Arenal Haut: It’s so nice to meet you! I would love you to start out by sharing a brief introduction, both about yourself and about NOVID.
Po-Shen Loh: So I am a math professor at Carnegie Mellon University, but I do a fair amount of work involving coming up with ways to align incentives so that large scale groups of people can get what they want in a way where everybody wants to get there. This is called game theory. These days, I work on this mostly in the education field. However, during 2020 and 2021, I was working on this in the context of trying to find ways to help people control pandemics.
I should emphasize right now. There's an app, but the important thing is not the app itself. It is the technique, because we invented a fundamentally new way to fight pandemics. There were lots of apps to control and contact trace. They tried to control the spread of pandemics, and every other app was roughly the same. It would tell you you were around somebody last week, and today, that person is now sick. So it asks you to please stay home so you don't get anyone else sick, because you might already be sick. Unfortunately, with that kind of an app, the incentives don't align, because if you think about it, it's asking you to please do something to inconvenience yourself to help society as a whole. Because it's telling you after it's too late. You've already been exposed, so you might already be sick. When you have things like this, the compliance rate is difficult, because you have to either find a way to enforce it or you have to cross your fingers and hope everyone's going to be altruistic, which unfortunately, is not quite true. It's just very hard to tell people to do things that are net negative to themselves in order to help strangers.
We're using basically the same framework all these other apps had, but with a twist. Our app will “ping” you (ex: Ping! 8), which will mean somebody just got sick, and they spend time with someone else (person #2), who spends time with another person (person #3), and so on, and it estimates that you are 8th in line to get the disease.
Why do we care? Why does anyone care? Imagine what would happen in a future outbreak if you had this information. It’s optional that you do anything with the information; you don't have to do anything if it tells you. Suppose you get Ping! 8, and then a few days later it goes Ping! 6. A new person got sick, and that person is 6 frequent relations away from you. And then a few days later, it goes Ping! 5. And then Ping 4, and Ping 3. Each of these is a new person who got sick, and now the disease is walking along this chain of transmission towards you. Few days later it goes, “Ping! 2.” How would the average person react? They might change their behavior.
Now, the important thing to think about is why are they staying home? Are they being motivated by the fact that they're just trying to save strangers? Why are they suddenly running and hiding? Who are they trying to save? Themself. So we devised a system of controlling pandemics that's powered by the desire to stay alive. At that point, this is a new non-pharmaceutical intervention that reverses the entire direction of action. If you tell people you are probably going to get exposed if you keep doing what you’re doing, that's very different information than saying you've already been exposed.
So that's what we came up with, and the beautiful thing is, what could this do to the spread of a disease? Well, suppose there is a disease that is spreading, and somehow, the people that anticipate getting sick isolate themselves. What does that do to the spread of the disease? It cuts it off. It's a completely new way of isolating the spread of the disease. And I will never say it fully isolates, because there's always leakage. Nothing is perfect. However, the beautiful thing is, you can still wear masks. You can still try to develop vaccines, develop cures, wash everything right, and disinfect. My point is that you want to pack every single intervention you have, and that will all contribute to reducing the arc of the disease. What I just explained to you is a completely new non-pharmaceutical intervention that was previously not available to the human race. But it is now. You basically would have automatically, self-modulating areas of extreme caution. What our technology would do is that the right area locks down because they want to.
AH: So what inspired you? What inspired you to start this project, and what prompted you to make this big shift from math and education projects to the public health sector?
PSL: It's a gigantic step, and the reason is because I am a Hertz Foundation fellow. The Hertz Foundation is an organization in the US which has been around for about half a century, and every single year, they hunt for about 15 extraordinary people in science, technology, engineering, and mathematics who are about to start their PhDs. If you get picked, you are offered a 5 year scholarship to study for your PhD, and you don't have anything else to worry about during those 5 years. The only external trade off is that you make a moral commitment that if there's ever a moment of national emergency, you promise you'll come and help.
I'm a Hertz foundation fellow. When I was picked in 2004, I didn't quite know what we would ever be mobilized to do. Because what does it mean? It's a moral commitment. You don't sign anything saying that they'll grab you from your house, and you know there's none of that. It's just that you agree that this is something you care about.
But then COVID happened. I remember the date. It was March 14th, 2020 (mathematicians remember that date very easily), when the Hertz Foundation community got an email from someone highly respected within the community. It was a very long message, strongly encouraging everyone to drop everything and switch to work on pandemic control, because he predicted this was going to be one of the biggest crises in our lifetime. And he was right.
So that's what happened. We were optionally mobilized, but encouraged to do so. I was in the position where I had the freedom to do so because I'm a tenured professor, so I can work on whatever I think is worth working on. That's when I decided I was going to rise to the challenge. What could I do? I didn't know what to do for a day, but then by the next day, I realized that network theory, the area of math that I work on, could be applied. This was the first pandemic with smartphones. So maybe I could use smartphones and network theory to come up with a new way to fight pandemics that never existed before.
I also happen to work at Carnegie Mellon. We have amazing computer science students who can program anything. So I put out a call, saying I have an idea. I want to do something with an app. I think we can save the world. Anyone want to join the team? And within a week, some of the most extraordinary people I can imagine came out of the woodwork and joined the team. It was so inspiring to work with these people during the height of the pandemic.
AH: In terms of developing the app and working with your team, what were the most rewarding and most challenging parts of the process?
PSL: Well, the most rewarding part is that we were able to work with people who are really, really, really good, so we basically felt that there was no technical challenge that was too hard. We created quite a lot of good technology along the way. It was very inspiring because there were all these people who were just churning up valuable breakthroughs in the sense that other people in the world hadn't been able to do things, and we did them.
But the most frustrating part was that we didn't control the smartphones. Apple and Google did (primarily Apple), and so they had a lot of power over who was allowed to do what. In fact, even after we made an app that worked, it was several weeks before our app could even be distributed for people to download. It was because there were a lot of barriers put up by Apple and Google that if you wanted to make an app about COVID, you needed to be specially approved by Apple and Google, and you needed to bring all kinds of documentation. It was very, very difficult. And that's actually not a technical problem; that's a human-imposed problem. Then, it turned out that Apple and Google themselves were also making something, so they made it very difficult for anybody else to innovate in the space and promoted their own services. So that was frustrating, but that also then made it more interesting. I don't mind challenge, and we did manage to push all the way through. Ultimately, we developed to the point where we were able to come up with something completely new, whereas everybody else was stuck in the existing ecosystem.
AH: Okay, so you talked a little about the barriers of getting this out to the world. But how have you engaged communities to adopt the app, and how do you plan to reach more people? And what has that reception been like?
PSL: That's been really tough, too. In terms of actually getting any community to adopt an app, the hardest part for us was that Apple and Google's system had already been very heavily promoted in all kinds of advertising and all kinds of media coverage. So most people assumed that our app was doing what everyone else's app was doing, and they didn't want that. Our main work was to explain to people that we actually came up with something new. So in the end, the biggest deployment areas where there were people using the app were a few universities, Georgia Tech and Carnegie Mellon University in particular. There were some other places that tried dabbling around with it. But what we found out is that without a very powerful and financially backed marketing campaign, it's not possible to convert.
So that's when we changed strategies. I became focused on getting the idea known in the scientific community, and then ultimately into the media, which ended up taking almost two years. But it was worth it. Because at this point, we don't need everyone to adopt our app. At this point, it seems like a lot of the people who are involved in this space know the concept, so when the next future lethal pandemic shows up, you'll probably notice that there is a number on every app, instead of just telling you too late. It won't be our app; there will be other people. And when you see that change, now you know where that came from.
AH: Do you have a favorite story or anecdote about the app’s impact? Have you seen it in action at Carnegie Mellon while you've been there?
PSH: Okay, so I will not say that I've had much impact, because people weren't that scared of COVID. However, I do know an anecdote that did happen. At one point, an undergraduate student came to me, and he said, “I just saw that my app said that I'm 2 contacts away from COVID. So now, what can I do to protect myself? What kind of mask should I wear?” I thought that was interesting. There's a person who is normally considered invincible actually caring about taking action to avoid getting sick. That's a success. It's only one anecdote, but it's a small victory that somebody who is ‘indestructible’ decides that they want to be careful. And of course that's with COVID. If there was Ebola or something much more dangerous, I would expect that to generalize to lots of people taking action when the disease is close to them, and therefore helping to clamp down on the spread.
AH: What future plans do you envision to expand on this new idea? And then what's next for you, both with the technology, but also with your other projects?
PSL: Well, one of the good things is that NOVID got other people interested in this space. At this point, it seems like there are other researchers thinking about this. So it doesn't need to be me driving it. That said, we are still keeping the app alive. It’s very important that there's always this demonstration that such things are possible. So we have that app running, and at the same time, it's not something that I'm pouring lots and lots of resources into, because we don't have a pandemic at a stage where lots of people are looking to deploy solutions. At this point, unfortunately, most people are just sitting around hoping that the thing will go away. I'm not sure it's going to go away. But in any case, that's not where we're putting our efforts anymore. I decided that I wanted to work on something that can help more frequently than once every 20 years. I pushed the NOVID concept all the way until it hit national newspapers, until it got into the eyes of governments, and until it got to the point where the scientists knew about it, so that this concept would be there for the future pandemics.
But after that was done, I went back to education and used what I learned about incentive alignment. Previously, I'd been trying to help people get better at math, and I was just telling people they should do it, and it was difficult. So after I learned about incentive alignment through doing NOVID, I came up with another way to do education, and it’s actually quite fun! We've come up with a completely new way to align incentives to teach students in 6th, 7th, and 8th grade who find that their math class at school isn't challenging for them. What we do is we teach them how to think, and this was also inspired by the work I did with NOVID. There were like a hundred other teams which all made apps, and they were all the same. There was one way of thinking. Then we came up with something different, and the idea will potentially save the world in the future lethal pandemics. So I wanted to work on helping more people learn how to think, and I decided we needed to make a way for people to learn how to think about and solve math problems. I went with math because math is the logic behind thinking.
So I came up with a way to scale up the teaching style which I use when I teach math. I say, here's a math problem you don't know how to do: think about ideas. So the class starts thinking of ideas, and the job of the teacher is to use the ideas that the students are generating to solve the problem using the students ideas. This is a very different kind of teaching. This is not coming in with a Powerpoint, but dynamically making a lesson based on student ideas to show you that you are capable of solving any problem that you see, even if you've never done it before. This is a very powerful skill. The only problem is that there are not enough facilitators for it. Because if you want to facilitate this, you need to know everything there is to know about middle school math, to the degree that you could teach it with zero prep. So this is a hard skill.
Here's how I aligned the incentives. It turns out there are a lot of people who have that skill. A lot – like tens of thousands. They are the people who are the high school math geniuses. The only problem is, they had no incentive to do this. So here's the incentive alignment: it turns out that there's a lot of them who would like to go to highly prestigious universities, and even if you're really good at math, you don't automatically get into these universities because you need to have other skills, too, such as interpersonal communication skills, interview skills. You need to have leadership. You need to have a personality. And you might have known some people before in high school who are really really good at math, and could have some help on those other dimensions.
So I run two programs. One program is for middle schoolers, teaching them math. The other program is for high schoolers. We work with some absolutely brilliant high school math students, and we hire professional comedians, actors, and actresses to teach them how to be interesting. And the net result is they win. They win big time. They practice these skills by live streaming the best live video math classes ever made. So now I run this international program where we align incentives, and we have enough people to facilitate simply because we came up with a new way that delivers something really valuable to the facilitators.
So that's what I actually work on. Now, I would never have gotten here if I didn't do NOVID because NOVID taught me about aligning incentives. It also taught me a ton about technology and user experience because of the extraordinary team I got to work with. I learned a lot of the skills that they had, and then ultimately was able to build this new project.