ACM ByteCast - Shwetak Patel - Episode 7

Episode Date: November 9, 2020

In this episode of ACM ByteCast, host Jessica Bell welcomes 2018 ACM Prize in Computing recipient and 2011 MacArthur Fellow Shwetak Patel. Patel is a professor of computer science at the University of Washington and a director of a health technologies group at Google. He recalls his beginnings as a computer engineer with an interest in both hardware and software, which narrowed to computing during grad school. They discuss how “smart house” technology and working in construction stimulated his interest in building sensors and how applied research enables his work to have a greater social impact. Patel also offers valuable insights on the benefits of academia as a starting point for innovation, the global implications of his work, and advice to people entering his field.

This is ACM ByteCast, a podcast series from the Association for Computing Machinery, the world's largest educational and scientific computing society. We talk to researchers, practitioners, and innovators who are all at the intersection of computing research and practice. They share their experiences, the lessons they've learned, and their own visions for the future of computing. I'm your host, Jessica Bell. All right, welcome to ACM ByteCast. And today we have a really interesting guest. I will let you

introduce yourself. Please say who you are, what you're doing now, and any sort of fun facts you have about yourself. Great. Hi, my name is Shwetak Patel. I'm a professor at the University of Washington in computer science. I'm also director of health technologies at Google. Interesting fun facts about me is that I was a trained electrician and plumber while I was in grad school, which informed a lot of my research when I was back in the earlier days. So people don't realize that I can do construction and something that ends up being a good conversation piece with people. Oh, wow. Is that part of something that ends up being a good conversation piece for people. Oh, wow. Is that part of why you ended up going into tech? Did you find that through construction? Or is there another? No, not really. I mean, a part of it was, so when I was growing up, my parents were in the hotel business. And so I, I shadowed them a lot in terms of like, you know, fixing things in the rooms and that kinds of stuff. So I was always very, like hands on and, and kind of mechanically minded. So if somebody would have

guessed, they thought I would be a mechanical engineer or somebody. And so, but computing was a passion for mine from the day from when I was little, but it was just something that I like to be hands on. So which you see a little bit in my research, but at the same time, it was just, it's just something that I was interested in, did a little bit of work for Habitat for Humanity, helping build houses. I grew up in the South. And so that Habitat for Humanity is a pretty big organization there. And then I just, you know, like to work on, you know, more, more mechanical things like, you know, HVAC plumbing, it was just like kind of an interesting thing for me. But little did I know, when I was a grad student, that stuff would be really helpful

for my research. So when I did a lot of electricity and water sustainability research that all that stuff actually was ended up being very helpful. Interesting. Yeah. Interesting. That's really cool. I always find it really interesting when people who are in computer science also have a very deep interest in a physical or hand-based either hobby or things that they really love. I really love cooking. And I actually find that there's some relationships between how I learn a new recipe or a new technique cooking and how I learn a new sort of like thought process or solution or something like that. So that's very interesting to hear. Yeah. Yeah. So, okay. So, so if the construction stuff kind of, okay, you're good at puzzles, you like how to like figure

things out with your hand. Tell us a little bit about how you then decided, oh, okay. I actually am really interested in computers and computer processing. And maybe you can tie that into how you came about your grad school research. Okay. No, yeah. I mean, I was always fascinated with computing kind of when I was little. I was fortunate enough to have access to a Texas instrument TI-49A computer where I was doing a little bit of coding on it. My parents immigrated from India, so I was born in the US. But when they immigrated to the US,

they didn't know much about computers. We're lucky that we got our hands on one of those. But my parents obviously didn't have access to computers in India, or didn't have much knowledge about them. So I was just kind of tinkering myself, which is a lot of computer scientists my age have done. It's like, they kind of figure it out themselves. But because I was also hands on, I also like to build them and like rebuild them and then build electronics. And you know, Radio Shack back in the day was actually, you know, electronics components store where you can actually go and buy components and build things. So early on, I was more of a computer engineer than a computer scientist. I was more interested in the hardware side of things. But then I realized the flexibility that software could provide in terms of what you could do on top of the hardware. So when I was little,

I did a lot of that kind of stuff, technology fairs, science fairs. And so I've always had this kind of mixture of software and hardware. But when I went to undergrad and grad school, I decided to go down the computer science path. Because for me, I was like, well, I think computing is really going to be the center of a lot of things in society. And that's what happened. And some's what happened. And, you know, some people would have guessed I would have been a mechanical engineer. Some would have guessed a doctor because I was very interested in more of the areas of like social impact and that kind of stuff. But computing was just, I was drawn to computing.

But throughout my undergrad and graduate career, I still had my hands in hardware too. So right now, you know, I do a lot of work at the intersection of hardware and software. And more and more of it's on the software side, but I really love the intersection of the two, the duality and the interplay between hardware and software. So not picking one or the other. I mean, I could have double majored in computer science and electrical and computer engineering at Georgia Tech, where I got my undergrad and my PhD. But I said, hey, I'm just going to get the computer science degree, but I will take the courses and do work in electrical engineering as well. So over that course of study, I kind of gonna get the computer science degree, but I will take the courses and do work in electrical engineering as well. So over that course of study, I kind of got a knowledge of both, even though my formal degrees are in computer science.

Okay. Yeah. And tell us a little bit about what you focused your research on in your sort of graduate degree career and sort of what led you to being interested in those topics. Yeah. So a lot of my influence in graduate school was actually influenced by my undergrad experience. So at Georgia Tech, I worked with Professor Gregory Aboud, who at the time was doing a lot of interesting research in smart home technology. So at Georgia Tech in Atlanta, we built this home for the ground up called the Aware Home. So this

was back in the day where smart home technology was just at the cusp of kind of being ubiquitous. And so grants from the National Science Foundation and even from the state of Georgia to build a home from the ground up to be a test home for doing research in the space. So this included things like smart home technologies, but also more importantly, elder care, looking at remote patient monitoring, those kinds of use cases. And I really got gravitated at that. Remember, I mentioned that I was really interested in home infrastructure and construction and hardware. And then you've got to build sensors for the home. Oh, that's kind of cool.

I can build hardware. Oh, and there's software on top of that where you can do machine learning. And so that's like, wow, that's the convergence of all three things that I'm interested in. This is awesome. So that kind of spilled into my graduate studies as well, looking at building new sensors and building software to support those sensors to do interesting things with them that have a socially meaningful impact. So it could be elder care, health care, sustainability. So that's kind of where I really got excited. It's like, look, I can build hardware, I can do the machine learning signal processing on top of that. And by the way, I can apply these to socially meaningful problems. And so that's where

a lot of my initial research was really around, how do you build sensors that are low cost and easy to deploy so you can get to scale, right? I mean, if you build sensors that are very customized and expensive, you just are never going to have an impact. So I was looking at how do you build simple sensors where you can use machine learning as the secret sauce to glean more information from it. So one of the early technologies I worked on that also was some of the work I did as a professor at UW was monitoring the energy usage in a home. So instead of putting a sensor behind every single appliance, what are some other ways we can do this that's easier to deploy? And so we figured out a way that you can actually monitor the electrical power line with a single sensor. And you can actually quote, listen electrically to the electromagnetic interference in the power line

to say, hey, you know, that was a toaster that turned on or that was a TV or that was a light switch and use machine learning to disambiguate what got turned on and off. And so at the end of the day or week or the month, you can actually do you can provide feedback to the homeowner or the users about their energy usage. And they can make decisions about how to reduce their overall footprint. So because one of the challenges is you don't really know, like things are out of sight, out of mind. A DVR is running, but you don't know how much power it's consuming you don't so so it's like how do you get that information in front of people to enable behavior change applications and but the challenge was that you're not going to install a thousand sensors in the house and people don't even replace their smoke detector batteries

let alone a sensor that's going to monitor some power uses on that appliance so so the idea was could we use machine learning to kind of get close enough but not it doesn't have to be perfect but with one sensor and so that's where that's where a lot of our initial work around sustainability and leveraging existing infrastructure to do sensing, that's where a lot of that stuff was born, was from our energy work back in the day. Very interesting. So hearing this and hearing your sort of love and interest in the hands-on part of all of this, can you talk a little bit about then why you decided to go into academia instead of perhaps working at a startup that does smart home or building the software or the platforms or these kind of things? Yeah, talk us through that process. That's interesting.

Yeah, I mean, it's a great question. It's actually the number one question I get in my office hours for new students. And just, you know, as I go through my journey here, I mean, if I fast forward all the way, I was able to do all of that. So remember, I'm professor at the University of Washington, I'm an executive at Google, and I also did three startups that were related to smart homes. So I kind of did it all. And part of it was because I was an academic, because I went down that path. For me, academia gave me the intellectual freedom to explore interesting new areas. And hey, you know, if something interesting comes out from a novelty standpoint that I can spin out a company or maybe license it to industry or maybe work with industry on a sabbatical or take a little bit of a leave, like the playground

is my academic role. So, and I always think about the university as a playground to do really fun, exciting, cutting edge work and industry as a way to scale it and get it out there, right? And so, I felt that I wouldn't be hamstrung if I were a professor because I can set the research agenda. I can get students excited about the things I'm working on and then try to figure out how do you get it out there and scale it. So, it was a good starting point because there's a lot of different paths technology can take in terms of how you have impact with it. You can open source it, you can create a company, you can license it. There's a lot of different things you can do. And I think the university provides you that flexibility as a starting point to innovate without question, and then try to

figure out when it makes sense to scale it out. So that's really the reason was I could literally have my cake and eat it too, so to speak. But it was the intellectual playground I needed to start from. Yeah. Do you think that sort of, I mean, academia is not exactly known for its flexibility within that. You think this is a, it has to do with the flexibility of where we are within tech right now, or specifically computer science, or why do you think that you've been able to have flexibility in a historically inflexible career? And then a second part of that question would be, do you think academia is going more towards this type of interdisciplinary type research? Or do you think specifically like, oh, that's just because computer science happens to be in this

interesting intersection? Yeah, that's a great question science happens to be in this like interesting reflection? Yeah, that's a great question. I mean, for me, it was really around. So I'm very impact driven. I was I've been very lucky. My students invite to have, you know, we've got great, you know, we write best papers and like those are great. But for us, for my students and myself, it's we're always focused on the impact. So how do we get this thing out there so people can benefit from it? And sometimes in academia, it's hard to have that impact because our funding models, the

way that our teams are structured, it's hard to have that impact. You don't have quite the engineering team or you might not have the right resources to be able to get it out there, right? And so for me, I was always impact-driven. How do we get this stuff out there? And so I fundamentally just thought about the problem space differently. But I think computing in general is changing because right now you're starting to see more and more intersections of academia working with industry for all kinds of reasons.

One is industry has access to some of the resources that are tough to get in academia. If you think about quantum computing, you need a quantum computer, right? If you think about a lot of the deep learning work, some of the GPUs or the clusters and those kinds of access of high performance computing infrastructure, you can work with industry and you can apply your topics to some industry relevant areas. If you think about access to hardware or really complicated sensors or hardware, you have to work with industry there. But industry also benefits because now because academics are really good at looking at the 5, 10, 15 year vision. Now you have a roadmap that you can provide industry. So it actually goes

both ways. So I think it's the convergence of very interesting technology that's coming out, but at the same time, industry being more and more willing to work with academia to be able to innovate on top of it. The other benefit is that we can innovate in unique ways, right? So industry often have their roadmap and have their business models, but we can start to, you know, innovate in different directions to actually, you know, potentially pivot roadmaps or, or maybe introduce new concepts that may not have been on their critical path. So it actually is can go both ways. And that's how I've looked at it is that, you know, I need to work with industry to be able to scale the work we're doing or, or just stays on the shelf. Yeah, yeah. I'm glad you brought up impact because I think particularly in your specific field,

impact is an interesting question because while we can learn so much with these new like sort of massive amounts of data that we're gathering, whether that's new kind of technology on sensor or new ways of like harnessing and gathering that data. So on the flip side, how do you as an academic think about impact

that could potentially be negative impact? And share with us your thoughts on negating that negative impact and some of the ways that you sort of consider that within your research. Yeah, that's a great question. I mean, the unintended consequences, we've seen a lot of examples of that, right?

And so the way I look at impact is multiple ways, right? And depending on the project, you can define it in different ways, right? You know, I mentioned, you know, open sourcing, licensing, startups, there's all different kinds of ways to have impact. But when I look at impact, it's, you know, how can you translate the work to something that to a point where it can be adopted by industry or society, and it has that benefit. But also looking at impact that you've also gone deep enough to help understand what those unintended consequences and

the challenges are, so that when somebody were to integrate that technology into a product or an industry, that you've created a framework that they've thought about those kinds of areas. So when we do our research, we do think about things like health disparities, inequities, looking at unintended consequences, looking at the security privacy side of the question. I mean, it's hard to tackle all that simultaneously. And in fact, there are other researchers that are way better than I am at that. But if I can uncover, at least make that, you know, and we're not going to hide it, it's just saying that we're going to uncover those things and provide a scaffolding around, hey, there's a research opportunity here and galvanizing a field around it, I think is

impact for us because, hey, we've actually put a lens on something that could be a problem down the road. Now let's have the research community come around, kind of put some effort around that. So I think being able to identify those early on is impact in my mind because you've actually addressed, at least got researchers thinking ahead of any type of major issue that could happen. And obviously, there are things that you just can't predict. But I think that's part of the impact equation is that you're able to tackle the problem broadly enough that you have the both the positive and the negative societal impacts of your technologies.

Right. I'm always curious, too. So I actually don't have an academic background in computer science whatsoever. I have a social science degree. And so I'm curious. I wish that I had gotten into computers earlier so that I could connect some of the stuff that I was doing in social science to. And I think academia is like could potentially be a wonderful playground for the like interdisciplinary mixing of a new technology coming out of a computer science engineering type degree with the context that something like social science brings. Have you seen any of that kind of work? What are your thoughts and opinions on that? Do you think that, like, for example, computer science degrees need to integrate that more? Or, yeah, just talk a little bit about your thoughts with that. Yeah, that's a good question. I mean, so my group in general is very interdisciplinary. You know, we work across the board with collaborators who are not in computer science for that obvious reason, right?

You know, when we think about technology adoption, making sure technology is not going to create a bigger inequity gap, right? And you introduce a technology and it only helps a small portion of the population. We're contributing to that, looking at how technology could, how it gets adopted or societal norms that are changing. How do you adapt the technology to the norms and so and right now you know obviously technology is like advancing so rapidly it's a hockey stick right you talk about the hockey stick of all this stuff right but then social norms are not adapting fast enough right so you've got a you've got your changes that are happening that are relatively flat right but at the same time technology is this hockey stick and you can't keep up so you've got to have this hockey stick and you can't keep up.

So you've got to have this collaborative, you do have to have these collaborative efforts with social scientists and looking at, even in specific domains of things like healthcare, like public health researchers and those kinds of researchers who have a good lens and a perspective on this historically to be able to see, well, what can we do now to be able to have that direct impact? can we do now to be able to have

that direct impact? I always call it an impedance mismatch. There's always an impedance mismatch from the technology being developed to the actual social and societal needs. And you can always force it. Like technology, you can always force it in there and there will be a portion of the population that will adopt it. But I think you've got to figure out that impedance mismatch. And in some of the research that we've done, we try to take that head on where we do the formative work to figure out who can benefit the most from this, focus on technology that will impact the people the most rather than looking at just the early adopters, right? That's the easy thing to do. Look at, build the bleeding head technology, the early adopters will use it, but why not like constrain ourselves from the get-go where it's

like, hey, it's the individuals that need it the most might not have the bandwidth, might have older generation phones. How do we retrofit this stuff? Like focusing on those kinds of things actually can move the needle more, I would think. And so those are the kinds of things we do. That's why a lot of the stuff we do is more what I call retrofit type stuff. Hey, I'll go to a student and say, hey, we got to get this pulmonary algorithm on the phone working on everything that's a feature phone onward. You can't just assume it's going to be the phone of last year, or better, it's got to be 10 years old or better, right? Something like

that, where you've got to be able to have that broad lens in terms of who could benefit from it. Yeah, yeah. Okay. So I'd like I'd like to pick up the thread and change it a little bit. You mentioned your current research group, and you sort of started talking a little bit about the healthcare work that you're doing. So, so let's talk about what you're doing now. What are some of your projects? Yeah, talk about that. Yeah, some of the, a lot of the work that I've done at the university has really been around looking at this wave of notion of remote monitoring, patient monitoring, looking at

care at home, just how to enable a better healthcare experience where, you know, right now the healthcare experience in general is this, you know, you might have a yearly or bi-yearly visit with a doctor and that's the data that's generated, right? And from that data, you have to basically predict, you know, what might happen or what might be the next course of action. But now with sensor technology, you can start to get a better pulse on one's, you know, health and physiology.

So, and then you could actually start to do applied machine learning techniques that could actually, you know, pre-symptomatically detect something's happening. Like that's the holy grail. And so, one of the things we've been looking at is like, you know, it's going to be hard to build all those sensors and people are going to wear something all over their body. That's going to be, I mean, that's going to take some time. But the thing that people have the most affinity to right now is the mobile phone.

You know, there's actually more phones than people in the world, which is quite an interesting statistic. But like if you think about it from a global health standpoint, and just globally, people have access to phone phones have a primary use case of communication and access to information. But if you can bring in healthcare into the phone, you have a high probability of compliance, you have a high probability of access and those kinds of things. So we've been looking at how do you leverage phones, not just for like, you know, oh, here's a checklist for things that you can do for your health, but how do you use the sensors that are already on the phone, which are already ubiquitous. I mean, every phone has a speaker, microphone, you got to talk on it, right? More and more have cameras, more and more have accelerometers. How do you use those beyond for just telephony and gaming for healthcare? So we've had a number of projects where we use all those sensors for all kinds of physiological monitoring. So instead of doing a respiratory measure once a year, or maybe every time you go to a doctor's office, you can do one daily, weekly, you can

be reminded to do one so you can get ahead of any conditions that might be happening. And so that's what we've done. In the last six years, that's a lot of the work that the students have done is turning phones into medical screening tools that can enable people to take the healthcare into their own hands. Right, right. I'm glad that you mentioned the global impact of it, because that was going to be my next question is, healthcare is so radically different if you look at countries all over the world. And it's very interesting to hear that you're thinking about this in a more global thing. What do you think some of the challenges are between applying this technology, depending on where someone is on the globe? What are some of the things that you think about within that problem space?

Yeah, I mean, there's a lot of different challenges there. I mean, one of the things we did was we purposely focused on the global health side, just because that provides the constraint that I think if you can solve that, or at least make progress there, it could have impact domestically as well. I think solving the domestic problem is sometimes harder to translate to emerging countries or developing countries. And so I think solving the global health, some of those I think could apply to everywhere. And so we purposely provided that constraint. And we co-work very closely with Gates Foundation, a lot of this stuff. But I think some of the challenges is access to technology, right?

I mean, yes, there is a significant amount of smartphone penetration globally. But if you look at it, those are two or three generation older phones. And so networking is still a challenge, right? You can't assume that you're going to go to the cloud every time you build an algorithm. So you have to do local compute. It allows us to really think creatively about, hey, look, these are the models we are constrained with. These are the phone specifications we're constrained with.

Oh, and by the way, in countries where air pollution is higher, the microphone is going to get gunked up with dirt. And so these constraints all were part of our design early on. And so we had to solve all those kinds of things first. But also how healthcare is administered is very different, right? In the U.S. versus other parts of the world where, you know, we do a lot of deployments in South Africa and India where you have community health workers who are the front line of defense.

So you have to empower them. Whereas in the U.S., you know, your primary care physician is your front line of defense. So it's just depending on who and how you empower and how that data, like if you have a phone that can do these, you know, this triaging, well, how do you connect that up with a caregiver, right? So, so that, those are the other things that we have to sort out, but we're trying to just get the fundamental thing sorted out, which is what can we glean from the phone so we can get some physiological information about an individual to help them through their journey. Right, right. Interesting. So let's say I'm a student and I'm starting in my

early computer science career, and I'm unsure what's happening with academia versus industry. Kind of this interest in social good, but I also really love the computing side of it. I'd love to hear your thoughts on how people should think about moving through their both academic and or non-academic career. And if you sort of lessons learned from from your movement up through that process. Yeah. So I and I went direct, I would say a direct computer science path, like everything's computer sciences, undergrad, PhD, faculty position, like it's all computer science. I also have a double appointment in electrical and computer engineering at UW as well. But in general, it's all computer science, right? But either through osmosis or

through just reading or whatever, I gained a lot of knowledge in energy and sustainability and healthcare. But these days, there's so many opportunities for minoring and dual degrees and double majors, and even some universities have really unique degrees you can piece together yourself. I think the bridging computing with, as you mentioned, sociology or public health or looking at these other sectors where I think it's really bringing in the public and the individual, the user, just society together in your kind of the way you approach a problem. I think it's just such a unique perspective that I think a student

could be very successful with that. I went down a conventional technical path, but brought in some of the other things. But I think the benefits that we can have for students that have this kind of broader mindset is going to be really, really great. I think computing could be where you can center around. I mean, you could, the thing I always talk, remind my students and just others about is that computer science is not going to solve all our global problems. I mean, yes, we love computer science and computing. It's great. I think of it as the instigator for change. So, it's the thing that can think, allow us to think differently about a problem. It could scale, maybe scale a solution. It can maybe accelerate the development or the implementation of a solution. It can make

things faster or it can allow us to scale it, right? It's the development or the implementation of a solution. It can make things faster or it can allow us to scale it, right? It's the instigator for change and the thing that's going to allow us to scale it. But computer science by itself is not going to solve our societal problems by no means, right? And so, I think people that think at the intersection of computing and these socially relevant problems and how you can actually get it adopted by society is a really powerful field to be a position to be in and i think students can i encourage students to be at the intersection of these fields especially now when when there's a lot of opportunities for impact right now right right right and it brings in that sort of that sort of question within

computer science of diversity both of thought and of background exactly exactly yep sort of question within computer science of diversity, both of thought and of background. Yeah, exactly. Exactly. Yep. Sort of coming to a reckoning with that and where we're seeing, oh, you know, really when something is too homogeneous, whether that's diversity of background, ethnicity, gender, or, you know, what you've learned, your content type thing. So I'd like to hear a little bit about you sort of as a, as someone who's now shepherding in sort of a new generation of what you do to sort of push that envelope. Yeah. People who have been either under, underrepresented support fields or backgrounds that have been underrepresented and how as a, as a topic, we can be better about this. Yeah. That's a great question. I mean, I mean, this is something that this field has been struggling with and trying to figure

out what we can do to just get broader representation in general. Because as I mentioned, we're applying some of these technologies to healthcare and sustainability and social justice. I mean, and that requires really everybody's perspective together coming in. And so the way to do that is to inspire i mean the next generation students and just kids that are coming up to us to look at computing is not just like you know it's not just like just gaming or it's just the internet or communication it's it's this tool now that can be used for applied across the board right i think and so

and luckily we're my group is very applied. And so, we can send this message at the high school level, even the elementary school level. So, my group does a lot of outreach work just because it's got this applied component to it. So, we host high school students, both need-based students that are in programs that might not have a computing program or don't have really access to technology classes, but also merit-based students, students that did really well and mixed them up in a summer program where they do research with my grad students. And so this is high. And some of these students actually end up publishing papers in computer science conferences, which is just, it just tells you the talent and the creativity, even at the high school level. We do a lot of outreach at the elementary school level as well, just to get students thinking about the broad applicability of computer science and trying to

break their mental model of what computer science is or what they think it is, right? And so that's one. The other one is like, I personally do a lot of work in just beyond just the technical side of things, but looking at policy, looking at things that impact computer science and other fields. So I've done a lot of stuff in that space where we want to make sure that policy is advancing so that we take into account some of the advances in computing. And as I mentioned, some of the research that we do, we try to make sure that we, as a principle, try not to create a larger disparity or inequity gap because of the things that we're building. So those are just some areas that we emphasize just practically. And the outreach is something that we've benefited a lot from as well.

So as we're getting towards the end of our session here, I'd like to sort of dive into two sides of a coin, which is the future. And I'd like to know first, or actually, you can answer in whatever order you'd like to. One thing I'd like to know is specifically to your field and the type of researching that you're pushing forward, what kind of scares you about it? What keeps you up at night? What are those big issues that you don't think are solved yet

that you're like, this needs to be solved before this field can really be effective and move on and not be detrimental? And then on the flip side of that coin, I want to know what is the thing that is the most exciting you think about the future of your field specifically? Yeah, I'll to know, what is the thing that is the most exciting you think about the future of your field specifically? Yeah, I'll start with the exciting. So one of the and this is something I

alluded to earlier, is that, you know, I start to use the word using computing and even electrical and computer engineering as tools for what we're doing, right? You know, for a lot of time, initially, in computer science, a lot of it was fundamental research to basically get to the point where we are now. But this notion of democratizing computer science in the sense that, you know, the tools that are out there to build a piece of hardware or, you know, prototyping a little sensor with an Arduino, or even building software, there's so many more tools that we can get more and more people to participate in developing, or at least contributing or even adapting what their work is to bring in a little

bit more computing. It's just super exciting. If you think about right now, during the whole pandemic, we had makerspaces where you had fabrication labs or pivoting to prototype PPE, right? I mean, it's just a matter of minutes notice, computing was able to enable a lot of this stuff, right? Independent of all the remote learning and the virtual classrooms, that was all computing-based. But I think everybody can now do something in that space because we're trying to build tools that are trying to become more and more ubiquitous. I think that's exciting. And that also is the downside of that is that, wow, how do we make sure that there's high quality output out there? And how do we make sure that it's done in a safe and ethically sound way? Those are some of the things we have to think about.

So I think that kind of leads me to my other, the thing that scares me is that, which I alluded to earlier too, is that society is moving at a glacial pace. At the same time, computing is moving very rapidly. So when you have a new technology that comes around the corner, it's so hard to really understand the unintended consequences until it happens. It's like, man, we're so in a reactionary mode right now. So how do we get to a point where we're not just firefighting all the time? So that's the thing that scares me is that computing is moving so fast that we just cannot keep up with the societal norms and the changes and the expectations. And that's the thing that kind of scares me. So related to that is also, how do we think about the education, computer science

education, right? I really think we need to rethink computer science education that has this broader perspective on not just the technical, but the broader societal impact, but also how we engage with industry as well. I think computing, I think we're at a point that we can take a step back and reinvestigate what does an undergraduate degree in computer science look like? What does an actual PhD in computer science should look like? Because I think the way computer science education will look 10 years from now, it's going to be very different. If we look at how industry engages deeply with academia or the duality there and the interplay between the two, I think we just need to take a step back and just double check

that. Yeah. Yeah. Awesome. That was a really interesting peek into the future. Yeah. So, okay, in our last couple of minutes, is there anything that you would want

to sort of have people know about your field? Is there anything that you think there's a big misconception or sort of like an action item you want to call out? I think what my final words are just directed to the students and the young researchers that are looking at computer science in the field. My call to action really is that, look, you know, we have the tools, we have the technologies, the world is in your hand

now to basically leverage whatever you can to, for social good. I mean, this is, we're looking to the next generation to hopefully build these technologies in a way that's going to have meaningful societal impact. And my entire career was all around applied research. And being applied, that's not a bad thing. And focusing on applying these things for what you're interested and passionate about, it should be the goal, right? And so I just encourage students to think broadly about computer science and go where your passion leads you. Great. Well, I want to thank you so much for joining us today. I think it was really wonderful to learn what you're doing, where you came from, and sort

of the exciting things happening in the healthcare tech space. I'm excited to see what happens. Great. Thank you so much. ACM ByteCast is a production of the Association for Computing Machinery's Practitioners Board. To learn more about ACM and its activities, visit acm.org. For more information about this and other episodes,

please visit our website at learning.acm.org slash b-y-t-e-c-a-s-t. That's learning.acm.org slash bytecast. C-A-S-T. That's learning.acm.org slash ByteCast.