她离开 Google，打造可能拯救数百万人的技术｜与 Mary Lou Jepsen 对谈 | EP #142

Today's episode is perhaps one of the most important episodes I've recorded in recent history. It's with an extraordinary entrepreneur, engineer, and designer—someone who is transforming our medical future. She is the CEO of Openwater, an advanced medical technology company that's developing not only diagnostics but incredible therapeutics to fight cancer, mental disorders, addictions, and strokes.

Her name is Dr. Mary Lou Jepsen. You may know her as the CTO of Intel, the director of engineering at Google and part of Google X, and the executive director of engineering at Facebook and Oculus. Along with Professor Nicholas Negroponte, she developed the $100 One Laptop per Child program. She has a bachelor's degree in engineering, a master's degree from MIT, and a PhD in optical physics from Brown. Professor Jepsen was named one of Time magazine's 100 Most Influential People, one of CNN's top thinkers, and one of Forbes' top 50 women.

If you care about your medical future, transforming the world, taking huge moonshots, or understanding what it takes to be an entrepreneur who impacts a billion or more people, Dr. Mary Lou Jepsen has your playbook. She's also an amazing human being.

Mary Lou, I cannot tell you how excited I am about this podcast. You are an extraordinary entrepreneur, technologist, and disruptor, and I want the world to know what you're doing because you're about to change health care for decades ahead. Thank you for taking the time. I want to go deep with you. I want to talk about how you're reinventing health care, how you're using exponential technologies to transform our lives, and, honestly, how you're making the impossible possible.

Thanks for having me. I'm so excited to show everybody what we've been doing through the pandemic because it's been a lot. We're going to unveil some things today that no one has seen. Thank you for featuring us.

You've been on an incredible mission. You've had extraordinary positions around the world. You were at a few different tech giants—Google and Facebook, where you were the head of engineering—and you were the Intel CTO. You were reinventing everything from holography to VR screens. You ran the largest consumer product development effort ever, and then you gave that up. You transformed yourself from that into what? What's your mission and passion today, using all the stuff that's coming down the pike for next-generation consumer electronics—VR, AR, LiDAR—and using the fact that infrared light, ultrasound, and electromagnetics penetrate our bodies?

With the manipulation we can now do, using Moore's law and the exponential reduction in feature size, we can make devices that are tiny. This is why I started Openwater close to 10 years ago. We're going to talk about it today.

Using these principles, I thought that maybe we could affect disease states on a cellular level—kill cancer cells without killing healthy tissue, fix strokes, and address neurodegenerative disease. Now, 8 years into this—it feels like 10—we have pretty strong results, and we're about to scale out and go into production with devices that anybody can buy to push this research forward in a whole set of disease states, including pathogen deactivation, such as COVID and other diseases.

For anybody listening, what we're about to go on is a revolutionary journey into how the technologies that Mary Lou has been pulling together are converging exponentials. This is the intersection of physics, AI, and chipsets that's turning what was once huge, bulky, and expensive equipment into software-defined therapeutics. I talk about the 6 Ds: when you digitize something, it dematerializes, demonetizes, and democratizes. That's exactly what you're doing to billions of dollars' worth of health care technology.

This laser, for example, was the size of a room and cost $1 million. With camera chips in your smartphone, we're able to see blood flow 20 times better than with a multimillion-dollar MRI machine, CT machine, or anything else we can find published in the literature.

It literally makes holograms. It records the phase of light. Here are the 8 camera chips, and the lasers are in there. It records the phase of light because the chips in your smartphone are so small that the pixels are the size of the wavelength of light. That means we can record the waves in the waves of light, and there's information in that.

With this laser, we made this system, which goes into production literally next month. We've already been in hospitals for 4 years with this technology. This is just one of the modules we're getting out to everybody.

If you've ever had somebody who has suffered a debilitating stroke—which is the second-leading cause of death in the world—or if you've ever had anybody with mental disease or addiction, or an aggressive cancer such as a glioblastoma, the work that you're doing is the chance to provide not just treatments but potentially cures for these things.

Potentially a cure. We have results just from this week looking at amyloid microclots with ultrasound at certain frequencies. One of the issues is that the microclots are too big to go through capillaries, so they kill off whatever is close to the capillaries that get clogged up with these microclots.

It happens with aging, neurodegenerative disease, acute COVID, and type 2 diabetes. We're clearing 80% of them and reducing the size from an 8-micron diameter to a 4-micron diameter. That's a really big deal because capillaries are 5 to 10 microns wide. If it's 8, it may not get through; if it's bigger, it can't get through. It's clogged.

So the potential is very strong.

Again, this is just lab work we're doing, but what we're looking at doing is basically taking something like this and putting it behind your knee. You're holding up something the size of a cigarette pack, basically.

Yes.

This has an ultrasound transducer. It's an 8-by-8 array, and we're able to focus wherever we want to using antenna theory. We're able to make resonant frequencies that allow us to selectively attack the microclots, like an opera singer can stimulate or even break a wine glass but harm nothing else in the room when she sings.

I want to slow this down. This is diagnostic-level ultrasound.

Yes.

I want to slow it down for everybody because there's so much here. What you're about to hear is, I think, the most important revolution in health care that we're going to see over the next few years, with a chance to democratize this at an extraordinarily affordable cost.

I know it is. We have to put it on the same track as other things. The 20- to 40-year Moore's-law cycle times are anti-innovation for the things that kill us, and we have to speed that up. We have great technology, but I think the business model we're using to speed it up is even more compelling. We want to bring other technologies into this suite so we can do more collectively, and we'll get there.

I want to tell your story to begin with. You're a brain tumor survivor, right? How much of that is your motivation? Do you mind telling that story? Where does it begin?

As a kid, I was pretty sick and in the hospital a lot. That got me really good at being on time because I knew I would likely end up in the hospital anyway.

I was finally diagnosed after I dropped out of my PhD in physics at an Ivy League school because I was in a wheelchair. I was really sick. I couldn't move half my face, and I was sleeping 20 hours a day. Then it got really bad because I couldn't even subtract. I didn't think I deserved a PhD in physics from an Ivy League school.

I had already worked as a computer science professor and had a degree from MIT. I was at Brown doing my PhD. I called my parents and asked them if I could come home and die. Brown had a medical school, so they let me see the professors there. No one could figure out what I had.

As I was headed out, one professor said, "You've got really bad headaches, right?"

I said, "Yes, very bad headaches."

He paid for an MRI and found my tumor.

Let's describe what an MRI was for comparison, because this is exactly the same size and shape. It's just 10 times more expensive today.

Because of Eroom's law—Moore's law spelled backwards—it's a giant room. I would say it's a minimum of 20 by 20 feet, shielded as a Faraday cage, with a large electromagnet and helium cooling, with a power center beside it. It's the most expensive room in hospitals and also has the highest margin—90% gross margin. It's where hospitals make their profit, so they're charging thousands of dollars for the MRI.

You didn't need one. You didn't know that you needed it.

I didn't know that I needed it. They found it, and Luckily, I only needed 1 operation. I've taken a dozen medications every day for the rest of my life and will continue to do so.

I still take medications. Every once in a while, I put them all in the same bottle and have a look at them, and I think, "I had to really fight to get these." You have to fight for your life often, and it focuses you. When we're here now, what do we want to do with our lives? There may be a positive outcome from that.

You went from recovering from a brain tumor to where? Where did you launch first in your career?

I'd already had a bit of a career, but I went to finish my PhD. Two other students and I got $4 million from DARPA to commercialize our PhD technology.

We started a company called MicroDisplay. We were the first people to make microdisplays for a wristwatch video, early VR, early smartphones, and projection systems. We set up mass production in Richmond, California, just north of Berkeley. Within a few years, we were shipping all kinds of novel devices—something that looked exactly like Google Glass, but in 1998. The software improved, but the hardware was already there.

We had a collaboration with an optical company called MicroOptical. I need to be clear about that: we made the screen part, and they made the optics. I did that for a while and got recruited at Intel to be the CTO of one of their divisions.

I like smaller companies better. I don't like the sharp elbows as much. I like everybody being in the same little boat, helping each other do the impossible, rather than having so many competing goals in big companies. The goal in big companies seems to be having as many people as possible working on something, whereas the ideal for me is something like WhatsApp—$19 billion and 50 people. That's a much more interesting exponential company.

You ended up at Intel, and that was an extraordinary resurrection from surgery to Intel.

Finishing the PhD was a lot, too. Then I left because Intel only had rail-to-rail processes. I ran into the new CEO, who had just arrived, and explained why we could never make silicon as good as anybody else.

All our processes were rail-to-rail. It was either zero or a voltage. To get the best thing for a screen, because we want to see grayscale, you need gradation—a voltage. I said, "Look, we could use anybody's silicon. We're Intel."

He said, "Come into my office."

I effectively pointed out the fundamental flaw in 2 sentences, literally in an elevator with the CEO. I needed a job, and I was happy. I saved them a few hundred million dollars a year on something that couldn't be in silicon, but everybody hated me anyway.

I put my résumé online because I had finished my PhD in physics and had already been a professor. I had taken a break after my master's degree and had been a computer science professor in Australia. I had also worked as a multimedia artist in Germany.

I want to get back to that later. I want to hear about your art career and your music career.

I only got 1 callback, but it was from MIT, which is pretty good. I applied to about 35 schools.

I ended up with Nicholas Negroponte. I had been a student at the MIT Media Lab in the 1980s. I did a master's degree there and made the world's first holographic video system with a team of graduate students. I loved the place.

In the final interview, I was supposed to have 20 minutes with Nicholas Negroponte, the legendary founder of the MIT Media Lab and, later, one of the founders of One Laptop per Child. I started that in parallel and co-founded it with him. I became the only other employee for the first year and basically lived on a plane with Nicholas while we made a prototype of the laptop.

For those who don't know, One Laptop per Child was the objective of getting a $1,000 laptop down to $100. You launched the entire tablet industry as part of that. Netbooks and tablets came after that and became a much bigger thing.

I actually have one right over here if you want me to grab one.

I can see it on the wall. Hold on.

The beautiful green XO laptop.

This wasn't just a stripped-down laptop. It was the lowest-power laptop ever made, the lowest-cost laptop ever made, and the first mesh-network laptop ever made. We wrote the first keyboards in Amharic and a whole bunch of other languages. No reading was required to use it because it was for kids who didn't know how to read.

How many were produced in total?

Millions. We created a multibillion-dollar, nonprofit, open-source project.

The lasting legacy is a few things. We transformed what a minister of education could do for children in another country. Intel and Microsoft nearly killed us. There was a 60 Minutes segment, but they spent exponentially more than we did to stop it. Eventually, they joined us.

We changed the equation. We also showed what a minister of education could do. The CEO of Google cites this—the Chromebook is its grandchild, or perhaps its great-grandchild—in terms of what you can do for education to help children cross the digital divide. It was also useful during the pandemic.

There's a lot of undue criticism of One Laptop per Child, but what you did was extraordinary. When did you get addicted to moonshots? Was One Laptop your first moonshot?

Probably holographic video. I think it was when a Nobel laureate stood up during my first talk and said, "That's poppycock. It'll never work."

It felt like I was being yelled at in front of everybody. It was probably only 2 minutes of insults, but it felt much longer. I was in my 20s. I went back to my hotel room and wasn't happy—probably crying.

I went up to him at the reception and said, "We all know you've done impossible things in your life. If there's an issue with this, could you explain why it's impossible? It's not sufficient to just say it's impossible. Could you explain why?"

I remember talking to my adviser at the time, Steve Benton, who ran the holography group at the Media Lab. He said, "When somebody tells you it's impossible, what it means is that they're a little bit jealous."

Let me set the setting here. You're giving a presentation on holographic video—your first presentation on a research project you're undertaking for your master's degree at MIT as a first-year student—and a Nobel laureate stands up and says it's crazy, that it will never work, that it's impossible.

Steve also said, "It means it's impossible for them." They may have tried before, but you can look at it with new eyes and find new ways through it.

You went on to build that?

Yes. I built the world's first fully computer-generated hologram with micron-size pixels in 1987, which was hard to do then. It was computer-generated on a supercomputer, which looks a lot like NVIDIA now, but it was a Connection Machine, an early parallel computer.

That was your first moonshot. Would you consider One Laptop your second?

I think we really did transform things. People don't remember it now, but the kids do. They're still working in the field, by the way.

These laptops have been working for 20 years because they're so low-power and durable. It's an incredible architecture. I designed it around the screen. Nobody does that. People think the CPU is the brain behind the operation. I'm like, there could be little green men inside the laptop—it doesn't matter if the screen isn't on.

If it responds to a stroke, you can shut the whole motherboard down most of the time and bring it back up in a single-digit number of milliseconds, and it seems like it's on. That's really important because half the kids in the developing world at that time lacked steady access to power.

We also had a screen with better resolution than the Apple Retina display, at the same time, for a $100 laptop. The computers at the time cost $2,000, and you had to buy $2,000 worth of software for them. People forget that, so it was a massive change in the cost structure.

What kind of battery life did it have?

Extraordinary battery life, and the batteries lasted a long time. We were the first ones—I went to BYD back then—for lithium iron phosphate batteries because lithium-ion batteries were burning. Lithium iron phosphate burns at 100 degrees Celsius, and we conditioned it so that it could last through 2,000 charge-recharge cycles, which was 10 or 20 times what normal lithium-ion batteries could do at the time.

There was a lot of innovation. We were talking about dozens of hours of battery life. It would last for a day or 2, but you could hand-crank it because it was so low-power, or a small solar cell would recharge it. We gave those out, too.

After One Laptop, where did you go next?

I started a company called Pixel Qi because I thought I had given up my job thinking that, since the laptop was built, we could bring in somebody who knew about education.

This is where being a woman in tech helps. People thought I knew about education. I knew children because I had gone to school, but it was time to bring in education experts. I decided to help the industry design more interesting things.

I left MIT because I was more excited about what I could do with the multibillion-dollar fabs of Asia. Despite the best postdoc I could get, I might be able to make 1 thing once, but I wouldn't be able to repeat it for a year because of contamination and everything else in the beautiful MIT labs.

I moved to Asia, started Pixel Qi, and made very innovative screen technology as the first fabless screen maker. I made a lot of screens for tablets, laptops, and smartphones, as well as other unique screens.

But you ended up at Intel, and that was an extraordinary resurrection from surgery to Intel.

Finishing the PhD was a lot, too. I left because Intel only had rail-to-rail processes. I ran into the new CEO and explained why we could never make silicon as good as anybody else. All our processes were rail-to-rail—it was either zero or a voltage—but to get the best thing for a screen, because we wanted to see grayscale, you needed gradation.

I said, "We could use anybody's silicon. We're Intel."

He said, "Come into my office."

I pointed out the fundamental flaw in 2 sentences, literally in an elevator with the CEO. I needed a job, and I was happy. I had saved them a few hundred million dollars a year on something that couldn't be in silicon, but everybody hated me anyway.

You had put your résumé online because you had finished your PhD in physics and had already been a professor. You had also taken a break after your master's degree and been a computer science professor in Australia, and then worked as a multimedia artist in Germany.

I only got 1 callback, but it was from MIT, which is pretty good. I had applied to about 35 schools.

I want to get back to that later. I want to hear about your art career and your music career.

I ended up with Nicholas Negroponte. I had been a student at the MIT Media Lab in the 1980s. I did a master's degree there and made the world's first holographic video system with a team of graduate students. I loved the place.

In the final interview, I was supposed to have 20 minutes with Nicholas Negroponte, the legendary founder of the MIT Media Lab. We started One Laptop per Child in parallel, and I became the only other employee for the first year. I basically lived on a plane with Nicholas while we made a prototype of the laptop.

So you created a $100 laptop by bringing a $1,000 laptop down to $100. You effectively launched the tablet industry, and netbooks and tablets came after that.

I actually have one right over here.

The beautiful green one—the XO laptop.

It wasn't merely a stripped-down laptop. It was the lowest-power laptop ever made, the lowest-cost laptop ever made, and the first mesh-network laptop ever made. We wrote the first keyboards in Amharic and many other languages. No reading was required to use it because it was designed for children who didn't know how to read.

How many were produced in total?

Millions. We created a multibillion-dollar nonprofit and open-source project.

We transformed what a minister of education could do for children in another country. Intel and Microsoft nearly killed us. There was a 60 Minutes segment, but they spent exponentially more than we did to stop it. Eventually, they joined us.

We changed the equation. The CEO of Google cites this—the Chromebook is its grandchild or great-grandchild—in terms of what you can do for education and crossing the digital divide. It was also useful during the pandemic.

There's a lot of undue criticism of One Laptop per Child, but what you did was extraordinary. When did you get addicted to moonshots? Was holographic video your first moonshot?

Probably. When a Nobel laureate stood up during my first talk and said, "That's poppycock. It'll never work," I think that was the moment.

I was in my 20s. It felt like I was being yelled at in front of everybody. I went back to my hotel room and probably cried. Then I went up to him at the reception and said, "We all know you've done impossible things in your life. If there's an issue with this, could you explain why it's impossible? It's not sufficient to just say it's impossible."

My adviser, Steve Benton, who ran the holography group at the Media Lab, said, "When somebody tells you it's impossible, it means they're a little bit jealous." I think the other thing he said was that it was impossible for them. They may have tried before, but you can look at it with new eyes and find new ways through it.

You went on to build it?

Yes. I built the world's first fully computer-generated hologram with micron-size pixels in 1987. It was computer-generated on a supercomputer—a Connection Machine, an early parallel computer.

That was your first moonshot. Would you consider One Laptop your second?

I think we really did transform things. People don't remember it now, but the kids do, and they're still working in the field.

These laptops have been working for 20 years because they're so low-power and durable. I designed it around the screen. Nobody does that. People think the CPU is the brain behind the operation, but there could be little green men inside the laptop; it doesn't matter if the screen isn't on.

If it responds to a stroke, you can shut the whole motherboard down most of the time and bring it back up in a single-digit number of milliseconds, and it seems like it's on. That's important because half the children in the developing world at that time lacked steady access to power.

We had better resolution than the Apple Retina display at the same time, in a $100 laptop. The computers then cost $2,000, and you had to buy $2,000 worth of software for them, so it was a massive change in the cost structure.

What kind of battery life did it have?

Extraordinary battery life. We were the first ones to use lithium iron phosphate batteries because lithium-ion batteries were burning. Lithium iron phosphate burns at 600 degrees Celsius, and we conditioned it so that it could last through 2,000 charge-recharge cycles—10 or 20 times what normal lithium-ion batteries could do at the time.

We were talking about dozens of hours of battery life. It would last for a day or 2, but you could hand-crank it because it was so low-power, or recharge it with a small solar cell.

After One Laptop, where did you go next?

I started a company called Pixel Qi. I had given up my job thinking that, since the laptop was built, we could bring in somebody who knew about education. This is where being a woman in tech helps: people thought I knew about education because I knew children and had gone to school.

I decided to help the industry design more interesting things. I left MIT because I was more excited about what I could do with the multibillion-dollar fabs of Asia. Despite the best postdoc I could get, I might make 1 thing once but not be able to repeat it for a year because of contamination and the constraints of the beautiful MIT labs.

I moved to Asia, started Pixel Qi, and made innovative screen technology as the first fabless screen maker. I made screens for tablets, laptops, and smartphones, as well as other unique screens.

You then went on to Google with Sergey Brin and the moonshot factory, right?

Yes. Sergey fell in love with the technology. I was also trying to work on brain-computer interfaces. As Google was starting Google X, he hired the whole company.

You were working on innovative consumer electronics, leveraging Android and many other capabilities of Google. You were doing things Larry Page and Sergey wanted you to do.

I'm not supposed to say what I did there, but you can read about it. There were some very cool projects, including large holographic walls and screens on every surface.

How do you do that, and why would you do that?

They weren't holographic, though. They were flat.

Something I've been doing during the pandemic is thinking that everybody wants a million-dollar view. All you have to do is make the screen at optical infinity, and anybody could feel like they were someplace else when they got home.

A lot of people spent time looking at a screen during isolation. I was at Mark Pincus's home, which has a beautiful view overlooking the San Francisco Bay and the Golden Gate Bridge. It had floor-to-ceiling, 30- or 40-foot-wide windows, and it was a breathtaking view. I thought, "I would love to have that on a screen. Why can't anybody?"

Do you think you could create a large-scale video wall that looks identical to a view out the window?

Yes. I call it my Venice Beach and Alley project. I don't know if I'll get into it.

Can I join you on that one? I love that idea.

I'd love to do it as an art project and then get someone else to turn it into a startup. I have a little studio where I work on things on the weekends just to clear my head. You could flip a switch and be on the surface of the Moon or Mars, or over the Eiffel Tower.

It's much easier if you're at optical infinity. If you close one eye and then the other, you see a disparity—a difference between the views. But if something is far enough away, it's easier to compute.

So you were with Sergey at the moonshot factory. Were you there when Astro was there?

I overlapped with Astro for a while.

And then you went on to Oculus?

Yes. Mark recruited me. He had bought this company for a lot of money, and it hadn't shipped anything. He bought it for the screens and the optics. It was as if he had bought me as a company.

I didn't actually want to go. Some things happened at Google that upset me, but I probably shouldn't go into the details. I loved Google as a culture, and I loved working for Sergey. He was fantastic to work for. Google has done so much for the world in so many ways—the investments and projects they've undertaken. People don't know how much Google has done.

Still, it was lucrative to go to Facebook. My most successful project up to that point had been the nonprofit, so it was quite lucrative to join them.

You joined the team at Oculus and reported directly to Mark?

It moved around. They called me "Game Changer," and I had another title, too. I was supposed to figure out how to change the game from what we were doing.

I knew it was a rocky road. They had just been bought and were trying to figure things out, so I did what I could. I invented some very cool sunglass-display systems and a bunch of other things that hopefully will see the light of day.

You are the screen goddess and miniaturization goddess. How far are we from wearable AR and VR glasses that are light and enjoyable?

It's a matter of will. They've spent $100 billion, which is a lot of money, and it's surprising how little of whatever they have in their labs has seen the light of day, if they've pursued it.

It's not that efficient. Reid Hoffman wrote a book about blitzscaling—when you spend a ton of money, it's not that efficient. Maybe that's what it's supposed to be, or maybe it makes the taxes come out right, but it's a lot of money.

Mark really deeply believes in it. I don't like the idea of covering your face with a giant mask or ski goggles.

That's why I want the million-dollar view. I don't want to wear it.

Exactly. But different people have different opinions. This has been going on since the late 1960s with VR and AR. There have been different waves of it, as well as the work at the Human Interface Technology Lab in Seattle.

You've been at the top of the entire tech stack—at Intel, Google, and Facebook. You witnessed digitization, dematerialization, demonetization, and democratization. When did you decide that you needed to focus on reinventing health care because it was so broken?

When I left Intel in 2004 and pitched at the Media Lab, I got the faculty position. I had been thinking about this, but then I got distracted by the $100 laptop and thought I could get that to work faster, even though everybody thought it was impossible.

When I went to Google, I was supposed to work on health care and brain-computer interfaces, but Sergey said, "We just wanted to know you were creative." He needed me to do other things, and I was happy working on them.

When I interviewed with Mark, his feet didn't touch the ground when I started talking about brain-computer interfaces and what we could do for health care. We had a whiteboard in the room, and I thought, "This is it. He gets it."

Then I started and he said, "You have to fix this VR thing first. I've spent billions of dollars on it."

That was in 2015?

I started in 2016 and left a year later.

What happened? You decided it was time to leave and build your dream company?

It was the fourth company I had started, and I had been in startups for half of my life. I'm good at startups, I think.

When so many people have so many different opinions, you spend a lot of time trying to educate software giants about technologies that aren't their core competency. Their executive management is focused on optimizing click-through revenue for ad sales. It was frustrating, and it was faster and easier to start my own company and build the thing without all the politics.

I'm not saying that critically. It's just the reality. With moonshots, you could call NASA a moonshot, but it was part of the Cold War. The Wright brothers' work was a moonshot. The invention of the birth control pill was a moonshot. These were small teams that somehow did it, and I think it's easier to do it that way.

So it's 2016, and you founded Openwater. Where did the name come from?

Peter Gabriel. He's an extraordinary rock star and human-rights activist. I knew him from my multimedia-art days in the 1980s. I ran into him at a conference and told him what I was doing.

He started calling me every day, saying, "You've got to leave Facebook. You have to do this outside." He wrote an essay about open water, about our thoughts flowing like water and having to take swimming lessons to learn how to deal with it.

It would really change how we interact with each other if we were transparent in all our human weaknesses, virtues, and problems. He strongly encouraged me and kept calling, so I said, "Okay, let's do it. Can I use the name?" He let me use it.

He has sweat equity and is also an investor.

I want to disclose to everybody listening and watching that I am an investor through my venture fund, and I'm a proud adviser to Openwater. I'm totally and completely biased, and I'm sharing this with you because of the extraordinary work that Mary Lou is doing.

Peter Gabriel's greatest contribution to society may be that he pushed you to get the company going.

He continues to do that.

Let's dive in, because the technology you've built and are now rolling out is going to save millions of lives.

Nearly 25% of the U.S. economy is going toward health care expenses. Thirty percent goes to hospitals, another 20% to doctors, and 6% to research and development. Insurance is only 8% or so, but it's huge. It doesn't move forward as quickly as it should.

We need to do something better if we care about people's lives. I don't think we're counting the 55 million people who die every year globally. Can you describe the state of the medical industry today? I want to set the comparative objective that you're about to crush.

The problem is the cycle time. There are some good cures, but about 30% of us are taken out by cardiovascular disease, another 25% by cancer, and neurodegenerative disease takes you out if you live long enough. Then there are pathogens and chronic diseases such as diabetes.

The treatments don't change quickly. It's now 26 years and close to $3 billion for a new drug approval, using the capitalized cost. For a novel medical device, it's close to $700 million and 13 years to go from an idea to developing it and getting it approved by the FDA. That's just approval, before reimbursement and becoming the standard of care, which takes the cost to about $1.5 billion.

Say you develop a treatment for a single rare disease. You spend $700 million, perhaps save some money, and a few thousand people have the disease. What do you charge per patient? The vast majority of humanity can't afford that cost.

What are we doing? Why are we funding this? The big funders of health care R&D—nine out of every 10 health care dollars in the U.S.—are NGOs and governments. They're funding things that may eventually work through trickle-down economics, but the cycle is far too slow.

What percentage of drugs that are prescribed actually work?

For me personally, it's about 20%. I check that because I'm missing part of my brain. I have a pituitary gland molecularly replaced with an age- and sex-appropriate dose.

For most people, do the drugs work or not? Do they cause harm? It's a real problem.

You assume that when the industry prescribes something, it will work.

Clinical trials cost $40,000 to $70,000 per patient, and they take years. For a bigger disease, such as mental illness or neurodegenerative disease, you have to do 10,000 or 100,000 patients. The cost becomes incredible and the time becomes prohibitive.

We have to change this if we want more innovation. We have to leverage the tools of our time. AI and Moore's law are 2 of the big exponentials, and there are others.

Clinical trials are exponentially slower and more expensive over time. That's been well documented. They call it Eroom's law—Moore's law backward.

That's the big problem that I think you have to change. If we can get more data than we've ever had before, it's less risky for a regulator to approve a new treatment or medical device. It's also safer for a doctor and patient to make a health care decision.

Why not collect more data? We're good at crunching data, and we'll learn even more through AI tools. But if you're making a new drug that has never physically existed before and putting it into somebody's body, it's hard to get that kind of scale quickly because you have to go through tests first.

You're using physics, AI, and chipsets not only to diagnose disease but to treat it.

We started in labs like these, developing different designs that modulate the phase of light and sound. Then we built out these carts around 2020 and put them in hospitals. We got great results in sensitivity and specificity.

Slow down for a second. You've been using and miniaturizing ultrasound, lasers, and cameras to see and affect what's going on inside the body.

Yes. We started with large systems that could manipulate the phase of light and sound so we could steer it wherever we wanted in the body. We could interfere with it to create wave structures, and we could resonate with it to selectively affect different cells with different structures, like an opera singer affecting a wine glass.

How big were these systems before?

They were the size of a room. We started with that in 2016.

And now they're the size of a headband?

Yes. We reduced them to this size and cost.

The desktop image you have is essentially the breadboard—the proof of concept that the physics worked.

These are big optical tables that float on air. They allow you to do experiments where you can see the phase of light.

How much did the scale and price change between 2016 and today?

These were multimillion-dollar systems. We went down to carts that cost $100,000 to $500,000, and then we reduced them to this size.

The console was the cart. This is the console that starts production next month, and this is the headset for it. It comes in different sizes. We also have a 6-pack of ultrasound transducers for the body. You can 3D-print whatever you want and strap it to any part of your body.

We envision putting this on the back of your knee to do both pathogen deactivation and cell rejuvenation, as well as amyloid microclot removal. We have some very good research results on that right now.

This is the box for our imaging system. We went from a multimillion-dollar system to a $10,000 system, and at volume it will go to the cost of a smartphone.

What is the order-of-magnitude reduction in price?

At volume, you can treat something for the cost of a phone call.

That's a 100,000-fold reduction in price.

It becomes really interesting when you think about cost structures. There are no shortages, which are a huge problem in medical equipment right now. It's a device with broad disease impact.

Let's dive into the major applications.

Here are cancer cells in glioblastoma. We had some great results with glioblastoma. The problem is that the surgeon can't get the whole tumor out. Some cancer cells hide among the neurons. You can't scoop out all the neurons.

All aggressive cancer cells share a mechanical property that normal cells don't have. It's the definition of metastasis. They have a big nucleus and a small cytoplasm. The nucleus is large because they're growing so fast and dividing their DNA quickly.

We exploit that the way an opera singer can match the frequency of a wine glass and destroy it while harming nothing else in the room.

We worked with 16 different types of glioblastoma, grew them in organoids, and performed sound sweeps over many octaves and rhythms to find the frequencies that killed the glioblastoma cells without harming healthy tissue.

You found their resonant frequency.

Yes. Then we did it in mice. This is the size of the tumor without treatment. These are our top 3 treatments. We gave a 2-minute dose at diagnostic level—lower than the level used on pregnant women and their fetuses in the Western world, where billions of them have received diagnostic ultrasound over the last 50 years.

The treatment destroyed the tumor. It needed another dose on day 5. The best one we tried was a 2-minute dose at a 10% duty cycle and 150 kilohertz, which is the frequency of a fish finder.

Glioblastoma is currently a death sentence. 100% of people do not survive it. I had a friend recently who passed away from it. If you've ever heard of someone having an aggressive brain cancer, it's the last diagnosis you want to hear. There is very little you can do, and from diagnosis it is typically months or perhaps a year before you die.

It's not long. We had some trouble getting into human trials because of safety requirements, so we switched to a different application.

At the University of Arizona, we did a study of 20 people with severe depression—really severe depression. We took fMRIs and saw overfiring neurons in the front of the brain. An fMRI shows the use of oxygen, which correlates with overfiring neurons. We quelled them.

Nearly half of our patients, in this first study and without even tuning the dosages, went into remission from severe depression. What's the best drug doing for us today? Much less.

How long did the treatment take?

Five minutes a day, every day, for the first week—5 days. The second week, 3 days, and the third week, 3 days. They are still in remission.

You could have this device at home. We also showed onstage how we align it because we're focusing sound on an exact place. We take a cell phone and capture a lot of pictures of your face, turn that into a mesh, and register the bone structure onto the MRI. We do that while you're wearing this, so we know exactly where the transducers are and can focus on the right spot.

This could also be useful for addiction.

All addictions. We can see the overfiring. Whether it's wanting a glass of wine in the evening or something else, we can see how to downregulate it.

This essentially leapfrogs transcranial magnetic stimulation, which is also approved for neurodegenerative diseases and other things. We're also looking at treatments for neurodegenerative disease and other mental illnesses.

The results are spectacular. You developed this early on for stroke detection as well.

The stroke-detection system is this unit. It uses an optical laser and high-quantum-efficiency camera chips that ship in smartphones and cost about $1 each.

The second-leading cause of death worldwide is stroke, specifically large-vessel occlusion. Large vessels block more flow downstream. You have a 2-hour window to get yourself to the right hospital, but even in the U.S., only 5% of hospitals can perform the procedure. By law, you go to the nearest hospital, so your odds of getting to the right hospital for the treatment you need are only 5%.

For a heart attack, an EKG is put on your chest to determine whether you're having one. You can't put an EKG on your forehead and determine whether you're having a large-vessel-occlusion stroke.

We created a system and tested it on 151 patients at Penn and Brown in the cath lab—the place where they do the thrombectomy. A thrombectomy is when you snake a catheter up your carotid artery and pull out the clot. It is literally a plumbing problem. Drugs don't work because the clot is too large.

The implications of having a large vessel occluded for too long are severe. If you don't get treatment within the 2-hour window and survive, you may not walk or talk again or have a job. The brain tissue dies.

The device can determine whether there is a clot and where it is?

We're able to see it with high specificity and sensitivity. With AI, we can also see seizures, which have a different mimic. We're looking at capillary blood flow as well because we can see blood flow very accurately.

We funded that for a few years. The vision is to put this device in every ambulance. The ambulance can know that someone is having a stroke and take them to the hospital that can put them in a cath lab. It can call the cath lab while the patient is in transit so the team can set it up.

Doctors know more about their Uber Eats orders than when they're going to get the next patient for life-saving treatment. We can use technology to improve that.

We have finished that level of clinical trials and sent it to the FDA. They want 10,000 more patients. At $40,000 to $70,000 per patient, that's a lot of money.

That's where you get stuck. You have great technology, but the business model is also necessary.

The technology you've brought together is a convergence of exponential technologies—new chips and cameras, AI, and 3D printing. It's the materialization of physics that enables you to see and manipulate what's going on inside the body and brain.

That's going to impact stroke, the second-leading killer on the planet. It's going to let us address glioblastoma and other aggressive cancers that kill us rapidly and don't currently have cures. It's going to enable us to help people with mental disorders and addictions.

All of this happens without harming healthy cells. The tissue samples went through autopsy at Charles River, and unlike chemotherapy, radiation therapy, or even surgery, they found no healthy cells harmed. The cancer cells and the neurons don't have the same resonant frequency.

We can selectively target them, and we can focus where we want in the body, unlike a drug that spreads throughout the body.

What people don't know is that one of your early visions was using this technology to read and write onto neurons—a version of a brain-computer interface without drilling holes in your head.

That's what we're doing. We're writing to neurons now and addressing mental disease, but eventually we can get to thoughts.

I think the business model, which goes to the name Openwater, is that you recently took a large grant from a well-known crypto technologist.

Vitalik Buterin, the founder of Ethereum, reached out. He's a mathematics genius and a very successful cryptocurrency entrepreneur. He's the most visible person in crypto. Satoshi Nakamoto, the person associated with Bitcoin, disappeared, so Vitalik is one of its public faces.

He had a lot of Shiba Inu coin and sold it when Elon Musk went on Saturday Night Live in 2021 because he realized he had about $10 billion worth of it. He donated a lot of it to dead wallets, but he was looking for help with COVID.

He called me and said, "If I could have done anything to help with COVID, I would have dropped everything in early 2020." He said, "I think you can do it."

We started talking at 10 p.m. on Friday nights in my time zone. He would ask really good questions, so I would spend the weekend thinking about them and writing a few pages. That continued until we realized we should take the company open-source. Maybe we could help with COVID and long COVID, as well as many other diseases that have been accelerated by COVID.

A study of 54,000 veterans showed that the risk of neurodegenerative disease doubles if you've had long COVID. The risk of heart failure goes up 173%, and the risk of stroke goes up 164%.

We think we can help with long COVID because we can see blood flow. If you drop COVID into a blood vessel, you get these microclots. Since they're 10 to 100 microns in size, they're probably not making it through the capillaries. I'm just a physicist, but it seems to make sense.

We decided to open-source the company and took a $50 million gift.

You decided to open-source all of your fundamental technology?

All 68 of our patents, all of our software, and all of our hardware are open-source under the AGPL. It continues to be open-source.

I think that breaks the cycle. The average capitalized cost to get a new medical device through regulatory approval in the U.S., averaged over every one completed in the last 30 years, is $658 million. In 2024, it's staggering.

What's really interesting is that 85% of that cost is device development, not the trials. Another 7% is sharing safety data. If you create a platform—a low-cost platform—we can reduce that burden dramatically.

We had those carts a year ago. We've now reduced them to this size and cost. The carts cost $100,000 to $500,000, and these devices cost $10,000, going to $1,000. People can buy them.

Open source is a distribution model, but it's also a trust model.

Vitalik said, "If you open-source this, I'll give you $50 million."

There was a lot more discussion about how we could help with COVID and other diseases. I went to Zug a couple of times, and it was a long discussion.

Vitalik used Shiba Inu coin. That's important. He didn't sell any Ethereum. People had gifted him the Shiba Inu because he was a famous crypto entrepreneur, and he wanted to use it for charities. We're not a charity, but we are open-sourcing all of our technology.

One reason One Laptop per Child couldn't succeed is that we had no way to make money. We sold it at cost, so we couldn't sustain it. If we had added $10, that might have solved the problem.

A for-profit entity that is open-source may be a better solution. I wanted to try it, and I convinced all my investors to say yes at the same time. At the beginning, I literally had to hold the phone far away because they thought "open source" meant "charity."

But it isn't. It was the best business model I could find—10 to 100 times more revenue and 10 to 100 times more margin than any other approach. It feels like Logan's Run: nobody makes it through the regulatory process because it takes so long.

Large companies such as Medtronic, GE, and Philips can take a thousand shots on goal. Even if some have trouble, some get through. We need to get more technology through the regulatory process because biology is unpredictable.

We want more data about biology. This is a way to get more data on the same platform, quickly and at low cost.

You've said that you're essentially using silicon and software to replace drugs.

We think silicon hospitals are within reach. We have good data on cancers, mental disease, neurodegenerative disease, longevity, and chronic diseases such as diabetes. We can activate certain cells, monitor them, and see what's happening.

We also have imaging technology that we put on the back burner for a while because we realized we could ship these devices faster. Ultimately, we think we can replace the MRI machine with a low-cost wearable that leverages the technology we're building into these 2 units.

The same technology, physics, and chipsets in these devices can provide different therapeutics and diagnostics through different software.

Exactly. We can put the technology into the hands of thousands of labs and scientists around the world, and they can find novel uses for it.

We're working with governments, ministries of health, large companies, and small companies. They can trust the system. If we overcharge or try to game things, they can go to another manufacturer and have it made. We still have a good business.

Right now, nobody else can make these. We've made the plans, but it's still hard to manufacture and design the systems. We're pushing the envelope.

Elon Musk open-sourced the patents for the rockets and Tesla's charging stations. Boeing could open-source all of its technology, but it would still be hard to make. We're pushing things forward, but people are dying in the process.

A million scientific papers have been published over the last 20 years about using infrared light, ultrasound, and electromagnetics to treat hundreds of different diseases. It's a rounding error to say that almost none of this technology has made it into people or the health care system.

That's because of the $658 million and 13 years required for approval. We have to break that mold and move to a different pathway.

When will these devices be sold to labs and governments?

We're taking reservations on our website because we can't sell a non-FDA-approved device without the right documentation. These are research devices. The first one will ship this month, and production begins in the second quarter of 2024.

Thanks to Vitalik's gift, we were able to shrink the devices down. We thought that was important because it's a general-purpose platform.

I remember the movie Brainstorm. They had a giant device that could read and write onto neurons, and then they shrank it down to a small device you could wear on your head.

That's what you're doing. You're about to unleash an entire revolution.

The idea is to flip the model on its side. Many different companies and health care organizations can treat hundreds of diseases in parallel. That lowers the cost of the hardware and gives us more safety data on the platform, which everyone can share.

Safety and efficacy are important, but they aren't enough to get approval. The open-source approach enables volume. We make money. It's a crass thing, but when you make more of something, it becomes cheaper.

For approximately every 10 times more of something that you make, it becomes cheaper. That's a slight exaggeration, but we share a portion of those savings as profit. It's the Android story. It's what Android did.

There is no quality advantage to a 10-unit build. That's what the FDA considers a 10-unit build. Twenty years ago, when I was CTO of a division at Intel, our minimum sample-size build was 10,000 units.

This enables innovation because the best products go through the most iterations. You can leverage the product, and you create a massive amount of data. We can use AI to see more things, and different people will do better work. We can move back and forth and break this cycle of 20- to 40-year development times in health care while people are dying by the millions.

That's why Vitalik helped us try it. I convinced all of our investors to say yes at the same time, and we signed the deal. We're open-source now and forever.

Openwater's open-source model is certainly appropriate.

I want to ask a couple of personal questions. I want to know about your childhood. I heard you say that your father helped you learn how to fix and build things. You had a farm, and there wasn't enough money.

My father started an automotive-repair business. He rebuilt car engines. I was a little kid, and I could shimmy under the car—which would probably bring in child protective services now—but you figure things out.

We plowed the neighbors' driveways when it snowed. That meant figuring out how to get the tractor started, how to attach the plow, and everything else. You learn to tinker and build.

My father grew up on a farm. Everybody was moving away because there weren't jobs. It was part of that transition as Americans left farms, especially in New England and the Midwest.

Who taught you art?

I loved art. To me, it was the same thing that led me to engineering.

The governor of Connecticut's sister was my art teacher in elementary school. My parents didn't like the art teacher, but I did. I thought it was fun.

I went to one of these schools where, not to date myself, I started kindergarten around 1965. The public elementary school in my district had a continuous-progress system. You could do what you wanted and be what you wanted.

I did math and art. That's what I wanted to do. I was doing calculus by fifth or sixth grade, but I spent a lot of time in the art room because I enjoyed that, too. Math is visual. I know there are music and math geniuses who think of it in terms of music, but I really liked art.

You also played in a band, didn't you?

I did, but I wasn't good. I was in a couple of small punk-rock bands.

I can see you as a punk rocker. What did you play? Were you a singer or did you play an instrument?

I was in a band. It was fun.

I heard you met Andy Warhol. When was that?

I took all these art classes when I started college. My parents hadn't grown up rich, and they wanted me to be able to support myself. They said they would help me pay for the best college I could get into if—and only if—I majored in electrical engineering.

I thought that was fine. I started, and I thought it would kill any ounce of creativity I had. It was so dry and boring. You spent a whole semester on F equals zero, then the next semester on F equals ma. I found it boring and dry.

I started taking art classes to maintain my sanity. I couldn't afford therapy, so I took art courses. I was at Brown as an undergraduate, and RISD was next door. People say the best part of Brown is RISD, and the best part of RISD is Brown. I think they're both great.

I ended up completing all the classes for a second degree in art, but they wouldn't give me the degree because I would have had to pay for a fifth year. I only paid for 4 years. Later, after I got my PhD, I received an honorary PhD, and they gave me the art degree then. I didn't have to pay for it.

One of the things I talk about is going from success to significance. So many entrepreneurs measure themselves by their stock price, the amount of money they've raised, or other elements.

Can you speak to the entrepreneurs listening who want to do significant, bold things in life? What's your advice?

Find a new way to do it. People keep asking about first principles. I always thought it was a first-principles question, like Maxwell's equations. But maybe the bigger answer is to read history—especially the history of science.

To fund my PhD, I worked with a history-of-science professor because my work was completely unfunded. I got whatever research assistantships I could. I built equipment and created kits for students. Later, I was asked to do that for elementary schools in Rhode Island, including Newtonian telescopes and Galilean telescopes.

There were reasons people didn't believe Galileo. The instruments were hard to look through. When you look at what the greats—Faraday, Galileo, Franklin, Rosalind Franklin, and others—had to work through, you have the impression that it was easier for them. It was never easy.

They decided to do it. They worked on it because they loved it and were passionate about it. You think you're going to die if you don't do it.

If you're going to do something big, bold, and significant, you have to love it. If you don't love the job, you should do something else. You have to spend all your time on it. I get up in the morning—or in the middle of the night—because I can't sleep and want to work on the thing I love.

You keep going. If you don't like it, you won't do a good job. I work all the time because I love it, and it's hard to stop me from working.

You're not working. You're playing, having fun, and fulfilling your purpose in life.

What else can you do? You should feel that way about it.

First principles are complicated. Oxford once gave people a fine for diverging from Aristotelian theory, back when Galileo and Newton were making their breakthroughs. It's a good thing they weren't at Oxford. We still have rules against thinking, which is crazy.

You have to find the barriers and see what's been overlooked. When I work on something, I look as far back as I can go, then come forward to the present day. I go back 50 years and see what people missed.

It's not just first principles. There are many principles, and you have to decide which ones to choose. You have to see whether you can find a new way through them given what we have now. People abandoned ideas 20, 30, or 50 years ago. Can we pick them back up and combine them with new things?

The other thing that bothers me is the question people ask startups: "How big is your company?"

How do you measure big? They usually measure it in one way—by how many employees you have. That's the wrong question and the wrong answer.

What should it be?

How big is your desired impact? How many people are you touching?

If they're interested in the company, they may want to know revenue and income. If you're on a board, total shareholder return should be the measure. But in terms of a startup's potential, it should be measured by the potential impact and the roadmap. That would be more interesting.

Maybe they're trying to assess burn rate, because the number of employees and where they're based can be useful for that. But it's not the right measure of impact.

Are you glad you didn't pursue Openwater straight out of Intel or One Laptop per Child, and that you waited until 2016? It seems like the technology became enabled only in the last few years.

It's really the convergence. We could have made some impact in 2004, when I finally had my feet underneath me after the brain tumor. It took years to recover, honestly, and to design a better version of myself. Getting the medications right was a big fight.

I could have started Openwater then. It was such an opportunity to partner with Nicholas, and then the opportunities to work with Sergey and Mark were so significant. We were supposed to do those things.

The reality of business is that they were responsible for big businesses, so I understood it. I enjoyed the work. I knew it could be applied to the body.

By waiting, we got more cycles of Moore's law, so it's easier to make these things. The manufacturing infrastructure is easier to use than ever before.

I've been through it so many times that I know you don't need a lot of people. You use contract manufacturing and teams all over the place. That communication transcends the traditional model of building your own factory and your own everything.

It takes a long time to build those systems. With contract manufacturing, you can turn things on and off quickly and move to different factories if you find issues. That's a detail, but it's important.

Mary Lou, from the bottom of my heart, thank you for everything you're doing. This is one of many chapters in the multivolume book that is Dr. Mary Lou Jepsen. I'm excited to see what comes next.

Are you on social media?

I'm on Twitter/X and Facebook.

What's your handle there?

Twitter: @MLJ.

MLJ, MLJ, MLJ. Okay, MLJ it is—3 MLJs, 3 MLJs in a row.

I again thank you for your brilliance, your perseverance, and for who you are. I know very few entrepreneurs who've got the spirit, the mindset, the perseverance, and the brilliance that you do. I'm grateful to call you a friend, and thank you for your time today.

I don't even know how to respond to such a generous thing, but I'm in awe of all that you're doing and a huge fan of what you're doing. I keep trying to support you, and I'm a member of A360 and all of that. I've learned so much from you, particularly through the pandemic. That's when I joined, because I was so isolated, and I was thinking, “How do you get back to the positivity, the thing that you talk about—the mindset?”

It's an incredible world ahead. It truly is. I think people need to see that. We hear about all of the problems and issues that are plaguing society, all the epidemics, obesity, and all of that. Yes, those things are true, and yes, we have to solve the health-care crisis. But we also have people and technology like you and Openwater that are giving us brand-new tools and giving us wings.

Thank you so much for highlighting us. I think it's going to be a much bigger story. We have those million papers; we need to bring them in, along with all this talent, and support them.

You're going to give the scientific crowd a new iPhone equivalent.

Yeah, that's the thing to build apps on top of. That's another reason why it's open source, because people say, “Well, I don't trust you.” It's competitive: everybody has access, and everyone can use it.

We haven't talked about the implications of AI on top of all of this, right? These systems are going to generate massive amounts of data. You might be able to understand a lot more about biology, and you'll certainly understand a lot more about safety and efficacy in the brain.

And the brain—read/write: 100 billion neurons, 100 trillion synaptic connections—and it's still very much a black box. You're building the telescope. I'm going to call it the telescope for the brain, or the microscope into the brain.

It really is. We can see. I showed live on stage at TED, I think in 2018, focusing through bone and flesh. We didn't get to use real bone and flesh because there was a rule against it in Canada, so we used phantom tissue. But we focused to a micron live on stage.

What does a micron buy you in terms of a neuron?

A single neuron, or groups of neurons. Groups of neurons are really useful for mental disease and neurodegenerative disease, so that is the focus of our first products. But we have a lot of technology that we've opened to the world. People can push it forward, and we'll help. We can do lots of different things.

I can't wait for my own. I can imagine everybody having one of these systems at home and finding the app: “I want to be happier. I want to get better sleep.” There's going to be anything that your brain implements or impacts.

Right, and you can get through clinical trials more easily because everybody can have this at home, so you can try it more easily. We have surveillance systems in our homes, like cameras and microphones, so we can see what the effects are, measure them, and collect more data.

My watch tracker or my heart-rate monitor is accurate to plus or minus 25%. But if you imagine that across millions, rather than—you know, you look at clinical trials, and people do 20 patients, 70 patients. I was looking at a company last night. They're 10 years old, and they've done 76 patients. Insane.

That's it. Yeah. How can you draw as many meaningful conclusions?

Yes, right. Or have the impact, given how much they've spent? Whatever they spent—$100 million—they've done 76 patients, and they've got a long row to hoe to get through approval processes, which probably will need 10,000 patients. And so you're just stuck. We can't get new therapies unless we can get more people to try them.

How do you make sure it's safe?

I mean, I think many of these things look safe. People say, “Well, you're using a different frequency. We're only using one frequency. What if? What if?” And it's like, “Well, yeah, okay, great. What should we do? Should we do this all in a hospital? Do we have to do this all in a university?” That's where you get to spending $40,000 to $70,000 per patient, and then the numbers become astronomical.

Or do you work with the Ministry of Health of a middle-income country that would like to own the regulatory approval? You get 10,000 or 100,000 of these units out there, and people say, “I want to be part of that trial,” and deploy them to their homes. They can do the trial there, or they can do it at the Ministry of Health, which then owns the regulatory approval. Then they do what they feel has been done to them by Big Pharma, for example. It gets interesting when a country can own the regulatory approvals.

I can't wait to see where you are in March. Next year, I want to come back and go deeper into the early results and talk about writing to and reading from the neurons of your brain.

Openwater.health, and on X, @MLJ—MLJ, MLJ, MLJ is your handle there. Have an amazing day. Thank you again for everything.

Thank you, Peter. Take care, my friend.