How are researchers using artificial intelligence to help with the pandemic? Learn how AI is helping scientists, researchers, and healthcare professionals quickly process all the data being collected about the virus to better understand how to treat patients and develop a vaccine. In this episode, APU business professor Dr. Wanda Curlee talks to Dr. Oliver Hedgepeth about the role of AI during the pandemic, how it’s being used to understand the economic downturn, and how AI fits into the larger picture of technological advancement.
Listen to the Episode:
Read the Transcript:
Dr. Wanda Curlee: Welcome to the podcast. I’m your host, Wanda Curlee. Today, we will be chatting about how artificial intelligence can help the 2020 pandemic and the economic stresses being experienced. Today, my guest is Dr. Oliver Hedgepeth, who is a Professor at American Public University. He has many years of experience working with artificial intelligence. Oliver, welcome to the podcast and thank you for joining me.
Dr. Oliver Hedgepeth: Thank you, Wanda. I think it’s going to be a really exciting subject. And people don’t usually think of the pandemic and artificial intelligence in the same breath.
Start a Business degree at American Public University.
Dr. Wanda Curlee: Absolutely, you have vast experience in AI, another term for artificial intelligence. How are researchers using AI today to help with the pandemic?
Dr. Oliver Hedgepeth: I started in about 1995 when I was the Director of the Army’s AI Center, the first AI center. And we started developing things, machine-based tools to help people sort through really millions of pieces of data to try to find a trend because humans can’t read millions of pieces of data.
And we found there were some Army things that had 17 books that you had to read through of data to find something about an Army tank and that was hard to do. And so we thought AI would help ease the problem. It’s gotten better, a lot better.
And that’s the kind of technology the science community has started today with pandemics. In fact, one of the things that’s most exciting is that the pandemic does create volumes of data. Data being, here’s some information of how a potential chemical helped one patient. We’ll use President Trump. He used the chemical. They used it on him and he seems to be okay.
They’ve used similar chemicals before the others and they dropped dead. So there’s two pieces of data, one lives, one dies. And so that’s two pieces of data.
The interesting thing about this data, you think it’s just out there and it is. And the data is written and validated documents, procedures that are journals, peer-reviewed journals, academic journals. And there are about 200 journal articles that define each type of example I just said. Yes, 200 that are published each day, each day.
Now just remember, 200 journal articles come to you each day and you’ve got to read through and find out there was some chemicals worked, some chemical didn’t work. That’s what’s going on today in the scientific world. That’s kind of opening of a program we’re going to be doing today. So the pandemic that’s infecting the American world is really exciting and the word AI is being used to help explain how we’re managing some of this data.
Dr. Wanda Curlee: So Oliver, you talk about 200 journal articles, which are peer reviewed. Although, I would guess they’re quickly peer reviewed nowadays, especially when it has to do with COVID. How can researchers keep up with that, even with AI, because there’s going to be information, as you said, that is contradictory. So what is a researcher looking for from AI and those kinds of things?
Dr. Oliver Hedgepeth: Hospitals are producing a lot of this data. And yes, there are 200 professional journals that they’re produced probably quicker now than a year ago, only because the push on trying to find a solution for this pandemic is so hard now.
I remember two years ago people were pushing these documents to get published on what’s happening to everything, bad diseases. But, March 16th the White House, Office Chief of Office of Science and Technology, issued something called a call to action, a call to action document, and they ask all the universities, MIT, Harvard, all around the world, to try to develop some new text data mining software. We have it out there. But they said, let’s do it. And I guess they gave them money to do it.
And since then, one company called the Allen Institute of AI. They were really pushing hard on time to tie together all these databases, these publications, and. Even Google’s involved in this. Wouldn’t you believe that? Even Google is involved in tracking this data to try to find valid thing, it’s money.
And if you were to do a search and I hope people listening do a search on the pandemic and artificial intelligence and you will find database systems like AlphaFold or DeepMind or Semantic Scholar and COVID Scholar and things like the Blue Dot. And out of London, BenevolentAI, BenevolentAI.
These companies are coming together and trying to develop their computer programs that look for specific things that AI can manage better than humans can only because they do it faster. And that’s one of the things that’s happening right now.
Dr. Wanda Curlee: Interesting. Very interesting. So what aspects of the pandemic can AI help with? I mean, it can’t do everything, but it could help in some areas. What do you, from your research, what do you think it’s doing the best at?
Dr. Oliver Hedgepeth: Well, the one thing is it’s doing a good job in terms of finding what does not seem to work. You may have a physician in Montana who says, “I tried this and boy, I think it works.” The AI system can track the data like that after it’s been published and find out that, yeah, it worked there in that one case for that one patient who was 39 years old, who also had a series of other medical problems maybe that helped him survive. But there may be other cases where it was done in London or some other part of the United States where a similar person, it didn’t work for them.
Here’s the thing about AI. Artificial intelligence will do a real good job of tracking data and saying, here’s a trend. But we still need humans to look at AI. Don’t rely on AI to say, this is it. You still have to kind of manage it. And that example I just said, it might be that, and not that AI is going to be wrong, but AI will say, here’s the trend. Here’s where you should be going.
And I’m always skeptical a little bit. I mean, I am really an AI proponent and advocate. But I remember when I started out in the 1980s and 90s in AI, we developed autonomous vehicles. I was working with autonomous vehicles, robot cars.
And the robot cars would ride down the road and that was really cool. It was riding down a black asphalt road. And it’s going around the curves and everything and it stops at a stop sign. And then it turns left at a shadow of a tree that’s crossing the black asphalt and it’s AI software said, black road. It looks like a black road. So we’re turn left, follow the shadow and smack right into the tree.
So AI has progressed a lot since then, but you still have to be careful of some of the aspects of it. The one thing is we can solve a lot of these problems, these trends a little faster with such software and even the banking system is getting involved in this. Do you realize that? Even the Federal Reserve Bank of Philadelphia for example. I ran into them the other day. Even they sit down at their conference tables and talk about how AI is helping the pandemic, mainly from their economic aspect because they want to make money.
Dr. Wanda Curlee: We hear about the vaccine and that is coming. Depending on who you listen to, it could be weeks. It could be a few months. But it sounds it’s coming fairly quickly. So do you think AI could help the research scientists determine who gets the vaccine first? I could argue that young should receive it first or I could then flip it and say those over 60 should receive it first. So will AI help those outside of the healthcare providers to determine who should get the vaccine first?
Dr. Oliver Hedgepeth: Now there’s a good ethical question for you. Ethics is involved. The doctors face these dilemmas all the time. That’s a real good question in fact. Probably one of the best questions that you can ask.
If you’re a physician and you’re a doctor, you’re already using AI software, smart machine software. You’re doing testing and the test comes back from an automated robot type system that says, okay, this patient who’s 95 years old and here’s one 65 and here’s one 35, here’s one 15 years old.
And you got to make a decision about which one might live longer and the AI system would say, here’s the one we choose. And the AI says this, “I choose this one.” The AI system might just choose to let the 15-year-old kid die.
It’s an ethical decision. And that’s where I come back to saying always. AI, it’s not a robot, it’s not something out of a science fiction movie, it’s a helpful tool. It’s a tool like a pencil or an eraser that writes down a number and then can erase the number. But it still needs a lot of human aspects.
There will be doctors who use the AI system to help them make a decision. In my lifetime, I’ve lived long enough to stand by the bedside of a relative and the doctor says, “Do you want to unplug him or not?” And the decision was mine to make, the decision was my wife’s to make.
And the doctor, smart as they are, they don’t make the decision. They tell you all the things that they say, this person’s basically dead.
I’ve lived long enough to have the examples of where the person was disconnected from the machine and everybody’s waiting around for them to die and say goodbye and say prayers. And all of a sudden their eyes pop open and they’re like, “God, I’m hungry.” It is amazing. We don’t know everything there is to know, unfortunately, about medicine.
But it is an ethical issue. I do expect that in a volume type triage, and what you’re talking about is really triage. If you look at MASH, which is a funny movie, it’s really real in terms of the decisions they make.
You get 20 combat soldiers and the doctors are going by saying, this one, this one, this one. Or you would look at the movies about Pearl Harbor and our soldiers are attacked. And the doctors, the nurses are trying to figure out which ones to bring in the hospital and which ones not to. And I remember decisions being made by nurses and them picking their lipstick and put a mark on the person’s forehead.
And basically that said, “Anybody who sees this, you make this person feel comfortable. You tell him he’s gonna live. We’re gonna take care of you.” But you know he’s going to be dead in about five minutes. It’s a triage type of question. And yes, I don’t see an easy answer to it.
Were in a combat situation now. It’s just, it’s unbelievable number. And that’s the triage environment.
The hospitals are still overwhelmed. The doctors and nurses are working non-stop. God bless them. I mean, I have nothing but respect. I have respect for soldiers, military people. But let me tell you, these doctors and nurses are like the military soldiers on the front line and I have great respect for them and they have to make these decisions. And I read a report from a nurse and it just breaks your heart to read it. She has to make a decision every hour of the day about someone to die.
Dr. Wanda Curlee: Yeah, it’s a tough field to be in right now. So, Oliver, we’ve talked a little bit about the different COVID treatments. Why do you think we have so many different treatments this time and how will AI help doctors to determine which are the best treatments for which ages and depending on symptoms?
Dr. Oliver Hedgepeth: This is a disaster. The pandemic is a disaster. But there’s a good thing that’s happening and it relates to your question immediately.
We have seen an increase in research that normally takes, usually takes three to five years. But I do believe that because of AI, that’s linking all these databases together, databases that had never been linked before together. Databases of doctor’s results, of scientific research on animals that’s going on at various universities and various research clinics that are trying out different chemicals.
I do believe because it’s 24/7 research on this and the data has been shared so much. People aren’t working faster, but the data is being analyzed faster.
This is the first time this has ever happened because this is a crisis and it’s a war zone. It really is a war zone. We’re under attack, the whole country and it really energized people. I’m a chemistry major from my undergraduate days, the three-dimensional molecule that this disease, COVID-19 looks like, they found out what it looks like instantly with a few AI things. They look at different chemical with how to model. They found it, the model exactly. Here’s what it is.
And they’ve been trying to study and how to break one piece of that little thing off and punch something in there that will stop it from growing when it gets into near the cell, the protein cells inside the body.
The AI started that. They know what it looks like. But how to break that one little piece so it doesn’t work is what’s going on. And AI is showing how that’s been done in one lab and it didn’t work. Another lab, it didn’t work. Another lab, it didn’t work. Another lab, it did work.
But AI again, is coming in that realm too. Instead of waiting six months, it’s projecting ahead, given similar tests that have taken place in the past many years, decades, here’s what might happen if you do this test, this test, this test, to come to a positive result.
Now again, I think the key thing, it’s like our human brain. AI is taking our brain and allowing us to think a million times faster, really a million times faster. And a lot of the data we’re talking about is really millions of pieces of data being analyzed per second.
Dr. Wanda Curlee: Yes. Yes, it’s amazing. So Oliver, when I was doing some research for this, I ran across an article about a 14 year old that entered some, it’s a science fair type. I can’t remember. But it was sponsored by a company. And she actually found a protein that none of the researchers had found that binds to the COVID and stops it in its tracks. And my son and I were actually discussing that and he said, “We don’t give teenagers enough credit for their ability to think out of the box.” So I wonder if maybe a cure, which I think AI will help with, will come maybe from teenagers. What are your thoughts?
Dr. Oliver Hedgepeth: All right, you got a good point there. A very good point. I am a scientist. I’m a chemist, I’m a physicist, I’m mathematician, very boring stuff. As a scientist, I follow procedures and rules to the letter and sort of these doctors and they’re producing this data for the AI system to rethink faster.
But a lot of solutions have come from thinking outside the box. I do believe penicillin was found by an accident too, if I go back. It was an accident. Thinking outside the box is important parallel with going through the normal procedures.
There are some conventional approach where you do multiple tests and do testing and then you inform patients and they’re admitted and then they get symptoms and you get a recovery phase and then you retest them again until you get to a positive result or a negative result and AI kind of jumps out ahead of that. So you’re right. It is going to be exciting.
Dr. Wanda Curlee: Yes, yes we have. So, Oliver, I want to switch gears a little bit. We’ve been talking about COVID and healthcare. But let’s switch over to what the pandemic has done to economic stress for many individuals in many countries. How will AI help this or will it?
Dr. Oliver Hedgepeth: Well, it’s interesting. I’ll start first with America and then move over to London because I’m talking to people in both. The Federal Reserve Bank of Philadelphia just held a meeting, I think a couple of weeks ago, I was reading about it.
I was looking up some pandemic AI material and wound up listening to a bank. They talked about the current COVID-19 and AI applications and how it’s solving many life-threatening problems. But what they started talking about was the economy. And the economy has seen increasing automation in industries. They were talking about how the AI stuff is spanning manufacturing and the food service and office work and about the COVID-19 pandemic, they said, “It’s ensured really that we’ll be transitioning to using more AI at really lightning speed.” And they were using the word lightning speed.
And what we’ve seen in these last, I’d say five, six months, is a lightning speed increase in thinking about the economy and investment. So the bankers are thinking, now here’s the banking side: The bankers are already, and the Wall Street people, are already thinking about how to invest money into certain companies that may be flourishing. Could be companies that are developing these vaccines or could be some kind of food or whatever that we need.
We’re seeing companies that are failing during this pandemic. The economic crisis is killing a lot of companies and it’s happening over in Europe. I’ve been working with a group in London on the economic circularity, circular economics, economic cycles that are going around where you don’t want to waste anything. And this pandemic is causing people to think about waste.
People don’t have money to buy food. So if we can develop food and other things that aren’t wasted, you’re using your money wiser. So it’s kind of a side effect happening here. Companies are thinking how to save money by not wasting anything.
For example, the plastic water bottle that we drink water out of, the purified drinking water. There is a company that will take the water bottle and actually 100% recycle it to be used in various other items that need plastic. It never goes to a waste area. It never goes to a trash can. And people are starting to think like that in Europe, especially, and a lot of places over here in America too.
So I see a lot of things that are happening on the business side, the economic side, and it means related to how the pandemic and AI is helping pandemic. AI is also helping them think through and identify where can you now use ground up plastic and some other clothing material item? Or some other food you can ingest somehow? They’d taken food or redo it, regrow it? How to make it survive against viruses that would take a corn crop down? Or how to find a virus that would kill a hog or turkey? You’ve got turkeys that one little disease and then 10,000 turkeys had to be slaughtered. And so it being used in many, many other areas. It’s just so exciting, that economic sense.
Dr. Wanda Curlee: So it sounds through this circular economy that maybe while some businesses are dying off due to whatever they can’t, they go bankrupt or whatever. But it sounds other businesses may be created because people are thinking more about how to be more fruitful and use things without throwing it away. Did I kind of capture that in a sense?
Dr. Oliver Hedgepeth: Yes. Yes, you did. The circle economics part is all part of what we call reverse logistics. Reverse logistics is reusing something. You throw something away, you don’t necessarily throw at a trash can, you just reuse it.
For example, I’ve got a coffee maker and I’ve got a dehumidifier at home and I’ve contacted a company and they want both of them because they want to clean them up and resell them on the market. They are not going to be torn apart and go into the trash and torn into little pieces of metal and plastic to go elsewhere. They’ll be resold on a market. The market might be in Mexico. Might be in South America. We’re finding that’s what circle economic thing is all about. And AI by the way is really helping out a great deal.
So I am so excited to see that this technology, it’s being used, but it’s behind the scenes. You don’t really know that your life is being driven and helped in many ways with AI.
Dr. Wanda Curlee: So I recently read an article that AI is helping small businesses during the shutdown. How do you see AI helping small businesses and large businesses for that matter?
Dr. Oliver Hedgepeth: Wow, that’s a good one. And it’s really exciting because AI is right here in front of us on our computers. The AI software is you put in your job description. You say, I need a job and bam, the software that’s grabbing your background information is AI software. It’s a smart machine. You’re looking for a certain job and it knows where to go look in different companies for your talent.
Your resume now is, you might say it’s shorter, you can write it physically shorter, one page. Here are the key things that are about me. But you’re already on the system. A lot of people apply for jobs don’t know is I used to work for AT&T. And that was before all this AI got as sophisticated as it was.
But at AT&T, we already knew who you were when you applied for our jobs. You applied for a job and you say, here’s who I am. Blah, blah, blah. At the time, we did have you come in, the best person come in. And you’re sitting across from a person with a computer screen and you don’t see the computer screen. But the AI software is pulling up everything you ever did from the day you were born. And they’ve got it right there and it’s legal for them to get to find all this information. And it’s being pulled from sources you didn’t expect. And today, it’s even more sophisticated.
All these companies know who you are. I talked to the Federal Reserve Bank now and then because I got some friends there, they know who people are. They want to hire them. They know because the background was already there. There’s enough information from Facebook and Chats and everything you’ve done. It gets pulled all together. So it will help you find a better job for you searching and the job may find you.
Dr. Wanda Curlee: Kind of winding this up, Oliver, do you think in the long run will AI help us through the economic stresses that we’re experiencing today?
Dr. Oliver Hedgepeth: People will lose jobs. But those people will be given other opportunities because what we’ll be doing with AI, AI is not getting rid of our jobs. AI will become another tool. It’s like this computer we’re talking on. There was a time, in my generation, where we were fearful of these computers, thought they’re going to take our jobs away. And they did. I remember looking at one of my textbooks that showed 100 accountants in a room, photograph, 100 accountants. Each one was a man. It was a man with a coat and tie on, had a machine, had pencils and paper and they were doing accounting, taxes or something. And the IBM computer came in. And a few years later, that same room is an empty, empty warehouse with one person, a woman, sitting behind the computer, typing away, doing all those accounting things that those 100 men used to do. They lost their jobs.
The thing is, that person behind the computer got a job and got a better paying job too. Technology will replace people. But you’ve got to understand how technology changes. This is normal. But in the end, we are better off. AI is better. Look what’s happening with AI and this quarantine stuff right now, it is better.
Look at when you go shopping. Right now, you can go shopping, sitting at your desk. You can call up the grocery store and there’s a young person who will be picking out your vegetables for you and your cans of soup, put in the cart and have it out front for you or drive it to your house.
Things are getting better. That’s technology is helping him to do that as well. So it’s going to be an interesting ride and I do believe we’re going to be better off. But I do believe the AI system, the AI technology is not going to go away. It’s going to make our lives better.
Dr. Wanda Curlee: So Oliver, thank you very much for joining me today. Do you have any last parting words?
Dr. Oliver Hedgepeth: Yes. AI technology is no different than any technology. We started out having a saw that was a handsaw to cut wood. And then someone invented electricity and they figured out how to put a chain on a piece of metal and call it a chainsaw. And you can cut a tree down in five seconds versus five minutes.
Technology has always been with us. And I encourage everybody to look up “how the pandemic uses AI” or just look up, “pandemic and AI.” And I did this morning and I found 7,230,000 hits. So it’s here. Thank you very much for this session.
Dr. Wanda Curlee: And thank you to our listeners for joining us. You can learn more about this topic and similar issues in artificial intelligence by reviewing other blogs and podcasts. Stay well and see you soon.