and even though government regulations and medication can prevent the effects of the virus, their slow implementation weakens their effectiveness. Every day, we are getting one step closer to a breakthrough in Artificial Intelligence, and AI may be exactly what we need to prevent a reoccurrence. Even now, AI has had a profound impact on mitigating the effects of Coronavirus and rapid developments in technology promise a safer future.

On December 31st, BlueDot, an artificially intelligent startup based in Toronto, alerted employees of an outbreak in China. Due to recent advances in computational power, the system “reads” through 100,000 articles every day across the globe in 65 different languages and compares it to past trends. Due to similarities between a news report in China and the SARS outbreak in 2003, it immediately flagged the article. Although this should have been enough for the government to issue legislation to prepare, the AI system did not stop there.

Employees switched the system to focus on passenger data from billions of airline itineraries after getting alerted and instantaneously, the system outputted which cities across the globe were going to get hit the hardest. Even with that, government officials were slow to act due to AI’s unproven track record. For example, Google’s AI massively overpredicted the 2013 cycle of seasonal flu based on their own data. Currently, we use an AI’s output as a mere suggestion in which governments take weeks to analyze. But in the coming decades, if the rate of improvements in Artificial Intelligence and Big Data remains constant, technology will be a lot more than just a suggestion. Even a small head start can account for millions of lives saved especially in cases of exponential growth. 

The basis of Artificial Intelligence is the data that they get from consumers like us which is why a powerful argument can be made for data being the most precious commodity of the 21st century. As the world becomes more interconnected every day, these systems can only become more reliable as they will have more data to work with. To help AI systems like BlueDot, big tech companies like Google, Twitter and Facebook have shared aggregate data regarding search data, posts about the virus, a user’s movement, etc. The challenge with all this data is feeding the right bits to the system to avoid it getting mislead by unrelated information that may sway it which is why humans play an integral role in AI companies currently as they have to adjust what data their systems rely upon. 

Another AI company named HealthMap based in Boston warned human employees about COVID before BlueDot. Like BlueDot, it relied on articles in China. To be specific, a forum called ProMed in Wuhan where local physicians discussed new pneumonia-like strands. After the detection, it used travel itineraries to simulate the spread. However, it soon became misleading since the Chinese government banned movements that these travel itineraries had predicted. ProMed switched to using location data of millions of phones and the four countries it predicted having the highest risk were the first four countries to receive cases of the virus. Now, they have added aggregate data from medical devices such as thermometers to the mix to get an even more accurate prediction of the virus. 

Harsh times like these prove the need for breakthroughs and along with thanking medical professionals and government officials, we need to be grateful to the programmers. As Gallivan says, “For earlier detection, it’s about creating a much smarter public health and medical system,” and in the future, we will see systems like these preventing major outbreaks like the one we are currently in.