You are here

Big Data and Deep Learning

Picture of screen with code on it
June 5, 2020

Ted Petersen Explains How AI Changes Covid-19 Tracking and Recording

... the line between data and artificial intelligence is sometimes hard to draw because a lot of modern artificial intelligence methods do rely on data, and so they tend to be data driven...

With states reopening and everyone craving a return to normalcy, tracking Covid-19 cases and predicting outbreaks becomes an increasing concern. Everyone debates using AI technology to help, but what would it do? With thousands of journals being published about Covid-19 from Universities containing research and possible testing it would be impossible for a human to read all of them, and then compare data. AI collects information from each and every journal published and compares it to all other published journals on the subject. It then filters outdated information and any misinformation and presents the viewer with the most accurate and important data for them to read.

The inner workings of AI

Ted Petersen explains that, “... the line between data and artificial intelligence is sometimes hard to draw because a lot of modern artificial intelligence methods do rely on data, and so they tend to be data driven, and you hear terms like big data. And another common sort of breaking news type turn is deep learning, which is all about data. I mean you have to, it all starts with data…  But usually artificial intelligence is maybe drawing conclusions or inferences from data that we would typically only expect a human could do… Of course we kind of get used to computers doing things that we don't expect them to do and so once that happens, it's no longer really artificial intelligence, like, like credit card fraud is predicted.”

As it relates more to Covid it “would use content tracking to record who that person had been in contact with. let's suppose I tested positive for COVID and gone to the grocery store last week. I don't really know any of the people at the grocery store that I saw. I mean just, there, there, by and large strangers I know some of the people who work there by name, maybe and I could identify them but a lot of people that couldn't identify so there's a, there's this kind of trade off, which I think comes up quite a lot now between this individual privacy versus sort of greater good.” This way of recording could keep track of how many people were possibly exposed and could notify individuals if they need to self-monitor for symptoms.

What could go wrong?

But there are also ethical concerns with AI tracking, and mostly with privacy. If people agreed to download the app and be used to track their symptoms and locations there could still be a possibility for error. With the wide range of symptoms for Covid-19 it would be extremely difficult to maintain or achieve 100% accuracy. Where privacy comes into play though is the access to possible medical information, constant use of location, and sharing of data with the cause of public health, but it would still require knowledgeable consent from the users. 

AI is more common than you think

Some good things to have come out of AI include credit alerts for fraud, creating transcriptions of live chats,  and could help in the medical field with predictions for outbreaks. “I think I was in Toronto. And I was staying somewhere paying for where I was staying with a credit card I bought the ticket to Toronto with a credit card, and I go to a store in Toronto and the card is basically flagged.” It’s not perfect but it decides based on the data it’s presented with. The AI technology would collect data from a certain area and record the trends they’ve seen in cases, need for certain supplies like ventilators, and calculate what they would need in the case of another outbreak and how likely an outbreak might be.

photo of Ted Petersen

About Ted Petersen

Ted Petersen is a computer science professor at UMD and deals with facial recognition and AI ethics. With his own experiences with AI and his research Ted explains both concerns with AI and benefits. He’s included information from the news and real-life situations that are currently using AI tracking for Covid-19.

About the UMD Computer Science Program

Ted Petersen's Homepage

This story was written by UMD student Bailey Jacobson, who is double-majoring in English and writing studies. Bailey works with Cheryl Reitan in University Marketing and Public Relations.