Thrive Capital 的 Nworah Ayogu 博士谈 Neuralink 为何比 SpaceX 更有趣

Thrive Capital has emerged as one of the most prolific venture capital firms of the past decade. Thrive has been at the center of the biggest AI deals lately.

Over the last 10 years, we've spent about $40 trillion on healthcare. We have not moved our life expectancy an inch.

Life expectancy is dropping. Americans are not living as long as they used to.

And if that is not—

Wow.

…an indictment of the failure of the healthcare system, I don't know what is. $40 trillion for no change.

More so than maybe any other time in history, the frontier of science and the frontier of what we've accomplished are moving faster than ever. We're seeing this explosion of AI and drug discovery, and I think that is changing the way in which we're going to deliver and make drugs. It's also going to change the entire value chain. Two-thirds of physicians in America are using OpenEvidence.

The company says about 40% of doctors in the U.S. right now are already using the chatbot. It's like ChatGPT for doctors.

OpenAI is now one of the largest healthcare providers in America. They have 300 million people asking ChatGPT questions every week—

Mm.

…about healthcare specifically. Healthcare is the key hero use case for most people. Neuralink is a more interesting company than SpaceX. As a nation, I think we need a 10X nurse a lot more than we need a 10X engineer.

Dr. Ayogu, welcome to Sourcery.

Thanks for having me.

We have so much to cover today. I think it's a really awesome time to go deep into Thrive and Thrive's portfolio in healthcare. There's a lot going on in biotech. You're up to 14 healthcare incubations, and Oscar Health was incubated when Thrive launched, so there's a long lore here to get into.

Our first investment was a healthcare investment in Oscar, so it's been part of our DNA from the start.

I just got back from Priscilla and Mark Zuckerberg's event for Biohub, where they came together with the Department of Energy and a handful of other big leaders. Isomorphic Labs was actually onstage there, too. It seems like we're in this complete inflection point of biology, drug discovery, AI, and healthcare, and there's just a lot going on there. I'd love to start with that and just get your pulse on why now.

1. The Biotech Inflection Point

I would say that, more so than maybe any other time in history, the frontier of science and the frontier of what we've accomplished are moving faster than ever. We're seeing this explosion of AI and drug discovery, and I think that is changing the way in which we are going to deliver and make drugs. It's also going to change the entire value chain.

The same way that with AI we've seen that trend flow through the entire supply chain—from needs around compute to needs around memory to needs around energy—we're going to see the same thing with biotech. It's not just drug discovery. It's going to be manufacturing, development, how you do your clinical trials, and commercialization. We're going to see that flow all the way through. We're seeing these new trends in N-of-1 medicine.

Mm-hmm.

That's the ability to create actual customized treatments for an individual that can treat conditions that we could never treat before. We've seen that already occur and already have an effect, and we're only going to see more and more of that.

The last thing I'd flag would be this interesting intersection between hardware and biotech as well, that's allowing us to unlock treatments we couldn't unlock before. I think of 2 of our companies, Neuralink and Fortell. When you look at the problems that they're solving and the capabilities they let us have, they're almost biblical in proportion, right? The lame can walk. The mute can speak. It's quite amazing when you actually see what folks have been able to unlock.

It's incredible. For those who don't know Thrive's portfolio in healthcare, I'll just name off a few companies like OpenEvidence, OpenAI, Cadence, Nourish, and a handful of others. But in terms of the trillion-dollar company, or that one being more of a traditional healthcare company, there's obviously an explosion of companies that are now really grasping the proliferation of using and utilizing data, and how much data can be shared between one another, whether it's AlphaFold or drug discovery—all these sorts of things. How do you make sense of the evolution of that and what buckets it's flowing into?

2. AI Rebuilds Drug Discovery

If we think of AI for drug discovery and where people often think of companies like Isomorphic and companies like Chai, the magic of it is saying, look, a lot of times you have a target. You have some molecule that's out there that, when we hit this target, changes something within the body.

But oftentimes, we need to understand what the right targets to hit are. Folks will call that biological risk. If we do something to this target, if we change it, what happens to the rest of the body? If we upregulate it or downregulate it, does it make you sleepier? Does it make you more awake? Does it make you lose weight? We need to understand what happens with that target, and then we need to understand all the off-target effects.

What are the side effects that happen? We want to do something—we want to make you lose weight—but we also don't want to make you nauseous. We don't want to make your heart rate go up. We don't want to give you liver toxicity. Those are the off-target effects. Biological risk is the thing we're trying to solve. We want to do the thing we want to do, and we want to make sure we don't do the things that we don't want to do.

Now, the flip side of that is that we then have to make a molecule to bind that target. When you talk about all these models—protein-folding models, molecular design, AI for drug discovery—people often think of that problem: How do we actually make this molecule? Some molecules are easier to make than others. You'll hear this term, “undruggable.”

Mm.

Undruggable means that the pocket is small, or sometimes there are molecules where I have to bind to part of it, because all these things are three-dimensional shapes. So if I bind to part of it, the molecule itself may move, so something else can bind. They'll talk about hidden pockets and things that make this stuff hard to do, because you have to be able to predict in real time what it is now and how the shape changes over time.

Those are the things that you're often trying to predict. With these drug discovery models, the idea is: Can we first create a model that will let us understand what is the best thing to make to bind this? That's step 1. The next step is: I've made something that can bind here. Can I also get something that will predict what happens after I bind this thing? After that, you also want to be able to predict what the unwanted effects are that I can get. That's the progression of the problems that we're trying to solve with these models.

I think the first step—what molecule binds well—is the thing we've actually gotten better and better and better at doing. But what folks have realized is that, now that we're able to do that better, there's still room to grow and room to improve. The actual thing that we want to be able to solve next is being able to predict what those effects are.

With coding and math, those things are very verifiable. I know whether my code compiles or not, right? For these things, I also have to verify: Did my molecule work?

Mm.

Does it bind to the target? I can actually test how well this thing binds. But whether this has the effect that I want is a more complicated question. I then have to be able to put it into a cell and see: Does this change the cell in the way that I want it to?

To figure out what side effects this has on a person, I need to first put it into a mouse to figure out: Does it do what I want it to do in a mouse? There's a saying we always have in research: We've cured cancer in mice 50 times. Everyone's like, “Look, if you want to cure your cancer, the best thing to do is to be a mouse,” right? Because we've cured that 50 million times, and we have not cured it in all cases in humans.

So oftentimes we'll put it in an animal and say, “Okay, how does it work in the animal?” Sometimes that's a proxy for a human, but the real best way to know how it's going to affect you as a person is to put it into another person. Those are the levels of data that we need to figure out. How do we test it in animals and humans more rapidly?

Other people are looking at how we actually get data that will give us good proxies for that. That's what we actually want the models to be able to predict. But before the models can predict what effect it will have, we have to have enough data to train them to make those predictions.

That's a bit of what you're hearing folks say when they say, “We need to generate the data to let us do that.” There's both data to let us verify the experiments we're doing, but there's now another vertical: What's the data that will get the models good enough to actually make those predictions?

Onstage at the Biohub event, Max was on this panel, and I think the question was, “What do you want to see in the next 5 years? What would be your crazy goal?” He said, “88% to 90% success rates.”

For a process right now that has 90-plus-percent failure rates—

Failure rates, yep.

—takes 10 to 15 years and costs billions of dollars. So what would change for that to happen?

It's one of the things I love about Isomorphic Labs. I think there are folks on Twitter who are like, “Well, you know, Isomorphic Labs, they're really good at protein folding,” but protein folding's not the only problem.

I'm like, look, Iso and the team invented protein folding with AlphaFold, right? That's why Demis got the Nobel Prize: because he created this thing—

It's hard to do.

—that created this whole field. And that's hard to do, which everyone is now enjoying and running at. It's amazing. We want more people solving these hard problems.

But AlphaFold came out in 2018, so they've been doing this for a while. They also realized a while ago, and have been talking at least internally for a while—

Mm.

—about, as we solve this problem, what are the other steps and processes to work on? They've been thinking about all of those other steps that you need to accelerate. They've been thinking about what it looks like if we're going to get a 90-percent success rate, if we're going to make this go from a 10-year process to a 4-year process.

It means that we need to get better at our molecular design. We have to get better at our in vitro testing, so testing in the cells. We have to get better at our in vivo testing, our testing in the animals. And we have to get better at our predictions of what works so that we may not be able to one-shot everything all the way through all the phases of clinical trials. Rather than having a 5-percent success rate, I can have a 60-percent success rate. That's amazing, because that radically changes—

Mm.

—the things I'm willing to do all the way throughout the process.

But it also means we've got to think a lot about what happens in the clinical trial process as well, around phase 1, phase 2, and phase 3. That is not just companies. That is us as a society thinking about how we optimize around this process.

Credit to ARPA-H. I think they've been doing a lot of work recently to think about what it could look like to drive efficiency around that portion as well. What we don't want to do is get very good at preclinical and still have the clinical process be the bottleneck.

I think we have governmental agencies who are thinking about that. We also have folks within these companies who are thinking about what needs to be true so that the rest of the system is ready to absorb these innovations as they go through. Even after we do that, again, the bottleneck continues to shift, and we talk a lot about how oftentimes you have to plan for the end at the beginning.

Mm-hmm.

So we will also need to make sure that we're getting better at how we manufacture these solutions as they're coming through. If we're going to be in a world where we can have new drugs coming out every 3 years rather than every 10 years, that also means our manufacturing processes have to get better and faster and have the ability to absorb that demand.

To focus on Isomorphic Labs a little bit more, this is a really interesting story of how you guys even got into it. Demis wasn't raising. I think Josh came over to him and was like, “Let's work something out.” I read this in the Colossus piece, which was great, about Thrive and Josh. Can you walk through how you guys got involved?

We've always had this philosophy that we want companies that are bilingual. We want companies that are truly the best at both of what they do. They speak 2 languages fluently. I think Isomorphic Labs is such a perfect example of that.

The AI researchers at Isomorphic Labs are world-class, which makes sense given that Demis also founded and ran DeepMind. These are some of the best AI researchers in the world. It's DeepMind, Isomorphic Labs, Anthropic, and OpenAI. They all hold the same bar when it comes to their AI researchers.

Those chemists and biologists who sit within Isomorphic Labs—I’ve met these folks. These folks are world-class. They are as good as the folks you would find at Eli Lilly, Pfizer, the Broad Institute, or the Wyss Institute. These are world-class chemists. They're truly world-class in both. They're bilingual. They're amazing at the biology, the chemistry, and the science, and amazing at the technology and the AI.

I think some of the magic of what they do is that at Isomorphic Labs they talk a lot about the Isomorphic Labs way and the fact that they're actually bringing folks together to be able to do both who are world-class. They have these amazing chemists who are integrating these AI tools, but using them to solve problems and bringing their knowledge of chemistry in the same way.

This intersection of both, and having that culture of excellence in both worlds, but also being able to get that cross-talk so you can overlap, is a lot of what makes that company so special.

And I think for Josh, he spent so much time working with Demis, letting them know that we appreciate what they're doing. To us, healthcare is a space that we know and are deeply familiar with, and we're as passionate about this space as they are. I think they're firm believers that a lot of the benefit that we're going to get from AI is going to come in the space of healthcare. That's also something that we believe, and I believe deeply to my core.

I think it's amazing, and it's a common theme through all of these AI-healthcare conversations I've been having, whether it was with Annie Lamont the other week, or even one I did way back when with Commure. It's equalizing the playing field for people in the middle of the country to receive the same kind of access to care that people in big cities or really expensive, hard-to-get-into hospitals can get.

I want to go into this bilingual part a little bit more, this model of what these companies hold for healthcare. A really great example of this is Neuralink. Can you explain the different types of talent within Neuralink? Because you have scientists, neuroscientists, engineers, hard tech—you have all of these different components coming together, and it's crazy, crazy pockets of talent.

3. Neuralink Restores Lost Function

One of my biggest critiques of healthcare is that I think we often do a very good job of extending life, but we don't do a very good job of improving people's quality of life, right? Of giving them that life back.

Neuralink now focuses on a lot of populations, but one of the first populations they focused on was folks who have ALS or paralysis. These are people who, whether through a disease, a car accident, or as veterans who have been serving our country but maybe were hit with an IED, no longer have the use of their arms and legs.

Oftentimes, these are also young folks—people who are 30 or 40—who no longer have the ability to work because you often need your hands and legs to work. They can't be outside playing with their children. What Neuralink does is say, “Okay, what could we do to give these people access to interact with the world again?”

They've created an implantable device that lets these individuals interact with and control a computer with their minds. They've created motorized arms that folks can now control, a sort of robotic arm to pick things and grab things. They have another trial where they implanted the device in the speech center, the speech motor cortex of the brain, so that someone who wasn't able to speak could speak.

The core of what they do is: let's use this implant to give people the ability to manipulate the outside world again. We may not be able to replace the arms and legs that you've lost, but we can give you that functionality back. That's what Neuralink does.

Now, when you meet the folks at Neuralink—and to your point about why they're bilingual—Neuralink has these amazing hardware builders, which makes sense. They're hardware folks who are at the same level as SpaceX and Tesla. These are true hardware builders who are building things with heavy complexity.

An example of that would be one of the things they have within the implant, which is very, very small: they have these microscopic threads, and their thread count gives you the ability to read from the neurons. I think at this point there are over 1,000 of those. They've increased the thread count as they went. That's amazing hardware in and of itself. Because the threads are so small, they actually created a separate robot to do the implantation.

Because it’s such a difficult procedure to do with something that microscopic, and they wanted it done with extreme precision and reliability, they created their own robot to do that. It’s not surprising that when you have access to folks who are SpaceX- and Tesla Optimus-level, you can also create a surgical robot to do this elite kind of neurosurgery. It’s wild to think about the fact that there are other companies for whom their whole business would be, “We have created a neurosurgical robot.” For these folks, this is one vertical of what they do and what they’re doing.

But when you go in, you also meet the neurosurgeons who are training with them, and I know some of these neurosurgeons. They went to Harvard Med with me. These were the best students and neurosurgeons in their classes, and these folks now are here doing that. These are practicing neurosurgeons who are co-developing these tools with them.

They have pathologists. I remember the first time I went to their pathology labs. I’ve never met a company before that insourced its own pathology lab.

Interesting.

Most places, you send it out. You get your slides, send them out, and go forth from there. What they realized, though, is that pathology is so key to what they’re doing because they want to understand, in humans and in their animals, whether there are any effects this is having on the brain. They want to be able to get that data and understand it because safety is so critical to what they do. Therefore, they’re like, “We are going to insource the pathology lab here because it’s what we do, and it’s critical to what we do.” So they brought in all these world-class pathologists.

This brings home the thesis that I think you see in so many of these Elon companies: “We’re going to vertically integrate through all parts of our supply chain that are important. If it’s important—if it’s mission-critical—we need to have it in-house.” Seeing them do that across things as diverse as not just hardware, but something like pathology, which most people don’t think of, I think that was that impressive level of detail.

That’s so interesting. I have to ask—I have to go deeper into this. Okay. So Neuralink is exceptionally great at hiring the world’s best talent, from pathologists and roboticists—

Neurosurgeons.

Neurosurgeons—

Radiology, because they’re scanning your brain.

Battery experts.

Battery experts. So this is the thing. When you actually look at the people using it, they’re using this all day, which makes sense.

Mm-hmm.

How often do you use your hands? You’re using your hands all day. Your feet—you’re using your feet all day. Maybe we’re not using them right now, but the idea that, yes, you need a battery that’s going to last all day because this is a device that people are using all day. Talk about product-market fit.

So how do they convince all these people to work on such a crazy moonshot, by the way? Incredible results and outcomes already, but these are relatively new. This was happening in the last couple of years: the first implant happened, and we started to see some of the real results.

Yeah. I think, first, D.J., who is the president and leads a lot of that company, is a very compelling leader. I think when you meet him, you can tell that his passion is there. In a lot of ways, I think passion is contagious. I also think excellence and talent density are contagious.

People who are truly N of 1 want to work with other N of 1 people. I also think they’re working on such an important problem, and they were category-defining. I think when Neuralink was first doing this, no one else was doing it. So they didn’t just step out into the lead. They created the category. Everyone who cared about this space, they brought them in, and now, as they’ve gone forward, their lead has only compounded.

I think the other thing—and maybe this is one of my hot takes—is that I actually think Neuralink is a more interesting company than SpaceX. The brain is vaster than space.

So, a $3 trillion company? What are we thinking?

We don’t make it to space, and we don’t invent silicon chips without our brains. That is the fundamental substrate that has unlocked all of this. We talk a lot about compute and intelligence. The brain is still a more efficient engine of compute than every chip that NVIDIA has ever made.

So, on the topic of hardware and robotics, let’s talk about Fortell. Explain Fortell and why hearing aids are just another extension that we need to continue to innovate on.

4. Fortell Reinvents Hearing Aids

Yeah. Let’s talk about hearing aids and hearing loss.

Hearing loss affects 1.5 billion people worldwide. We all go to concerts. We all have headphones in our ears at all times. I would only expect that number to increase as we go forward. Unfortunately, you and I are probably both going to have hearing loss when we get older. It’s also an issue in that it’s the greatest modifiable risk factor we have for dementia.

Mm-hmm.

Folks who have hearing loss also have a one-third increase in their likelihood of having dementia. Think about that for a lot of reasons, but your ability to hear things is anchoring for the brain. Vision and hearing are the brain’s main sensory inputs. A lot of people theorize that losing that, as well as the social isolation that comes from not being able to hear, are some of the main contributors to dementia.

So if you want to keep your brain sharp, one of the most important things you can do is not lose your hearing. Now, most hearing aids just amplify all sound, right? That’s actually not what hearing is. Hearing is both the sound itself and your brain dampening things in and out. It’s part of why, when you’re in a coffee shop and having a conversation, you can have the conversation because your brain knows, “Focus on the person I’m talking to and tune everything else out.”

Most hearing aids just increase everything, which is why it’s so hard for folks with hearing aids to go out to dinner or be on a crowded street. Anywhere where there’s background noise, it’s actually very hard for them, which is why those folks often will just stay home where it’s quiet.

Matt DeJong is an amazing founder. He’s an amazing hardware builder at a company called Butterfly, which makes the handheld ultrasound. He’d had grandparents who’d unfortunately suffered from hearing loss for a long time, and he’d seen the devastating effect it had on them. So he’d spent 15 years thinking about this project and a lot of time working on it as a side project. He was fairly obsessed with a space that most people aren’t obsessed with.

Around 5 years ago, he brought his team together to actually work on this full-time, and a lot of it was driven by changes he saw in 2 verticals. One was the changes he saw in silicon and in chips. So we’ve got this story where they all went out to TSMC, and they taped out their own chips, which is so hard to do. I think you’ve talked to a bunch—

My God.

of these kinds of folks—

Yeah.

who work at the new kind of chip fab companies. It's so hard to tape out your first chip. It's unheard of to tape out your first chip successfully the first time. But in a lot of ways, because of the way a hearing aid works, they do need their own custom chip because hearing aids are very small, right? It's the aesthetic. No one wants this giant thing—

I know people who have gotten—

on the side of their ear—

them stuck a couple of times.

Yeah. So you want something small, but it has to last all day because you're using—

Mm.

it all day. So you need the battery life to last. You actually have to have very fast inference because you're translating incoming speech, right? So that's difficult. There's just a lot of complexity that you have to have, which is why they had to design and tape out their own custom chips.

The other thing was that they were thinking a lot about how to do the sound processing such that we know what we want to amplify versus what we don't. They talk a lot about how he got this idea for semantic noise processing during the pandemic from Zoom. One of the things that folks were doing with all of the noise-canceling technology they were developing was saying, “Look, we're not just going to— We're going to dampen the sound up and down by knowing what kinds of sound this is.”

So different kinds of sound have different tones. Being able to tell the sound of a conversation that's in front of you from a car going by—what the actual noise itself is—and processing and understanding that, then using it to dampen the sound up and down, is one of the strategies they used.

For them, they took this hard technical project and problem and decided to build a company around it, and that was the company Fortell. I think for them, they've had this amazing uptick in consumer validation—

Mm.

—and technical validation. When they had their beta, they had a range of folks, and I think everyone in their beta loved it. I'm always careful with my healthcare companies not to divulge information they haven't already divulged, especially with their patients, but these details are public. Steve Martin and Henry Kravis are both folks who live very active lives and are hard judges of quality.

Yes.

Anybody who's worked for them or with them would say that these are hard judges. They have loved the product. They've both been users, but they've also passed it along to their friends and said how amazing it is and how much it has changed their lives.

They had one of their beta customers who, after the beta ended, was told, “Okay, the beta's done. Thanks for letting us have this. Can we have this back?” And the person said, “Absolutely not. Over my dead body will I give this beta product back to you.”

No.

The other thing is that there's this group that's essentially a third party that rates all hearing aids. They have a rating scale that they use. A 4.5 was the highest rating that a hearing aid had ever gotten on their 5-point scale. Fortell recently did it and got a 6.

What?

They broke the scale?

They literally broke the scale, and I think that shows, both from a technical perspective—

Mm.

—as well as from a consumer perspective, just how well it works.

Mm.

And I think that's a perfect example of creating this hardware product to meet a need in which you understand the unique problem you're solving, and then going forward and saying, “Okay, what are the advances in hardware and AI that will let us solve this in a very different way?”

I think we're seeing that. The market is loving it. It's impressive to see people applying these tools to problems that are really debilitating for people's quality of life.

My mom has hearing loss. She underwent chemo and lost her hearing from it. This was years ago. So what is the process for people who want a Fortell hearing aid? Is the experience different? How do you make it more customized?

Yeah. Me and Josh joke that we're very popular with the 70-plus crowd now because we have so many people who will call us and say, “Hey, my mother, my father, my uncle really needs this. Can you help us move forward on the wait list?”

Mm-hmm.

Right now, they have—

A wait list? Come on.

I know, I know. But they're working on ramping up, so the wait list should be going away very soon.

Mm.

Right now, they have an office in New York. They also have their website, where they fit folks. They want to make sure that people get to try it before they buy it. They'll be expanding rapidly from there.

They've created this great product, and now they're doing everything they can to make sure they have enough of it for everyone to utilize.

Incredible. We've gone so far and haven't even gone into your background yet. Dr. Nworah, please tell me your life story.

5. From Physician To Investor

I was born on a snowy day in Columbus, Ohio.

No, I think for me, I talk a lot about how, in a lot of ways, I've spent a bunch of time making good things big. I spent a lot of time at startups trying to scale very effective models, and I've also spent a lot of time trying to make big things good.

So, bringing that kind of startup mentality of tackling hard problems rapidly to large-scale problems, I trained as a physician in internal medicine. I originally studied neurobiology and health policy in undergrad, so I liked both the social science and the science. I studied medicine and business in grad school, so I went to business school and medical school. I did residency, so I'm a trained internist and physician.

After that, I joined the founding team of a company called Cityblock Health. It's a Thrive portfolio company. What Cityblock does is provide integrated care for high-cost, high-social-need populations: full-stack social care, behavioral healthcare, and medical care for the patients who need it the most.

I joined that company in its very early days and saw it scale out to become a national care infrastructure. Then I spent a bunch of time at Amazon, launching what I guess is now Amazon Health. We turned PillPack into Amazon Pharmacy, and I served as our first chief medical officer there.

Then we launched a business called Amazon Clinic. Our first vertical at Amazon Pharmacy was asking, “What's the space in which Amazon will have an advantage in healthcare?”

Yeah.

We were like, “Amazon has this amazing logistics network.”

Mm.

They're very good at getting things from point A to point B, right? That would be our advantage in pharmacy. If we're going to have a delivery pharmacy, it would make sense that we should be able to get these medications from point A to point B—from the pharmacy to people's homes.

There was a lot of work on pricing and selection that we did, but really, that was the core of it. We said, “Okay, once that started working and we saw it was going well, the next question was: If people trust you with their medications, will they trust you with their care delivery?”

That was when I launched Amazon Clinic and ran that business. It was a virtual-care business saying, “Let's handle the low-acuity, more transactional portions of care,” because we felt like that fit more with the shoppable model you had on Amazon.com.

Once we got adoption there, we said, “Okay, the next step should be: If they'll trust us with their low-acuity, transactional care, will they also trust us with their low-acuity, relational care?”

That's what drove us to buy One Medical and integrate it in. All those pieces together became Amazon Health. It was about bringing a large tech company into a new space, with that space being healthcare.

About two and a half years ago, I joined Thrive. I've known Josh and Karim since college. I was familiar with their work in healthcare, and I was familiar with them as people. Thrive had always been a group of people and investors that I respected, but they'd been making some pretty big bets around AI. I was a big believer in AI as an opportunity.

Mm-hmm.

So I came over to focus on all things healthcare.

Incredible. I'm so interested in all those lines, and now we're getting to the age of trillion-dollar companies. I'm sure you've incurred a lot of challenges along the way with the healthcare system, but as we get into the age of trillion-dollar healthcare companies, is policy still going to be a bottleneck? Or what are the bottlenecks going to be? How do these technology companies reach that level?

6. Healthcare Becomes An AGI Trade

Yeah. I think, in a lot of ways, if you're truly AGI-pilled, as they say, you should actually be long healthcare.

Mm.

I think we're only going to see more and more of these trillion-dollar companies in healthcare for a few reasons. One is that as these capabilities progress, what are the things that we as humans are going to want to spend our personal money and our societal resources on as we go forward? If you take AGI to the extreme, we're all going to want more time, and we're going to want to be able to enjoy the time that we have. So the things that we know will be increased buckets will be healthcare, because that is the way in which we increase the time and the quality of the time that we have on this earth.

Mm.

And then some version of entertainment. We can be long those 2 things. So I think that macro trend, if you believe in AI and AGI, will continue to go forward.

I think there's also a point you could make that since time immemorial, we've also been doing that. I had a mentor once who told me—this is a bit of an aside—that there are 3 professions that have existed for all time. He said they were pastors, doctors, and teachers. As long as there are humans, you'll need people to care for the body, people to care for the mind, and people to care for the soul.

So I could say that going forward we'll spend more on healthcare, and I'd also argue that going backward we also did the same. I think that's one reason: the macro tailwinds are there.

I think the second is that we talked about this frontier of science. We're seeing it expand in such unique ways. Keytruda is one of Merck's drugs. Keytruda's 2025 revenue from that one drug was larger than the revenue of all the major labs combined.

Oh, my God.

That's one drug. Biotech is already big business.

What does Keytruda do?

Keytruda is what we call a checkpoint inhibitor.

Okay.

Think of it as an immune booster. It can boost your immune system. But the interesting thing is that companies are continuing to find new use cases for a lot of these medications.

You've probably heard about Merck and Moderna's new cancer vaccine, which I think is an example of the kind of amazing frontier science—

mRNA.

—that we're getting. mRNA.

Mm-hmm.

Exactly. Moderna, the COVID darlings of mRNA. The way that they do it is, with melanoma, you take the tumor out and have your tumor surgery. What they do is say, "Okay, we're going to take this tumor," because each tumor is unique. Your mother's cancer is likely different in a lot of ways from someone else's cancer.

We talk about cancer as this broad thing, but even melanoma, a specific type of skin cancer, is different in each person because tumors mutate and change in different ways. So what they do is take the cancer out, sequence it, and say, "Okay, what are the parts of this tumor—its expression, its own genome—that make it unique? We're going to create an mRNA vaccine to target that exact cancer."

They say, "We take the cancer out. We're then going to make this vaccine, inject it into you, and also give you Keytruda to give you an immune boost at the same time—

Mm.

—to make it easier for your immune system to really recognize that cancer." Then any microscopic pieces of that cancer that are still floating around in your body, your immune system is now boosted to wipe all of those out.

Your immune system also has memory. So if that cancer happens to come back in 3 or 4 years, your immune system will remember it and be able to fight that cancer while it's still small and get rid of it. That was the thesis of cancer vaccines, and that's exactly how it played out.

I think in their phase 3 clinical trial, they found that people who had that vaccine were 50% less likely to have a recurrence of that cancer in 5 years. So I think that's an example of both the frontier and the types of new treatments we're able to create, but also the fact that a key portion of these new treatments is now Keytruda, this immune checkpoint inhibitor.

Mm.

I think we're going to see more growth because the frontier is able to tackle more and more problems.

I think the third piece goes back to the Lilly example. For that, maybe it's worth zooming out a bit on why I think Lilly won. Lilly wasn't first. Novo was first.

Mm-hmm.

They had a drug with better weight loss, but at the end of the day, when GLP-1s first came out, the drug was so popular that there were shortages.

Yeah.

No one could get them.

Yeah.

Lilly leaned into its manufacturing capabilities and innovated across the value chain. We talked about that when we were talking about ISO: You have to innovate across the value chain and project forward to where the bottlenecks will be.

They were able to have more supply and get supply on board faster—a whole year before, I think. They took a year before Novo could unblock its supply. Lilly got there a year beforehand, so it had more supply out there.

But I think the other thing was commercial. We all know Lilly Direct. They were able to say, "Sure, we're a pharma company, but let's think in an innovative way about how we partner and how we lean into care delivery," and they unblocked that commercial aspect.

I think that's actually why they've been able to be so successful and be the trillion-dollar drug. They're saying, "We're not just going to focus on the science. We're going to be great at the science, but we're also going to think about manufacturing. We're going to think about commercialization."

I think that's actually the opportunity. We're seeing more and more of these companies, the same way that in software, hardware, and AI we're saying it's easier to build with AI, so we're seeing companies that have more surface area right out of the gate. The companies are more ambitious.

Mm-hmm.

I think we're seeing the same thing now in biotech. We have companies saying, "It's actually a lot easier to build things in all these spaces. Therefore, we're going to have more surface area. We're going to do the science and the manufacturing. We're going to do the science and more of the care delivery."

That makes their businesses more defensible but also gives them a much bigger TAM. I think that's a big reason we're going to see more of these trillion-dollar companies: people's ambitions have gotten larger.

We often say that in a lot of ways, what we're seeing now is that our problems haven't gotten smaller. The tools we have have gotten bigger.

Mm-hmm. And you helped OpenAI start up its healthcare platform. How did that get involved?

7. AI Enters Everyday Healthcare

Our time with OpenAI was around 2 and change years ago.

Mm-hmm.

We were talking with Brad and Sam about healthcare writ large, and there was this push that they should spend more time with leaders in healthcare because this was going to be a key space for them. Therefore, they should spend time with incumbents both to understand how to work with them and to understand the space better.

I organized a series of meetings with different leaders across healthcare. There were separate events with some of the top biotech CEOs, some of the top payer CEOs, and some of the top health system CEOs, along with the OpenAI leadership, to talk through the state of the art in AI, the problems these people were facing, and how to approach them.

I remember with the health system leaders, o1 wasn't out yet. Reasoning was a brand-new thing; it hadn't happened yet. But because we were close to them, I'd been able to play with the reasoning models myself, and I'd started putting in New England Journal of Medicine case studies—think of them as House, M.D.-level difficulty, really hard cases.

I would put the first half of them in, because the way they tell the story is: Here's their history, here's this, here's that, and then something happened.

And then they'll tell you about what the expert said and what it was. So I'd cut it off right there at “And then it happened.” I'd just put that portion in every week, and I was blown away by the fact that every single week, they would get it right. These are cases that neither me nor my colleagues, who I'd say are some of the best doctors out there, maybe we'd get right 2 times out of 10.

OpenAI's o1 was batting 10 out of 10, which was amazing to me. I remember going to that meeting and saying, “Look, I think these models are amazing. These models can diagnose.” And everyone was treating me like I was a heretic. It was like, “This is blasphemy,” looking at me like, “This is impossible.”

I remember we then showed them and did a couple of them live. Seeing the eyes of some of these healthcare CEOs, some of whom were also doctors, was just amazing. You could see it start to click for them: “Oh, wow, this is the capability of these tools. This is what's happening. This is what's coming down the pipeline.”

The same thing happened with one of the biotech CEOs. We were like, “Okay, what do you think of these tools? Let's talk through them.” And they were like, “Well, I think they can be really good for customer service. I don't really think, though, these tools will ever be useful in drug discovery.” Now, 2 years later—

Mm-hmm.

One of those same CEOs has announced billions of dollars of spending in this. So it's also been very interesting—

Huh.

Seeing the progression of folks' belief in the tools in such a short amount of time.

Yeah.

That's a really good point. So how has the tool stack changed for these—I don't know—whether it's physicians, doctors, scientists, or anybody in healthcare? How have the tools changed?

Yeah. It's really interesting because I think you see it differently across each of the verticals. If we take healthcare services, I think where people are applying the tools has been very different, and there's a distinction between where the system is applying the tools versus the individual.

Right.

We'll say the system writ large. I always talk about how there's a front office, a back office, and an in-office. That's a division I like to use for healthcare. For the back office, we've seen tons of adoption in revenue cycle management. You can think of that as how doctors and health systems optimize the amount of money that's coming into their offices. It's them coding and putting in the right codes so they can get paid the right amounts.

We've seen an explosion of tools doing that, which makes sense, right? People will adopt the tools that help them make more money very rapidly. We've seen an explosion of tools there. We've also seen more and more tools start to pick up in the front office, so think of all the folks who are on phones, calling people, scheduling people, and following up with you to say you're due for your vaccine.

I think as the voice agents have improved, we've seen more and more folks adopting tools in that space. The interesting part has been that we've started to see more and more tools getting adopted in the in-office space.

Mm-hmm.

Scribes were one of the first ones to really take off vertically within the in-office space. I think the other is OpenEvidence, one of our companies, which I first heard about from friends of mine from residency. To me, it was fascinating because I've never heard doctors rave about a clinical tool. We normally call it health IT.

Mm-hmm.

These things are like Brussels sprouts, right? People are forced to use these tools. Someone says, “You have to use this pager,” or, “You have to use this enterprise tool,” and they drag doctors kicking and screaming. They monitor you for how much you're doing X, Y, and Z, and they hit you over the head and flog you for using these terrible tools.

Yeah.

Therefore, the thesis was always, “Doctors just don't like technology.” Well, no, because all the doctors have iPhones. They all have Macs at home. They just don't like the bad technology that they're forced to use all the time.

This is a use case where I thought, “People are willfully telling me about something with joy in their eyes. I'm very confused.” I tried the product myself and thought, “This is differentiated. This is amazing.” Then I learned more and more about their traction.

I remember Dan had actually gone to college with me, so I called up a buddy and was like, “Hey, where does he live?” I shot him an email and said, “Hey, Dan, it's good to see you. It's been a while. Can we catch up? Why don't I just come by your house?” I hopped in a car and drove down to his house. We had a meeting that day.

My clinician hat is what let me actually see, “Oh, wow, this is a differentiated product.” It's one of those things where, if you're a clinician and you see it, you see right away why that's an N-of-1 product. For them, the numbers are amazing: two-thirds of physicians in America are using OpenEvidence.

To put it into context, Epic—which everyone always thinks of as the behemoth in healthcare, and it is—

Yeah.

Epic took 40 years to get to the scale it's at today, and it has less than 50% of physicians using its product after 40 years.

That's crazy.

OpenEvidence got to two-thirds after 3 years.

Whoa.

That is the amazing kind of scale we can get in this new age with a user population that is notoriously hard to gain adoption among.

Because we haven't talked about this yet, could you frame out the numbers behind the healthcare system at large? Even that kind of parameter around OpenEvidence is one thing, but writ large, the United States' healthcare problem and the system we have—what are the numbers behind that?

Oh, it's wild. We spend $5 trillion a year now, right? That's projected to go to $9 trillion by the end of the decade, so $9 trillion a year in spending. Right now, it's $5 trillion.

It's the number-one employer in America. Twenty percent of all jobs are healthcare jobs, and I think that alone is crazy to think about. It used to be manufacturing. Our economy is now driven by healthcare.

We talk a lot about AI and AI's effect on jobs, and where some of the job changes come from. That alone is insane, but it's also the fact that, as we project forward, people project that 30% of the new jobs we create will also be in healthcare.

Hmm.

It's our largest job engine now, and it's only projected to become a larger and larger share. It's the largest line item in the federal budget—6 times the size of defense. We talk about the defense resurgence and how it's time to build big business in defense.

Yeah.

That's great, but again, healthcare is 6 times larger. It is quite impactful, both in terms of the impact it has on folks and just the size of the industry itself.

There's a trend in Sourcery interviews, if you notice. Now we're in the biotech age.

Welcome.

We were just in the defense world. Now we're going into biotech.

8. Healthcare Must Improve Outcomes

It's funny. When you think about impacts, this is another thing I think about. There's all the money that's in healthcare, and there's all the jobs that are in healthcare.

Hmm.

The impact it has on the nation's economy can't be understated. But when you actually look at the outcomes as well, that's also fairly interesting. I think a lot about life expectancy because we—

Mm-hmm.

We always talk about how we spend all this on healthcare, and I think in a lot of ways we'd actually be fine spending on healthcare if we had better outcomes, right? I always say, look, I'm actually fine with the trade, and maybe this is just me. If someone told me, “Hey, we're going to spend $1 trillion more on healthcare this year, but everyone gets to live to 100,” I'll take that trade any day of the week.

Yeah.

Right? Cut somewhere else. It doesn't matter. If we're giving everybody an extra 20 years of life, all day. All day. I will choose that for the nation. I will choose that for me. I will choose that for you. I'll choose that for my parents and your parents.

But that's not the story we have, right? We're spending a ton for very bad outcomes, some of the worst outcomes in the developed world. I was looking through the life expectancy trends because I think that's the outcome, right? You've got to stay oriented on the outcome that matters.

It's not just the spend; it's the outcome. Right now, our life expectancy is about 80 years. That's where we are as a nation. Between 1900 and 1950, we went from a life expectancy of 50 years, which is wild, to a life expectancy of 70 years, and a lot of that was a mix. Some of that was increases in sanitation and water, but a lot of it was that we improved infant and child mortality—

Yeah.

So people weren't dying during childbirth, children weren't dying at an early age, and we had antibiotics. We had a bunch of really good antibiotics and vaccines that were created, and people got them. All of a sudden, as a nation, we lived 20 years longer.

Our next big spur was between 1990 and 2010. We managed to go from living to about 70 to getting to about 80, and the big spur of that was cardiovascular disease.

Oh.

Half of that was reductions in smoking—

Mm.

But the other half was all good science. There were statins to lower people's cholesterol. We were talking about your cholesterol beforehand.

Yep.

Right?

Which is good.

It's good. It's good. Great cholesterol.

Thank you, Nworah.

She authorized me to disclose her information, I hope. Post-hoc authorization. But there were also improvements in blood pressure control—

Mm.

So we reduced the number of strokes. It was half medication, half people smoking less. That gave us the next 10 years.

For the last 10 years, we've spent about $40 trillion on health care. We have not moved our life expectancy an inch. And if that is not—

Wow.

—an indictment of the failure of the health care system, I don't know what is. $40 trillion for no change.

Oh my God.

It's interesting because if you break the numbers down, there are stories to be told within them. In Hawaii, they live to be 80. In West Virginia, the average is 72. The richest 1% versus the poorest 1% has a gap of about 15 years if you're a man.

Wow.

So it is a tale of two cities. The other thing I think about for this past decade is that we've actually had huge decreases in cancer mortality.

Mm.

There are so many things that were death sentences that aren't anymore. Breast cancer used to be an almost fatal diagnosis, and we've seen that across a bunch of our cancers—

Yeah.

—where we've had this huge uptick, and it's been amazing. Again, the frontier science has gotten there. It's just been balanced out by things like an increase in overdoses.

We've had that balance of two worlds. But I really think if you look at this next 10 years, this should be the period where, given where we're at with science and treatments now, if we're doing our jobs right, we should look back at this next 10-year period as a period where we went from 80 to 100.

I think the thing that's really interesting when you think of science is this trend of N-of-1 medicine.

Right.

The canonical example people often point to is Baby KJ, who had a rare genetic disease where the baby couldn't process ammonia. More than half of babies with this condition die, which is devastating. As a parent, I cannot imagine anything worse.

At Penn, when they realized this was happening, the only real treatment option was a liver transplant, which is a very complex and difficult thing to get. With organ transplants, we don't have the supply of organs. So what they did was design a custom treatment to gene-edit that baby right after birth.

When the baby was 7 months old, they delivered the gene therapy. It was successful. The baby is now 2 years old, walking and talking—

That's crazy.

—but to be able, in 6 months, to say, "Okay, we're going to diagnose it, we're going to create it, and we're going to manufacture this treatment for just this individual," and have it work is amazing. That family got to take home their child and is getting to raise that child.

The fact that we now have these N-of-1 medicines that can cure these rare genetic problems is amazing. The science has gotten there in so many of these things, but the bottleneck shifts again. I think they say it took $1.5 million to actually make the treatment and deliver it.

Wow.

Again, this is one of the things that you grapple with in medicine and health care as we talk about health care spending. A million and a half dollars so that this child doesn't die but lives—is that worth it? I would say yes. Tell me to increase my taxes, cut somewhere else, and tell me what I need to do. I am all for it.

But again, that's the debate we have in health care because, in aggregate, how do we afford these things? Part of it is deciding what we pay for, what we don't pay for, and where our values lie in these important life-or-death decisions. That is the story of health care.

Every cancer treatment is—

Mm-hmm.

—life or death. Every transplant is life or death. Every gene therapy is life or death. What is the appropriate price tag? It sounds weird to say, but what is the price of a life—the life of a child? That is the question that you're asking as a health care system on a regular basis.

The other side of it is that we have to be able to innovate across the supply chain and across the stacks. How do we make it so it's not $1.5 million to develop this? What infrastructure needs to exist to make that more scalable, more repeatable, and cheaper?

That is both the excitement of the frontier and the challenge. The science is getting to the point where we can do things like this—things that people had dreamed about for so long and that have such massive impact. We can do these things, but at the same time, as we do more and more of them and do them more rapidly, we also have to think about the rest of the system.

How do we make sure the manufacturing is effective? How do we make sure that we know how to pay for these things? We have to think across the entire supply chain. I think that's what's so exciting.

I'm so curious, from your perspective, seeing all these companies on the private side develop new innovations across many different types of health care, what have been the biggest outcomes or the biggest impacts that you've seen on outcomes within the portfolio?

One of our companies is Cadence. What Cadence does is remote patient monitoring. They send patients home with hardware. They'll mail hardware to a patient's home—a blood pressure cuff, a scale—so that they're getting continuous data from these patients who are at home.

These are fairly sick patients. Beyond the hardware, they also have a system behind them. They have AI monitoring these signals at all times, and they have people who are also watching and interacting, whom the AI can escalate to.

The other thing they've done is that some people believe in building a different way. Some people believe you build outside the system; some believe you build with the system. Cadence is built with the system. Cadence works and partners with health systems. They're in about 7,000 sites of care today, with over 100,000 patients whom they're actively monitoring at all times.

They integrate into the EHRs of these patients, so the doctors are also aware of what's happening. They're monitoring you at home, but they're also bridging your care with your doctor, and they have AI monitoring and overseeing all of it.

So that's Cadence's model. Think of it this way: let's say you see your doctor 4 times a year, right? That's 4 hours if you're lucky. Let's be real: it's more like 1 hour because most visits are 15 minutes. But let's be generous: 4 hours, because your doctor spends an hour with you every time, out of the entire year. Most of the other 8,000-plus hours of your year are spent outside that doctor's office.

Mm.

So if we're actually going to change your care, we need to be monitoring you at all of those moments, and that's Cadence's special sauce: being able to actually monitor you at all of those times outside. If you measure your blood pressure and it's high, right? Or let's say you have heart failure. For heart failure, we need to know your weight so that we can make sure your fluids are right. When you step on that scale, we can see, “Huh, your weight's higher than it normally is,” or your blood pressure is too low or too high. They can intervene right away.

And with that, they've been able to see—and they've had this published in the New England Journal of Medicine, the Journal of Cardiology, and the Mayo proceedings—they've found that, one, they've been able to reduce people's hospitalizations by 25%.

Wow.

They've been able to increase the number of folks who actually have their blood pressure controlled, which prevents you from getting strokes—

Mm-hmm.

—by over 80%. They've been able to reduce people's A1Cs if they have diabetes by almost a full point. I think one of the other things that's been amazing is that they're also saving folks money and saving the healthcare system money. They save Medicare $2.5 million every week. Every week: $2.5 million.

So they're both reducing costs and increasing outcomes. And this isn't outcomes like, you know, you feel better, which matters. Patient experience matters, but these are hard outcomes in terms of whether you're in the hospital, whether your blood pressure is controlled, what your A1C is, and whether you're having fewer strokes. These are hard, hard outcomes.

It's amazing the way they do it because people talk a lot about how doctors interact with AI and what's the system. Is AI better? Is the doctor better? Who's better at diagnosing? And what they've said is, “No, no, no. The real question is: How do we integrate these 2 things so that everyone is doing what they're best at?”

So they have these new AI agents now who are driving some of the medication titrations. What they've found is that the AI will learn all of that doctor's practice patterns over time. And what they have now is 98% concordance with the prescriber and the doctor. When the prescriber looks at what they would've done, 98% of the time the doctor agrees with the recommendation that AI is putting forward for the prescription. So they're truly working syntonically together.

I think the other interesting thing is that they've found that they've had a 70% increase in getting ahold of folks with AI. What they've found is that when you measure your blood pressure, because the agents are always monitoring it, if your blood pressure is high and I call you an hour later to say, “Hey, your blood pressure was high. Are you feeling dizzy? Because if you're feeling dizzy, this might be a more serious thing, or maybe you need to take this other medication,” if I wait an hour, I might not get you because you might be playing bingo, you might be at Trader Joe's, or you might be playing with your grandkids. You're not going to answer the call then.

The key thing is that I need to call you right when you're taking the blood pressure measurement—

Right.

—because now you're in that mode. With AI, they can do that. They can intervene at the moment they get the signal. They can get the signal, process it, and go through your past charts to say, “Are you someone whose blood pressure just runs high?” A blood pressure of 160 for you would be a problem. 160 for someone who's always at 160 is something that matters, but we're not going to change your medication right away for that.

So they have that broader context, and I think that's the power of what they're building: They've been able to integrate your doctor, your EHR, and your healthcare system with their agents and the devices in the patient's home. They're actually able to bring those 2 worlds together.

That's incredible.

It's like, if you asked a 10-year-old, “What should healthcare be like?” I feel like Cadence is what they would draw.

It's so funny because all this stuff just seems so intuitive. It just should have already been like that.

It makes sense.

Okay, we're going to keep the doctor hat on, but we're going to put the investor hat on again. It's been on, but I want to really put it on now. So, talking throughout the portfolio, especially the investment thesis, I had this conversation with Annie Lamont. You were in one of the first waves of these huge exit environments for healthcare companies at Amazon.

What are we going to see now? Are we going to see labs acquiring companies? Are these big providers? Are these Epics? Are they still going to be in the mix of acquiring them? Where do we land with where the legacy industry is and how much they're going to adopt versus the competitive players on the market?

Yep.

9. Healthcare Companies Find New Scale

I think that is the magic of what we've seen with AI. We've seen it in core tech, where you have these amazing new companies that are growing up and able to be truly dominant standalone companies. But the AI wave has also pushed incumbents to compete in a way they haven't before.

Mm.

Right? We've seen Stripe making acquisitions. We've seen Google making acquisitions. We've seen all of the large entities across tech and fintech. They've been moving fast because they've realized they need to tool up in this new time, in this new period.

I think we're going to see a similar trend play out in healthcare, where we're both going to see existing companies have enough surface area and move fast enough that they'll be able to IPO and become large standalone companies. But we're also going to see that incumbents are not going to stand flat-footed, and they're not. Lilly launched Tune Labs with NVIDIA. A lot of these large entities are moving on their own. I think health systems are doing the same. You're seeing the large incumbents, the Oracles of the world, move fast as well.

Incumbents aren't flat-footed in the age of AI, and they're going to either try to build or buy the capabilities they need to compete in the new world, because everyone sees that the new world is changing. I think folks are also seeing that the attack vectors are different.

The same way that I talked a lot about how these biotech companies are seeing that, “Huh, you know what? Care delivery is a space that we can go into to expand our TAM. That's a logical extension.” We're seeing the same across different portions of the value chain. So there are a lot of these large distributors, the McKessons of the world.

Mm-hmm.

McKesson has gone out and bought large oncology providers, right? McKesson is a distributor of medications, right? They've gone out and bought large care delivery assets. That is a new thing that we hadn't seen before, but people are rethinking the lane that they sit in because they're realizing that, one, when it's easier to build, you start to move faster into these relevant adjacencies.

Talking about exit and the exit environment, there's one company—we'll come back to this conversation we had in the beginning—that's very important to Thrive, and that was Oscar Health. Oscar had a crazy comeback story. At one point, it was down 95%. So what happened with Oscar?

Yeah. Oscar, the first company that Thrive ever invested in. Josh co-founded Oscar when he co-founded Thrive. And I remember I was a know-it-all med student at the time, and I was talking with Josh and Mario about Oscar. I told them, “This is a bad idea. You can't. The last health insurer that was created was in 1977. That was UnitedHealthcare, right? Pre-internet. You can't just make a new health insurer. It's not a thing that's done. It's too hard. It takes too much scale. It takes too much capital.” I was like, “You can't just build things.” And I was very glad to be proven wrong.

I think, in a lot of ways, their whole vision—can we build this for the consumer—

Mm.

—using the ACA as the time to actually build a consumer-centric health insurer—was true, and it was amazing.

To your point, at one point it was down 95% from its peak value in December 2022. Now it’s back up above its IPO price. I think it’s increased 100% year over year, which is quite amazing because health insurers writ large have had a pretty hard last year and a half. And it’s called Oscar.

You could argue—I was telling Mario, I was like, “You guys should call it Alphabet, because you guys are kind of defying the gravity of health insurance writ large.” Part of that is just the way in which they built it. They built it with a tech-first infrastructure, and this infrastructure that they built a decade ago.

Mm.

They built their own claims stack, their own claims processing. But this infrastructure they built a decade ago has really been the thing that’s enabled them to move so fast with AI today, because they had a modern tech stack, and they had that tech DNA and builder DNA ingrained within them already.

So I know we went through so many companies. This is why I love talking to you, and we’re definitely going to do this again. We talked through so many companies. I want to double back down on OpenAI.

Yeah.

Because I think it is a bit underrated, their foray into healthcare and what’s going on there. So tell me a little bit more about that.

10. OpenAI Becomes A Healthcare Provider

Yeah. I think it’s interesting. You could argue in a lot of ways that OpenAI is now one of the largest healthcare providers in America, right? They have 300 million people asking ChatGPT questions every week about healthcare specifically.

Mm-hmm.

Right? In a lot of ways, that makes them the largest healthcare provider in the world. And I think when you think about that, it’s pretty amazing. They’ve not only ushered in the AI age writ large, but I think when it comes to healthcare, the idea that all of us now have access to a world-class physician in our pockets at all times—2:00 a.m., 7:00 a.m., 8:00 p.m.—we all have access to that knowledge at all times. That is probably one of the largest changes that has happened in healthcare in a long time.

And for most people, I think we talked about this again: healthcare is the key hero use case for most people, right? For Silicon Valley, it’s coding. For the rest of America, it is healthcare. Find me someone who does not have a story of themselves or their loved ones where they have put all of the information from some complex medical journey into ChatGPT to get an answer.

They partnered with Boston Children’s and the Harvard hospitals to take about 300 pediatric cases that were unsolved, and they were able to get to new diagnoses for about 20 of them. That’s quite—

Wow.

—amazing when you think of some of the capabilities. The fact that all of us have access to that at all times—

Mm.

—that is a paradigm shift for healthcare, and it’s one that people are using. We all know those use cases. And I think that’s the part that’s so inspiring: as you think about, one, the impact that has now is huge, but there’s also the impact of what that does going forward and how that changes the rest of the system, how that changes where the bottlenecks are.

Because if we can now understand more ourselves, we can treat more ourselves, what does that mean for where the healthcare system itself can focus its resources? Is it able to then deploy more folks to making sure that we have more accessibility for the procedures that we need?

Can we make sure that your average surgeon—they spend half of their time on nonoperative cases. What are—

And even now it’s wild because there’s a 3-month, 4-month wait for an orthopedic surgeon in LA.

Yeah.

Right? But if that surgeon was now seeing only cases that that surgeon needed to see, if you were able to optimize that supply and demand, all of a sudden maybe it means there’s a 1-month wait now rather than a 3-month wait.

And by the way, I’m happier because I don’t have to wait. The surgeon’s happier because they get to make more money just operating, but they’re also doing the thing that they’re trained to do. And I think those are the types of ways in which you think about the downstream effects of that kind of explosion of knowledge and access.

I often say, we talk a lot about the 10X engineer in tech. I’m like, how do these tools not just enable the consumer, but how do they enable the rest of the system? Because as a nation, I think we need a 10X nurse a lot more than we need a 10X engineer.

Incredible place to end it. Dr. Ayogu, thank you so much.

Thank you.

And I can’t wait to do this again. I’m sure we’ll have to do this again soon, because now the inflection point has happened—

Yep.

—and we’re hitting the ChatGPT moment of biotech.

Exactly. Exactly.

Well, thank you so much.

Thank you.