26: Andy Coravos — A Brave New World of Digital Medicine

Flux Podcast
25 min readFeb 13, 2020

Subscribe and listen to the episode on iTunes.

Andrea (Andy) Coravos founder of Elektra Labs

Alice Lloyd George: To kick off the year I have Andrea Coravos with us. Andy is the founder and CEO of Elektra Labs, a company focused on a future of consumer medicine that’s both data-driven and secure. The company just closed $2.9 million in seed funding. Over the last few years we’ve seen an explosion of wearables and sensors in healthcare. How accurate are these monitoring tools? And how secure are they from hackers? Elektra is cataloguing these distributed devices and has created benchmarks for pharma companies, doctors and device makers. In this conversation we get into everything from the evolving role of doctors to the rise of digital therapeutics and where the FDA fits in to all this. Andy spent time as an EIR at the FDA, so I couldn’t imagine a better guide to the brave new world of digital medicine. Welcome Andy, it’s great to have you here.

Andy Coravos: Thank you. I’m excited to be here.

ALG: First off — is that an accurate description of what Elektra is building?

AC: That was a good way of describing it. The core thing is a lot of people are more honest with their phones than they are with their doctors. So there’s an opportunity to capture that data and provide a more meaningful example of the lived experience. But what that data looks like can be variable. There are many consumer devices that don’t yet show that that data is meaningful. We’re trying to elevate the ones that do have evidence around their technologies.

Eric Topol is one of the most innovative thinkers in medicine

ALG: It feels like we’re in a much longer-term shift towards the future you describe. One part of the shift has been the rise of citizen medicine. Eric Topol, a doctor and prolific writer in the space says we’re in a “Gutenberg moment.” Just as the printing press took learning out of the hands of the priests and the elite, the mobile Internet is doing the same for medicine. Hence that concept of “the patient will see you now,” meaning that patients have greater access to information and are increasingly in control of their own healthcare decisions. Paired with that is computation. We’ve already seen computers replace doctors for some tasks, but what Elektra really addresses is a second part of this shift — distributed diagnosis. Are we at the beginning? Where do you think we are in this overall shift?

AC: We’re always at the beginning for technological development. There’s the conference talk — “these tools are going to replace doctors” versus what they’re really doing, which is augmenting and shaping and changing the role of the doctor. Today our doctors are probably the most expensive data entry people we have. The amount of time they spend filling in medical records is burdensome and painful. The goal is to allow a high-end human brain to perform at the top of their license and strip out the routine things. Some of the tools themselves are advanced and small, the compute power is good. We have high-end gyroscopes and accelerometers that understand movement. Many of these things aren’t new. We’ve had glucometers for a long time. What’s happening is that they’re shifting in form factor and they’re now able to collect information at a local level without having to send it back to a centralized location to make a decision on it.

ALG: Since the form factor has shrunk so much, can you help give a sense of how many of these devices there are?

The growing number of wearable medical devices [Source]

AC: We just pulled some of the data. About a hundred and sixteen million wearables that were shipped last year. The thing that’s nice is that different tools do different things. Sometimes you need a high end MRI or CT scan. But in many instances collecting longitudinal data is really powerful. For example, right now for Parkinson’s patients they use something called the UPDRS, which is like a tapping test where you’re seeing some of the symptoms of Parkinson’s. If you’re in a study, you’re only collecting that maybe one time a month, maybe a couple of times a year. If you managed diabetes using one blood sugar measurement a month that would be crazy. Your blood sugar is changing constantly throughout the day. If you’re measuring things like tremor or heart rate, collecting this longitudinal data one time is a powerful shift in understanding someone’s lived experience.

ALG: It seems like the major difference is having continuous data and patching different types of data together. Is that what Elektra is effectively doing, decentralized clinical trials?

AC: I have to say, because it depends on the listeners, there are no blockchains. In healthcare, when people say “decentralized” what that means is that you’re collecting the data outside the bounds of a clinic.

ALG: Distributed.

Software-enabled clinical trials

AC: Distributed. The challenge with distributed and ontologies are always an interesting thing in new fields. People have always felt that clinical trials are distributed. You would do them distributed across countries, but they’re still centralized in a clinic. There’s two ways to think about what “decentralized” means. One way is where are the data collected, remote versus at the clinic. If you call in somebody with telemedicine maybe you’re taking a survey or talking with your doctor. The two of you are in separate locations but you’re still at home. The other type is virtual, where you might be collecting data from a sensor that doesn’t have any human intervention. But you could use a sensor in a clinic. Maybe you come in and visit your doctor and they’re using sensors there. Decentralization is a combination of where you are and how the data are being collected.

ALG: And with Elektra you’re not actually touching the data that comes off the wearables right?

AC: Yes. All these wearable data they collect things that you could consider digital specimens. We have blood, urine, stool — genomic specimens. With Elektra we look at a couple of different factors. Often people say, “what tool is the best one to measure heart rate or blood pressure?” But that’s like trying to say, “What is the best type of food?” Do you need sugar? Do you need protein? What are the different tradeoffs that you’re making? We can’t just score food, in the same way you wouldn’t just score a wearable. We look at accuracy, usability, can you wear it in the shower, what’s the battery life. People often stop there and miss two other major components. All these tools are connected to the Internet. Anything connected to the Internet is not if it gets hacked but when. So how do you handle the cybersecurity components of these sorts of technologies?

ALG: Yikes.

AC: And the data rights around them too.

ALG: You’ve actually created a database of these digital biomarkers, the Atlas platform? I was having a look. It says you’ve got over 650 wearables documented there.

Atlas platform

AC: Yeah. Some of them are biomarkers, some of them are digital measurements. But effectively what we make is a label. The way that you would have a nutrition label for food that breaks out the different components, we make labels so that you can compare each of these. Right now we work with pharma companies that are collecting biometric data remotely. Long term as the FDA starts clearing more of these tools, we’ll likely need pharmacies distributing and handling these things. With pharmacies today you have something called a formulary, which is how you keep track of all the different drugs that are in the pharmacy. But we don’t yet have a formulary of these types of connected technologies.

ALG: You’re talking about digital therapeutics?

AC: Yes and no. This is where the whole world starts getting interesting. There’s two ways to think about what a digital tool is doing. In a clinical trial you have a drug and that’s doing an intervention. Then you have things that measure whether or not that intervention is successful. Those are called “end points.” So you might test a drug and see how it impacted blood pressure. Today what we spend most of our time on are the measurements. Our things are not intervening, per se. You would still have a traditional intervention, but we measure that digitally. There’s a whole class of things where the software itself is intervening, which is mind boggling. Those are digital therapeutics.

ALG: What’s an example of a digital therapeutic?

Adam Gazzaley, UCSF

AC: An example — and I’m biased because I worked for one of the companies that was building one — is a team out of UCSF led by Adam Gazzaley. They were looking at different ways to change and shape the brain. They developed a technology that looks and feels like a video game. Kids play this technology for about 30 minutes a day. It’s a 30 day protocol and reduces their ADHD symptoms. Instead of taking Ritalin or Adderall you play this game. The fundamental difference is these types of technologies, they can put them into clinical trials. The thing that I’m excited about is that a lot of traditional drugs that we have today just suppress symptoms. For digital therapeutics it’s looking like the effects are longer term. If you look at the kids six months after they have done the therapy it’s still 80 percent as effective, even when they aren’t doing the therapy anymore. It’s restructuring part of the executive function in the brain, which makes it more long term and stable.

Akili Interactive

ALG: That requires a radical shift. Just taking your example, people really don’t think of video games in that way. Digital exposure to things that can rewire the brain is a new way of thinking about healthcare, and obviously in terms of measuring and prescribing those treatments it is very different. What will a pharmacy or doctor who prescribes these types of therapies look like?

AC: The measurement might be the same. You might want to measure a digital intervention the same way you’d measure a drug intervention. Some things do change — it’s pretty new to think about a tech company fundraising like a pharma company. That company Akili Interactive has raised and run clinical trials like a pharma company would. They submit their evidence to the FDA. Software is considered a device so they submit their evidence to the devices group. They might be one of the first prescription video games.

ALG: I want to try it. And what about pharma companies? How are they responding and are they getting into digital therapies?

AC: One of the challenges in the pharma industry is that are fewer blockbuster drugs. They have to think about new ways to develop their tools over time. There’s a lot of pharma companies that say, “Hey we have really good drugs. But they’ve failed. We think it wasn’t that the drug failed but that it wasn’t measured well.” A lot of them are looking at decentralized clinical trials to collect more longitudinal data using digital tools. That’s probably the most popular way they are using it. There are also pharma companies that say, “Hey, should the intervention itself be digital and should we start building out an entire asset class around digital therapeutics?” Companies like Novartis have invested a lot there. They’ve been working with Pear Therapeutics. It is a whole new asset that they are looking at.

Pear is a prescription digital therapeutics company

ALG: Pear was the first one to actually get FDA approved?

AC: They are the first to get FDA cleared. So there is “cleared” and “approved.” They’re two different regulatory pathways.

ALG: You spent time at the FDA?

AC: I did. So now that the FDA has more software they’ve been thinking about how to regulate it. They started to hire software engineers. I served as one of the engineers on their team around software as a medical device.

ALG: What was that experience like? The FDA is thought to be quite slow and bureaucratic to approve medical devices, but it sounds like they are being proactive when it comes to digital and health?

AC: Move fast and break things is great for tech. Move fast and break things in healthcare means people die. It is important to have a governor and a system that impacts people’s lives. The question is, being too slow also kills people. If you have an opportunity to bring something to market and it’s too slow, that’s also harmful. Figuring out what is fast enough but also the right governor is important.

ALG: Do you think they’re going fast enough? Or too slow?

AC: The answer is always both. In some areas I don’t think it’s moving fast enough and in many other ways it’s shifting. I’m impressed with how the FDA has taken a position on software. If you think about drugs, you’re not going to tinker and change a drug over time. But a software company might ship multiple updates over the course of a day. Software needs to change. If you make software “stable” and you don’t ship updates, you’re going to end up with Blackberrys for your major tools.

ALG: I don’t want a Blackberry pacemaker.

AC: I definitely don’t want a Blackberry pacemaker. If something has a bug or security vulnerability you want to be able to ship updates. So this is a wild shift for the FDA. They’ve always been product focused. One of the things that they’re working towards is piloting a program called “pre-cert.” The idea is you pre-certify a company and under certain boundary conditions that company can ship updates.

ALG: I feel like there needs to be a separate FDA for all this though. It seems like a different mindset.

AC: That probably would take a literal act of Congress. Given that our Congress is not particularly functional right now, we have to figure out how to work within the current bounds of 21st-century cures and software defined under devices.

ALG: The other thing I want to get into is their increasing presence at DEFCON, a hacker conference that happens in Vegas every summer. Can you share what your last visit there was like?

AC: For people who aren’t familiar — DEFCON is in its 27th year. It is probably the biggest underground hacker conference. Most people don’t bring phones or devices. If you do you bring burners that you don’t mind people playing with. DEFCON has about 30 villages that sit under the main DEFCON. One of the ones that was in the news last year was the Voting Village, where they bought a bunch of voting machines. Then they hacked them and found the voting records underneath and sent them to Congress and said “Hey, fix this.” One of the villages is called the Bio-hacking Village, where people have been looking at things like the security of pacemakers and infusion pumps. There is some traditional bio hacking, but in addition there is a lot around the security of medical devices.

ALG: So it’s hackers demonstrating what it’s like to hack a pacemaker.

AC: Yes. Everything that’s connected to the Internet can be hacked. In tech when you find a vulnerability, it’s standard that you would go and do a coordinated disclosure, tell the company there might be a bug bounty. Then the company ships an update. That generally was not happening with device companies. For instance, pacemaker companies try to maximize battery life. The way that you maximize battery life is you don’t do anything computationally expensive. Turns out encryption is computationally expensive. So a number of them weren’t encrypting their protocols. There were trivial attacks where you could just ping a pacemaker and wake it up into a high power mode and drain the battery. When those white-hat hackers found it, they would report it. But instead of shipping updates the companies were saying, “OK, we’re going to sue you now.”

ALG: Yikes.

AC: The security researchers said, “What? Why?” and the companies said, “Well, you’re tampering with our device. We can’t just ship updates, we’re regulated by the FDA. So we now have to sue you.” The FDA found out and said, “Whoa. One, that’s a myth. You can ship security updates. And two, what are white-hat hackers? Where did you all come from? And can we meet you?” Then they started attending DEFCON.

ALG: And are they pulling the white hats into the FDA now?

AC: That was part of how I got involved. When I had the opportunity to serve they had a role called entrepreneur in residence. The FDA does things like public workshops where they will bring in members of the public or experts to look at different types of devices and they’ve been spending a lot of time working with researchers particularly in pre and post market guidances.

ALG: And what happens to the offenders who are below the bar for what Elektra would deem secure. Do they fail FDA approval or get kicked off a health app store? What happens?

AC: There are certain people who are “offenders” who do things that are bad, and then there are a lot of people who are just trying to figure out what is important and how do they make themselves better. In some instances there’s a floor — you should have a coordinated disclosure program and not sue a security researcher. But above that there are certain tradeoffs that people have to have. What we try to do is articulate what is the floor versus where is somebody on the spectrum of maturity.

The Wall of Sheep at DEFCON

ALG: I’ve always wanted to go to DEFCON, it’s been on the agenda. Though I’m pretty sure I’d end up on that Wall of Sheep, the wall of newbies who’ve been hacked. You said you’ve brought a lot of new people — for people who haven’t yet attended, what is it like? Is it doable?

AC: It’s totally doable. The researchers are some of the most welcoming communities. If you’re someone who can easily get nerd-sniped it is the best place to go. So for people who don’t know, the Wall of Sheep is one of the villages. They put up fake Wi-Fi access points, so they’ll put in things like HH Honors and Starbucks and things that people’s phones and devices automatically connect to. Then they take your email and password and leave it on the screen. I feel so bad for all of the people who were visiting Caesar’s Palace normally.

ALG: And they don’t even know DEFCON’s going on.

AC: No idea DEFCON is happening.

ALG: I guess it’s a good learning curve.

AC: But there’s so many things there that are interesting. This year they had a tamper-evident packaging workshop. With all the doping that was happening they were looking at how people were doing different things to urine samples, there’s lock picking. It’s a lot of people who are testing around the edges and figuring out how to secure systems.

ALG: It feels like we’re soon going to be living in a world where we’re constantly measured and tested. It’s terrifying to think that there isn’t much of a security layer around that yet. You’re at the forefront of it, but this is a massive amount of data. For all the founders who get access to this data and are are applying machine learning to it — how should they think about filtering and getting the right insights? Given the scale is enormous.

AC: Yes. One of my pet peeves is when somebody says that you can de-identify data sets. At this point, there’s a lot of data sets are nearly impossible to de-identify. There certain ways that you might do it from a legal perspective, but if you were to de-identify “my genome” it wouldn’t actually be my genome anymore. A lot of these things are uniquely identifiable for people. I’m uniquely identifiable with 30 seconds of walk data. What that means is not that no one should have access to your data, but just who has access and when and how are you dealing with governance around the accessing it?

ALG: Are you someone that has wearables? Do you personally track things? Are you a quantified self person?

Digital phenotyping — assessing people’s wellbeing based on their interaction with digital devices [Source]

AC: I do track a lot of things. Then there’s things from personal perspective that I don’t track. I do not like tracking voice or anything that’s recording. There’s a lot that you can pick up in somebody’s voice. Researchers are finding that you can pick up early stage Parkinson’s. In a world where we have a poor functioning insurance system and there’s pre-existing conditions, and a lot of machine learning is just predictions, which may or may not be right, I would like to limit my exposure. I love tracking my steps and waking up to a buzzing alarm at a certain time. So there are things that I’ve chosen to share more of my data in exchange for benefits that I find useful.

ALG: Do you wear the Apple Watch?

AC: I have alternated between different devices. I have been testing out the Apple Watch yes.

ALG: It’s crazy to step back and think how well the Apple watch has done. When it came out there was a lot of excitement and also a lot of skepticism. In their fourth quarter earnings wearables was huge. Across Apple Watch, Airpods and Home Pod they brought in $6.5 billion in revenue for the quarter, a jump from $4 billion the prior year. As iPhone revenue starts to dip, wearables are a booming part of their business. And they’ve set a bar for others. What do you think we can learn from the success of the Apple Watch?

AC: One thing I like about Apple is they have made an emphasis on not just privacy but understanding ownership and who gets access. With any company the more you say you’re doing something the more you’re under scrutiny and they’ve definitely had some issues. But one thing I’ve liked is how responsive they are when issues arise. A lot of people in healthcare say, “Oh this had vulnerabilities.” But everything’s going to have vulnerabilities. The question is not how many do you have but how quickly can you patch and update them over time. Apple has been pushing towards that responsiveness.

ALG: Responsiveness is important. I saw you tweeted out something related to this — the Hippocratic Oath for connected medical devices. Is responsiveness one of the criteria there? Maybe you can explain what that Oath is?

The Hippocratic Oath for medical devices

AC: The idea behind the Hippocratic Oath is, doctors take the Hippocratic Oath to do no harm. Should the security engineers, data scientists, software engineers who develop these tools also take a Hippocratic Oath? And what would that look like? There’s a number of things within the Oath that might make sense. One is that no software is built from scratch. We all use third party components. If you have a vulnerability in one of those components you possibly have a vulnerability in the whole product. Listing out the ingredients is an important factor. There’s a Hippocratic Oath for connected medical devices lists out ethical and security considerations.

ALG: It seems like a great initiative and conceptually the right direction. Hopefully companies will start to go along with it. Particularly as more parts of the supply chain are sourced from overseas. What else can people do to help?

AC: One is to stop saying “ownership.” We don’t really own data. You might own your car, you might own your house. But data is infinitely copyable. It’s hard, if not impossible, to own it from a legal perspective. You may have better opportunities around rights and governance. We have good protections around physical specimens like blood urine, stool, genomic data. We don’t have good protections around digital specimens. Historically a lot of health information has been outside of HIPAA. It might be considered wellness data. We need a significant revamp of thinking about protections for digital specimens.

ALG: Interesting. I want to turn the conversation for a moment towards your background. We’ve touched on your time at the FDA. You also spent time in private equity. How did you make those jumps? Tell me more about your path to founding Elektra?

AC: Sure. While I did serve in a government role all of these ideas are mine alone. I do not represent the United States government. I come from a family of innovators and tinkerers. My grandfather was one of the first people to develop fiber optics. We always had things in the house that we were tinkering with.

ALG: How was he involved in fiber optics?

AC: He was at Lincoln Lab and was developing early technologies around it. I had no idea as a kid. We had all these crazy fiber optic flowers that looked cool and it wasn’t until I got older that I realized how amazing the work he did was. We’re a family of tinkerers. I was coding different things in school. I wish I could go back to my 18-year-old self and tell her that she would like software engineering, I didn’t think it was for people like me. I had a server and was making Websites for people —

ALG: People like you meaning?

AC: Like women. To be fully honest. I didn’t even consider doing that when I was in school. I studied economics and was doing a lot of data science work. It wasn’t until I was in a private equity role that I realized, “I think I would prefer to be a software engineer.” I was in my later twenties and decided to go to coding bootcamp, which was one of the best decisions I made. Then I got a full stack experience. I worked at a YC company and then a digital therapeutics company. I was interested in this idea of how do we make better healthcare for people at home, and how do we make safe and effective care? Particularly around the data rights and security side. Elektra evolved out of that.

ALG: That’s an incredible journey. What pushed you to make that jump from sitting at a private equity company wanting to build and start from scratch and learn these tools?

The Crossroads of Should and Must by Elle Luna

AC: I paid an executive coach a lot of money to tell me that I could jump off the platform. I was nervous. I thought, “I’ve worked so hard.” Even when I quit the firm said, “What are you doing? This makes no sense you are going to be competing with 12 year olds. This is not a skillset that you have.” For me it was like that Elle Luna crossroads of should and must. I knew I’d regret not doing it. There’s something satisfying about building. I feel grateful that I now get to spend less time on the services side and get to make stuff.

ALG: Yes it feels good. And your parents are also in health healthcare?

AC: They are. They do more interesting things than I’m able to. My dad’s a dentist. My mom is a nurse. I am abysmal with blood. They think it’s hilarious. They tell me about things that they do and I get cringey and weird. They always wondered how I would end up in healthcare if I couldn’t deal with blood. I figured it out.

ALG: Your horsepower got you there without any blood involved. Well being a blood-splattered doctor is a noble profession, but there are even fewer people focused on solving what you’re solving, so it’s also a valuable impact to be having, and at scale.

AC: I hope so. Why did you decide to do this interview?

ALG: Why did I decide to do the interview? I should back up and say, I met you at a weekend that Erik Torenberg — shout out to Erik — organized in Tahoe. I didn’t know what to expect. There were a lot of amazing people there. I’m grateful that I went. It was both inspiring and fun. You were the one of the people that stood out to me as being awesome, sharp and brilliant.

AC: I felt the same way so this is cool for me.

ALG: The second time I saw you was at Dialog, another great event in Utah. We were sitting in a room with people having debates around different topics. You owned the room. You have a clear framework for thinking about things.

AC: Thank you. Tell that to my first employer who said, “You are so nonlinear of a thinker. We are going to keep you in the box.”

ALG: Linear thinking is overrated. As someone who’s a lateral thinker I’m going to say that. Being able to make the connections between a lot of different ideas, industries and people is valuable. It’s a lot to hold in your head. It’s also something that humans can do that machines can’t do right now. It’s a unique attribute. It probably makes you a powerful CEO and good at your job.

AC: Is now when we talk about Burning Man?

ALG: Ha. It was running through my mind that you must have lots of optic fibers in your burning man costumes?

AC: An homage to the family.

ALG: That optic fiber legacy. The other connection is that you are friends with Rachel Katz — shoutout to Rachel — another talented healthcare founder in San Francisco. She is an old friend from my China days.

AC: Hitchhiking across China in trucks is probably one of the most badass things I’ve ever heard of. And then on the side she made a graphic novel. She is a full on Renaissance woman.

ALG: Truly the Renaissance lady of our times. Sadly I didn’t find her at the burn either. Have you been to burning man many times? What’s your favorite aspect?

AC: Ironically my favorite part is the ability to disconnect. Fortunately I’m healthy so I don’t have to depend on connected tools and when I allow my brain to sit and chill I think about things differently. In the frenetic economy that we’re in and constantly touching our phones, having that space — whether you call it meditation or running or whatever it is — to allow a brain to form new connections is powerful. Because there’s no Internet there you have to.

ALG: During the past burn I put my phone away for eight days. My brain did feel different when I came back and looked at my phone again. It’s hard to explain what that digital detox feels like. Though there’s an increasing number of studies. Am I right that you do quite a bit of meditation?

AC: Yes. It’s an important part of my practice. Once or twice a year I’ll go to a monastery and sit for a number of days. I did an eight day a while ago.

ALG: Is that a silent retreat?

AC: It was. I broke it twice. I locked myself out of the room and then I ran out of shampoo. Both of those things I felt like was worth asking.

ALG: It required speaking to solve those things. Where do you go?

AC: It depends what kind of practice people have. My practice has mostly been out of a Japanese lineage called Soto Zen, and so I’ve been practicing at the SF Zen Center. There’s a place in upstate New York. We should go. It’s a three day.

ALG: I would wave at you for 10 days in silence?

AC: No that one’s part silent, part not. We’d have four hours of silence and then could chat.

ALG: Is laughing allowed? I would be laughing a lot.

AC: Buddhist monks laugh so much. They’re so happy. There’s a lot of giggling.

The Book of Joy

ALG: That’s one of the things that surprises me the most about enlightened, mindful folks. I reread the Dalai Lama’s Book of Joy often. It covers the week he spent with Desmond Tutu. You’d expect these moralists and monks to be very serious, but they spent much of the week in hysterical laughter. Which is wonderful and such a validation of all the principles they discuss.

AC: It’s a good life.

ALG: It’s a good way of living. I should do a retreat. Though I’d probably be rubbish. I’m not sure I have the discipline for a week of silence.

AC: We’ll go. It’ll be fun. Also there’s no shoulds. There’s only choose to or not.

ALG: Do or do not, there is no try. Or should.

AC: I hate the word “should.” Small tangent. It never goes well. “I should have eaten better, but I didn’t and I chose not to.” Should is not a helpful word.

ALG: It’s also not a word that’s in the present. You’re judging yourself in the past or creating expectations for yourself in the future. Maybe I’ll try to eradicate it from my vocab.

AC: My dream wearable is one that locally processes my voice and says, “Oh, you said ‘should.’” Zzz. And then I just stop it.

ALG: It would zap you.

AC: Yes. Pavlovian.

ALG: And then change your brain wiring. You would stop thinking certain ways.

AC: What people don’t realize is that software really does impact behavior change. You think about the mood experiments that Facebook was doing or the way we see things on the Internet. The attention economy is all brain changing. Software is a powerful tool. It could be used for good or less good.

ALG: Are you a social media and Facebook user or do you prefer to stay off it?

AC: Both are true. I definitely use Facebook and have been putting controls into making sure that I’m using Facebook in a way that is proactive and not just, “Hey, I’m lonely. I’m going to infinite scroll and somehow feel more lonely. Why are all my friends married and have babies?”

ALG: Well I’m sure we’re in for some sort of social media correction to come. You probably know Tristan Harris and Time Well Spent. Are you spending time with any of these attention economy initiatives?

AC: It might sound weird but I think a lot of the attention economy and Buddhism and digital medicine overlap in a meaningful way. If you think about what a digital therapeutic is, it’s just changing your brain. They’re not going to directly lower your heart rate or replace chemo, all the digital therapeutics are brain related. But whether you’re packaging it in a software product or using it to gain ads, all of these things are brain manipulation tools. Some of them might help your brain work better. Some don’t. Following these types movements is powerful and in many ways intersect quite a lot.

ALG: I’d definitely like to see more founders building effective solutions at this intersection. Particularly borrowing older Eastern mindfulness concepts and packaging them for consumers. Have you seen anything interesting in this cross-section and if not, would you build it yourself?

TED talk by Judson Brewer on behavior change

AC: One of my favorite researchers is Judson Brewer. Judson is super interesting because he’s part philosopher but also a psychiatrist. He takes old Buddhist texts and then modernizes them into Western language. He takes things like dependent origination from Buddhism and turns them into Western tests. He’s been building behavioral modification tools that stem from Eastern religions and then are tested through Western methods. He has a company called Mind Sciences that’s doing a lot of this work.

ALG: Interesting. Consumer facing?

AC: This one’s consumer facing.

ALG: I’ll check it out.

Andy and her Elektra Labs teammates

AC: It’s great. I’ve done a smoking cessation one, a mindful eating one and an anxiety one. For me the anxiety one was great. I didn’t realize that the ways that I was using social media — when I felt sad or lonely and wanted to scroll — is in many ways how people use drugs or use food. That helped me catch that trigger beforehand and shift my behaviors. So that has been a meaningful product for me. Digital therapeutic.

ALG: Super interesting. There’s much more I want to talk about but we’ll have to wrap for now. It’s been such a pleasure Andy, thank you.

AC: Thank you.