Tal Zaks——连接科学、医学与回报——[Invest Like the Best，第406期]

I know firsthand how complex the tech stack is for asset management firms. And seemingly every new tool and data source makes the problem even worse, adding more complexity, more headcount, and more risk. Ridgeline offers a better way forward. One unified platform that automates away the complexity across portfolio accounting, reconciliation, reporting, trading, compliance, and more, all at scale. Ridgeline is revolutionizing investment management, helping ambitious firms scale faster, operate smarter, and stay ahead of the curve. See what Ridgeline can unlock for your firm. Schedule a demo at ridgeline.ai.

Hello and welcome, everyone. I'm Patrick O'Shaughnessy, and this is Invest Like the Best. This show is an open-ended exploration of markets, ideas, stories, and strategies that will help you better invest both your time and your money. Invest Like the Best is part of the Colossus family of podcasts, and you can access all our podcasts, including edited transcripts, show notes, and other resources to keep learning at joincolossus.com.

Patrick O'Shaughnessy is the CEO of Positive Sum. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of Positive Sum. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum may maintain positions in the securities discussed in this podcast. To learn more, visit psum.vc.

My guest today is Tal Zaks. Tal is a physician scientist turned biotech executive and investor who served as Moderna's chief medical officer during their COVID-19 vaccine development, giving him an extraordinary perspective on one of modern medicine's pivotal moments. His combination of medical expertise, platform innovation experience, and investing acumen allows us to explore the interconnected challenges of turning scientific breakthroughs into viable medicines while generating venture scale returns. We dive deep into the lessons from Moderna's mRNA platform, examine how emerging technologies might reshape drug development, and the fundamental question of what it means to make a healthy person healthier. For investors, entrepreneurs, and anyone interested in the future of medicine, this discussion provides a window into both the immense potential and profound challenges of advancing human health. Please enjoy my conversation with Tal Zaks.

So Tal, maybe to begin, you could give us your 1-minute summary of your career, just so people have context for where you're coming from in all the incredibly interesting topics that we'll get into today.

In a nutshell, I'm a physician-scientist who spent all of my life figuring out how to translate the wonderful innovations of science in our era into better medicines for patients. I was fortunate to have been trained by some of the best. I trained for more years than I would recommend anybody, having done an MD, a PhD, a postdoc, a residency, a fellowship, and 2 internships.

I finally went out to do something in the world and spent the first part of my career in drug development, mostly as an oncologist in oncology drug development. Then I spent the second good chunk of time at Moderna as the chief medical officer, developing it as a platform as opposed to any one medicine. For the past several years, I've moved over to the investor side to help align the translation of science into medicine with an actual return on investment, realizing how important it is to get those 2 in alignment to do good in the world.

Your intersection there of investing and deep science and personal, hands-on work as a physician is exactly the Venn diagram that I've been searching for to have this kind of conversation. It's really about the current state and potential future states of medicine and therapeutics, and the investing returns that might be earned from paying special attention to those areas, which I think is the key distinction between the therapeutics themselves and the potential for returns.

Maybe you could give us the equivalent of a State of the Union on that topic. How do you see the world of medicine today versus your career? If you compare today's snapshot to everything you've seen across your working career, I think that would be a great place to start, just to give us context for where we are today.

1. Medicine Needs Returns

I think we're overall in a good place. I'm optimistic because the advances in science and technology have been so robust and amazing. That is tempered, I think, by 2 opposing forces. The first is that the translation of all that wonderful science into medicine is probably as challenging as it ever was in terms of the unpredictability of what makes a good medicine, and we'll come back to that.

It's also challenging in terms of the public perception and the willingness to pay for innovation, which I think are challenges in the current political environment that are dampening some of the prospects, or optimism, if you will, of what is possible. I was invited to give a short talk at one of the panels of the National Academy of Science, Medicine, and Engineering back in the middle of 2024. It was a panel that asked, “How do we better align investment and innovation with the unmet need?” So they asked me, “Given your intersection, what can you tell us?”

One of the points I made to them is that investment in innovation requires a return on investment. I stated this publicly. I said, “You guys use the term reimbursement. Reimbursement is what I do when I put in my expenses to get paid for a meal or a flight. We don't do this for reimbursement. We do this, A, because those of us in the trenches really believe in the mission, but B, because there is a need to return capital to shareholders.”

In a funny way, I've been with this VC firm, OrbiMed, for a little over 3 years. One of the executives from a large pharma company, who shall go unnamed, came in and knocked on our door. He wanted a role as a venture partner with us. In the interview, he told me about his background, and at a certain point, I sat back and said, “Let's call him John. John, what is it that you think we do here?”

John looks at me and says, “Well, you know, you do clinical trials, phase 1, phase 2, increase value.” He looks at me, sees my face, and says, “Well, clearly I didn't exactly answer your question, so where did I go wrong?” I said, “Look, on the tactics, what you're saying is right, but the answer to what we do here is very simple. People give us money, and a number of years later, we need to give them more money back. Otherwise, we're not going to be doing what we're doing.”

Framing the potential in terms of return on investment has become the focus of this phase of my career because, without it, all this wonderful progress will be for naught. It's not by chance, I think, that the modern armamentarium of medicine has come from the United States. I think it's a combination of the infrastructure we set up on the public side—the National Institutes of Health's basic research—as well as the infrastructure we set up on the commercial side and the ability to garner a return on investment.

The part that is often lost on the public and the people who debate this return is the very long tail of benefit to society that what we do brings. I happen to be on the board of directors of Teva, and I'm very proud of that because Teva is one of the largest and highest-quality generics manufacturers. People forget that once the brand price erodes and we're in generics land, that is now a benefit to society, all of society across the globe, at pennies.

My father, who passed away a few years ago, was diagnosed when he was 50 with his first MI, and he lived to his mid-to-late 70s. He did that because of a generation of drugs that, even in his time, were half generic, but today all of them are generics. The benefit that we bring and what we do is not just the short term during the patent period, where all the angst is about pricing—and it's legitimate, I get it—but somehow society also has to take into account the tail effect of what it is we're doing that is being left for the next generations.

What is your assessment of the potential for the world of therapeutics to explode in a similar way to what we're seeing with artificial intelligence and some other areas of technology? I'm curious to ask you this because you have exactly the right balanced perspective: an excited investor, I'm sure, but also a practical realist from your time designing and rolling out drugs at traditional pharma companies and at Moderna, which is a very neat platform that we'll talk about.

What do you think is that potential? Should we be excited for a thousand flowers to bloom and an explosion of progress in therapeutics, or do the realities and complexity of biology mean that it will be slower than the most excited people think it might be?

2. Biotech Defies Prediction

It's a question I've been asking myself quite intensely, as you can imagine, for the past few years. In fact, I joined the venture community because of the excitement around where that branch of technology has the potential to bring us. But one of the things that I think is often lost is the difference in the predictability and the nature of investment in biotech versus some of the other technologies, tech in its various manifestations.

The challenge is that it's still hard for us to predict what is going to work. If you look at what makes a return on investment, it's basically 3 things: how much money you are required to invest versus what you can get at the end, how long it takes you to actually get there, and what the probability is that you will actually arrive at that destination.

And that's basically what determines your return. Now, that's where biotech and tech are very different, and we can come back to that. But in biotech, you have to account for 2 variables that are extremely hard to predict. One of them is whether the biology will pan out, and the other is whether the pharmacology will pan out.

Biology means, is this protein that I think is involved in disease actually involved in disease? And the pharmacology is, okay, yeah, it is. Now I'm going to change that protein's function. Am I actually able to do it with the drug I have at hand? Does it get to the right place in the body? Does it have its effect? Is it tolerable from a safety perspective?

Those risks of biology and pharmacology are very hard to predict, and some of it has to do with the nature of the beast, if you will. It's not a system we designed, so we're constantly learning about it, versus engineers. But some of it—and this is one of my biggest lessons from Moderna—is also a different mindset.

Engineers and physicians are almost diametrically opposite in how they think, and I had the good fortune in my life to work with, I think, one of the most brilliant engineering leaders, Stéphane Bancel, the CEO of Moderna. What it taught me is that engineers have 2 things that physicians struggle with: one is vision, and the other is process.

So what do I mean by vision? Well, if you look around our life, take a look out your window if you live in the city. Everything you see is a function of the vision of people who came before you. It would not otherwise have been there. That's why we love nature so much, because it's uncluttered by the visions of other people. But if you live in any other environment, you're constantly faced with a vision.

So you have to ask yourself, “Well, what's my vision for the thing I will leave for the future?” Then to enact it, okay, requires a process. So what does it actually take to get there?

Here's an interesting thing. If you ask an engineer an engineering problem in their domain, say, “I want to get to the moon. Is it possible?” anybody who knows the domain will tell you, “Yeah, it's possible. Here's what it's going to take.” They can even map out the resources and the likelihood of getting there.

Well, you ask a physician, “Hey, I want to cure cancer. Is it possible?” And the answer is, “I don't know. I mean, I can think of some approaches, but we've got to go try them.”

One of the first things I did when I joined Moderna, and probably the scientific thing I'm most proud of in my time there, is actually not the COVID vaccine. It's something called a personalized cancer vaccine. We've learned enough about cancer to know that everybody has a different cancer, everybody's immune system is different, and we know that mRNA is a phenomenally good platform for making vaccines. So can we immunize people against their cancer?

To do that, we'd have to figure out what each individual needs to get immunized against, and so it would be personalized. That's a very challenging, complicated task. It's expensive. So we set out to do that, and as we were starting down the path, I remember we had a phase 2 study that was running. That's basically a study where you give half the people your vaccine and half the standard of care. The question was, are we actually going to be able to prevent some cancers from returning relative to the comparator?

Stéphane used to stop me in the hallway and say, “Tal, is this thing going to work?” And I'm like, “Stéphane, I honestly don't know. What I know is I feel I've been put on Earth to do the experiment, and we've been fortunate to align the technology and the resources, including the financial ones, to run the experiment. But I'm not going to predict whether the experiment's going to work.”

People have been trying to do cancer vaccines. I've been trying to do cancer vaccines since I was a postdoc, and so far we haven't. I think this one has a chance of working, and here's why, but until we test it in the clinic, we won't know. We've been fortunate that that phase 2 actually read out with quite a successful readout last year. In fact, the phase 3 is now enrolling, and hopefully we'll have a personalized cancer vaccine on the market before too long.

Sorry, so just so I understand that one, which sounds incredibly exciting: I could get tested—blood work or something else, or genome, or something, some combination of things that's unique to me—and then a vaccine would be created to reduce the odds, or eliminate the odds, that I would get the cancers I'm most prone to get?

So that's still in the future. The way this works is with people with early-stage cancer. Specifically, we started with skin cancer. Cancer has been diagnosed, but it's early. Likely, a surgeon will cut it out, and then there's some probability that it will come back. What can we do to improve the odds that it won't come back?

I see.

In that context, at least as far as a randomized phase 2 goes, when you give a personalized vaccine to those patients, it cuts the recurrence rate of cancer by about half—by about 50%—which is quite significant.

Going back to this, what sounds like the key bottleneck, which is that we just don't know how to predict whether or not something will work: can you break that down further? I'm trying to wonder if there's a version of the future where it does become easier to simulate, model, and predict—more like an engineering challenge versus a complex biological system challenge that we can't predict.

Practically or theoretically, what is in the way of us predicting things better and therefore drastically improving the efficiency of our efforts?

3. AI Improves Drug Discovery

I'd break it down to 3 phases, and I do think that the world is improving. The first phase is: is this target for intervention actually relevant to the disease? That's basic biology, and I think we're making great strides there in understanding biological processes. Some of it is big data, some of it is just old-school grunt work, but there are a lot of tools that have been developed and are being deployed that are making it much more accessible for us to understand disease better.

The second has to do with what's called drug discovery. Okay, so I've got a protein whose function I want to alter. What is the ability to actually discover a new chemical entity, a new protein entity, or a nucleic acid entity that will actually interfere there and do the pharmacological effect?

There, I think you're seeing a very significant deployment of these modern AI tools across the industry now, and it's very quickly becoming, to a certain degree, commoditized, with the advancement of AlphaFold predicting protein structures and people applying the same kind of tools in the chemical discovery space to come up with new chemistries. I've seen people even apply these tools to figuring out these lipid nanoparticles that will shepherd mRNA into different tissues, to come up with better formulations and ways of bringing that medicine into the right places in the body.

So I think drug discovery is getting a leg up, and a significant one, from the various applications of AI tools. The part where we're still behind is in what's called development, or clinical development—putting it in people. There's no shortcut here. We don't have a holistic model of a human being.

We have made progress in understanding what natural outcomes are for people with high-quality, big data sets to which you can apply machine-learning tools. So our ability to predict outcomes on the control arm is getting better, and people are leveraging those tools to make trials shorter and smaller.

But in the end, the gold standard is and will remain: “Okay, I have to put it in people and see how those people respond, and make sure that the people who get this have a better outcome than the people who don't, and I know I've got a new medicine on my hands.” I don't see the tools that we have today changing that in the near future. Now, is there a version of the future where they will? Yes, absolutely, but that will require some level of evolution of our healthcare system.

Could you just describe the basic investing process for you when you're looking at a company and a product? I know it's quite distinct from me looking at an AI software application or something. What are the key variables that, in most investments you're analyzing, you're looking at? Is it typically just a team that's going after one specific target or one disease? What is the nature of an atomic unit of an investment or an investment diligence process?

4. Therapeutics Investing Follows Milestones

That's a good question. For us, again, I term us sort of as the collaborative spectrum of venture capital. We will invest at any stage, starting from seed all the way through public PIPEs. So we've got a pretty flexible mandate.

For us, I think the key elements as we look at an opportunity are: what is the thesis that this team is trying to pursue? How much capital is required to get to the next value inflection point? What is that next value inflection point? What is the likelihood of achieving it?

If it's already a drug in the clinic, okay, what's the transition? If it's an early bet on a target, what's the likelihood of that target panning out? If it's a different strategy, then we look at that. So it will depend on the stage of investment, but ultimately it boils down to: what is the thesis this team is trying to prove for making the world a better place? How soon will we know that we've achieved some milestone that people will recognize as a value inflection—the value created?

One of the most recent opportunities I had to walk away from was a vaccine where it was an interesting idea and an early concept, but it became clear that the phase 1 data probably weren't going to be de-risking. You were going to have to take it all the way to phase 3 to know.

Okay, now you’re asking me for a quantum of capital and a time horizon that is just not commensurate with the risk involved. I have to see my way to building value incrementally over the life of the investment and the additional capital that will be required. So it’s got to be the content—the drug, the theme, the scientific content. It’s got to be the capital requirement, the value, and the people.

What is the talent around the table? What are the missing pieces? Am I going to be able to help you uniquely from my experience to do that? Because I’m here not just to deploy capital, but actually to deploy my experience as well. Granted, it’s at the board level, but nonetheless, I’m not going to be making investments in fields that I know zero about or that I’m not interested in leveraging what I do know. So I think these are the 3 elements.

Now, in terms of process—and I don’t think in that regard we’re very different—it’s really interesting. Being a venture capitalist, a partner at a firm, you’ve got 2 very different sides of the same coin. The first part of the job is deciding where to deploy capital, so that’s the diligence process. You look at an opportunity, you’ve got a team of junior folks helping you out, you bring in external experts for the stuff you don’t know, and it takes 4 to 6 to 8 weeks. Then you come to a decision on whether that’s worth the investment.

If you decide to invest, the coin flips. Now the diligence team’s not going to help you. They may dive in and out once or twice, but actually you’re there on a weekly, quarterly, monthly basis with the CEO, with the management team, and with your other board members. That’s why, for me, looking at the syndicate of who my partners are for this investment is so critical, because this is a long-term commitment.

There’s a lot of capital. It’s not over, and it’s likely not even over with this investment round. They’re going to need more capital down the road. I’ve got to make sure I’ve got a reserve. I’ve got to make sure I’ve got other colleagues around the table who are also going to be able to be in it for the longer run. So making sure that syndicate is the right one, that I’ve got colleagues around the table on the board who bring complementary experiences and expertise to what I bring, so that we can help the CEO and this young management team actually get there—those are the other elements that then become that second part of the job.

Patrick O’Shaughnessy

It sure makes me wonder about the role of capital that is not rational, whether that’s government or otherwise, that’s willing to fund things that sound like that vaccine story you mentioned, where you’re not going to get the appropriate readout until phase 3, and it just doesn’t make sense to provide them with capital. It sure makes me wish someone in the world, whether it’s philanthropies or otherwise, were intentionally funding things that are irrational explorations. That’s an interesting takeaway from your summary.

When I look at investments, I will often tell the folks, “Look, there are 3 pockets of capital in my book. There’s government, which is what’s there for the greater good, and I hope those investments are rational. There’s philanthropy, and then there’s us.” I’m not in the business of philanthropy, and I’m not a government, so for me, the lens has got to be what it is. But there are investments that are fit for philanthropy, and there are investments that are right for governments to make because there’s not an immediate return. Either the capital scale or, more often, the time scale and the risk are not commensurate with the money I need to return to my LPs.

If you were to think about the changes in general that would most accelerate our ability to identify a target, understand a disease, and create a therapeutic, it sounds like it’s some mix of, if we could somehow test these things not in humans but in simulation, that would be one big one. The healthcare industry itself would need to change in key ways. Do I have that roughly right? Are those the 2 things that most impede our ability to go faster?

You do, and I think your roundabout way of asking the question is coming back to what I think is one of the central tenets that makes this endeavor so different, which is that the number of different disciplines that need to come together to solve this is quite wide. I’ve not found many, or maybe any, other industries where you need so many different types of expertise around the table to solve the problem. I spoke about physicians: You need biologists, you need engineers, and, of course, you need the financial understanding of how this works.

The number of disciplines that actually need to come together is vast, and frankly, it’s been one of the fascinating threads of my career. To the degree that I’ve been successful, I think it’s because I’ve always been curious—not to approach it as the physician in the room, but to try to understand the language of the other functions that need to come together. If you look at pharma, a project team that tries to make a drug will usually get kicked off around a target, and if they’re successful, 10 years later you will have a drug.

But if you look at that project team, you will rarely find anybody around the table who was there at the start, and yet the thing works. So the great success of the pharmaceutical industry has been to build these multidisciplinary teams and enable them to progress this very complicated process. Yes, you can cut it short in time if you have the right technology, and we can come back to mRNA vaccines as the obvious example of that. But ultimately, you do need these different disciplines.

Sometime during med school, I chanced upon a really interesting book by a Canadian philosopher called John Ralston Saul. It’s called Voltaire’s Bastards, and the subtitle is The Tyranny of Reason in Western Civilization. He made the point that ever since, quote-unquote, “the Age of Reason,” we have mistakenly assumed that science in itself and reason will be an ethical force for good, which, of course, it isn’t. Science and reason are science and reason, but it’s the humanistic moral compass that needs to frame how we deploy them.

The other point he makes is that in this age of reason, the way that experts have survived and grown is in silos, and that silo is enshrined in language. Every discipline develops its own language over time, and it becomes a barrier to common understanding. If you’re trying to solve a multidisciplinary problem, you very quickly realize that one of your greatest barriers is language. The finance world knows this very well. We coined the term “Fed speak.” That’s clearly an example of language as a barrier constructed on purpose.

But we all fall into the trap, and what I realized early on is that physicians are probably as guilty, or guiltier, than anybody else in doing that. I’ll give you an example. Back in the day, if a patient came to me and his platelet counts—these are the blood-clotting elements we have in our blood—were low for reasons that the physician didn’t know, it would lead to a skin rash because when blood doesn’t clot, it will accumulate under the skin. So people would show up with a skin rash.

The patient would come to the physician, and the physician would look at him and say, “Hey, yeah, you have idiopathic thrombocytopenic purpura.” Oh, God. That sounds scary. Now, what did I just say? All I said is, in Latin, you’ve got low platelets and a skin rash, and I don’t know why. Idiopathic means “I don’t know why,” purpura is a skin rash, and thrombocytopenia is low platelets. So just by using Latin, we made ourselves sound smart, but the physician had no clue what it was.

Today we’ve learned a little bit about the biology, so it’s not as bad as I make it out to be, but you get the picture. When I came into the industry, I was always super curious to understand the language of the other functions that had to come together to solve this problem. Yes, I’m the physician in the room, but the people doing the biology, the chemistry, and the manufacturing are just as important.

By the way, one of the reasons I’m now in the investor seat is because I realized from my time at Moderna how critical it is to understand the language of investors and the expectations that they have in order to align across this mission to be successful. It is a multidisciplinary challenge, and that’s not going to change. We’ll be able to make it more efficient, yes. These tools are already making an impact in drug discovery, and as we integrate and have better and better ways of looking at human data, I think it will also make a dent in development.

But it’s interesting to sit back and figure out how we can actually radically change it. I don’t know that I have a good answer, although I’m trying, and this is especially germane given the experience we had with COVID.

One of my favorite books that I read last year was For Blood and Money, about the development of 2 cancer therapeutics. One of the takeaways from reading that was the story in which the primary investors made incredible amounts of money. But the story was quite chaotic. The things that had to go right were many, and many of the bounces of the ball seemed just crazy and random and impossible to predict.

Sheer persistence, a lot of luck, and a lot of factors had to come together for that group of people to make a lot of money. A couple of people were a few degrees away from making a lot of money and made very little. I’m curious, with stories like that seemingly all over the therapeutics world, what the best therapeutics investors do that bad investors in therapeutics don’t do.

If there are stories like that happening everywhere, it’s tempting to think you just have to get really lucky versus be really good.

So what is the combined role of skill and luck in making money as a therapeutics investor?

5. Success Needs Room To Adapt

I joined one of the best VC firms I could find, OrbiMed, because they have a track record of actually being successful, and I wanted to learn the answer to that question. Luck certainly plays a role, but here are a few thoughts that I think are not as obvious.

We all recognize the importance of learning from failure, and I listened to your podcast with Jared Kushner the other day—his version of “What is God trying to teach me?” I thought that was extremely well put. For many, many years, my favorite quote of all time was Nelson Mandela’s, who said, “In life, I’ve either succeeded or I’ve learned.” Coming out of the COVID success, it actually made me realize that’s a misguided quote because it suggests that you haven’t learned as much from your successes.

I would argue that the best investors actually learn more from their successes than their failures, and that’s not trivial. When you fail, there are a whole bunch of things you can point at as causal elements. But when you succeed, as you say, what was it? Was it luck? Was it talent? Was it getting the right people around the table? It’s a combination of factors, and I think the good investors develop this sense of pattern recognition of what works.

As I’ve tried to uncover it for myself, frankly, out of my own curiosity, I think in the domain in which I function—which is venture capital, and it’s very different from private equity or some other corners of investing—for venture capital, and specifically for what I call collaborative venture capital, which is the type of venture capital we have, which is to say that, yeah, we will often seed companies, but we very quickly look to syndicate deals. We look to work with other investors. We look to broaden the investor base but also broaden the competencies we have around the board. For me, the answer has been a combination of the talent of the people and the content that we believe has a leg to stand on.

When I call ourselves collaborative venture capitalists, it’s because, at least for me, it’s always critical to look at the talent that comes together as much as it is the content. Yes, I get excited by the science of what they’re trying to solve, but I have to get excited about the people who have the experience and the wisdom to navigate and understand what it takes. Because, as that book points out, and you correctly point out, it’s almost never a straight line.

If it’s not a straight line, it means that you need people around the table who have the experience and ability to look around corners. It means that you need enough of a capital and strategy structure to give you some degrees of freedom of movement. I can tell you that early on in my career as an investor, I made an investment in a small company. It was a very rational thought: it was for a certain idea of a drug that would have a certain effect against cancer, and it was a small team, and it was very linear.

It didn’t pan out, and one of the lessons I took from that failure is that this was probably not well enough funded or given enough opportunity—with the team structure, the financial structure, or the scientific structure—to have those degrees of freedom to adjust. If you look at investments that have been successful, I think you’re giving them some degrees of freedom, and you’ve got a management team that has the right functions to be able to do that.

For me, it’s another learning from Moderna. If you go back and look at the formative years of Moderna, the strategy was always to develop this technology of mRNA, understanding the engineering potential as much as the medical and biological potential. In fact, in the early days, we were more certain of the engineering benefits than we were of the medical benefits, in the sense that it wasn’t clear whether this would be optimally used for rare diseases, oncology, vaccines, or something that we had yet to discover.

It was all an interplay of delivery, medicine, and biology. These were the big risk factors. Once you could engineer the mRNA, you could do it again and again and again and again. You could do it reproducibly and cheaply. That was clear; that was the engineering benefit. But what kind of medicines could you make from it?

I think the brilliance of Stéphane and that initial team, before I joined, was to set up enough degrees of freedom to go and explore those opportunities in parallel. When it was clear early on that a vaccine was the straightest shot to proving the pharmacology of this technology, we went after vaccines. We didn’t drop the other elements, but we made sure to have the right capital-allocation strategy.

I remember that, in the early days, the loudest conversations we had around the executive table were about the relative capital deployment of these various applications. I was fortunate to work with some of the smartest people I’ve ever met. I’ve already mentioned Stéphane, but at that time we had 3 physicians on the executive team: Lawrence Kim, who trained as a physician but then became a finance executive at Goldman Sachs and joined us as our CFO; Stephen Hoge, who’s the president of the company to this day, a physician who then spent years consulting with pharma and joined as the president of the company; and Stéphane.

It was a very robust dialogue about where we should invest the marginal dollar and where we saw this technology panning out. To take agility to the extreme, in December 2019, when somebody started coughing in Wuhan and that first Wall Street Journal article came out, the person in the company who first picked it up was the head of vaccines. The next person who actually took it and ran with it was Stéphane himself.

He saw right from the start the importance of chasing this. It wasn’t clear to the rest of us, frankly, whether this would peter out like prior vaccines. If you remember January, February, and even the beginning of March 2020, common wisdom would have been that this was going to peter out and not be a big thing. Stéphane was absolutely operating in a very different mindset. He saw the opportunity, and with his own will and management team took the company there. The rest, as they say, is history.

Obviously, you’re part of one of the most central medical episodes, certainly in our lifetimes and also in recorded history, in the development of that vaccine and rolling it out. What is your postmortem on that whole process?

I think parts of it are exactly what everyone wishes: something happens, and we’re able, with technology and a prepared platform in mind, to address it incredibly fast. There are all the stories about how quickly the vaccines themselves were developed, and then most of the time delay was just testing them. That seems amazing. Everyone wants more of that.

Then there’s been maybe a hangover of side effects from the vaccines, which I personally don’t understand in any great detail. I’m sure you do. We all wanted these rushed because we wanted to get back to our lives and not have loved ones die.

I’d be curious—this is such an interesting real-world experiment of moving very fast to address a huge problem—what your postmortem analysis is of the whole thing, having been at the dead center of it.

6. The Vaccine Speed Advantage

First of all, we were very well positioned at the beginning of 2020 to get there. What most people don’t realize is that by the beginning of 2020, Moderna had already tested the ability of mRNA to generate neutralizing antibodies in humans against 8 different viruses. COVID was to be our ninth, and our success rate was 8 out of 8. That is unprecedented in drug development.

It’s a function of the platform nature of this, and so if you get the antigen right—that thing you’re trying to immunize against—you’re going to hit. We were very well prepared, and the other thing that helped us along was that we had already been collaborating with the NIH before then.

The government got to know us—BARDA and the NIH—during Zika. People forget that, but in 2017, Zika was the big scare. We partnered with the government and started to develop a Zika vaccine. By the time we got there, it was of no interest, but at least the NIH took notice of the rapidity and potential of this platform.

I’ve told this story before, but in September 2019, Stéphane and I went down to see Tony Fauci and the NIH, and we were talking about the latest vaccine. It was something against cytomegalovirus. Dr. Fauci looked at me and said, “So if what you’re telling me, Tal, is that you’ve got the best vaccine platform I’ve ever seen—” And with a bit of chutzpah, I said, “Yes, sir, and let me tell you why.”

The outcome of that meeting was 2-fold. Number 1, the NIH wrote a paper. They published it in 2019—you can find it online—with Dr. Fauci and his team citing mRNA technology as a leading platform in our readiness for a pandemic. This was November 2019.

The second thing was that we had agreed with the NIH that we would run a demonstration project in which the NIH team would pick a virus—something nobody had ever heard of and that we hadn’t sequenced before. They’d send us the sequence, we’d make a quick batch, they’d test it in a phase 1 trial, and we’d start the clock to see how fast we could go.

That was the outcome of that meeting, and then suddenly someone started coughing in Wuhan in December. Two months later, the rest is history. We were well prepared as a company with a platform.

There was a recent Economist piece on personalized cancer vaccines and how they all benefited from COVID and mRNA vaccines. What people don’t realize is that the opposite is actually true. When we set out to do the personalized cancer vaccine, in order to treat somebody with cancer with a personalized dose, you have to make a small batch and have a very quick turnaround time because people with cancer can’t wait.

And so we had set out years before to build a manufacturing process that would allow us to churn out a small batch in a rapid turnaround time. Guess what? That same-size batch is exactly what you need to run a phase 1 trial with just 2 doses, because a cancer patient is going to get doses for 6 months every 3 weeks. And so that served us well to be prepared to move so quickly.

So that's what enabled us to start quickly. But how did we get to the endpoint quickly? Well, that was really 2 factors. The first was the private partnership, and this had to do with the NIH, the CDC, and the FDA under Operation Warp Speed actually getting the government's act together, giving us guidance, and working with us to be able to move quickly.

What usually takes weeks and months of back-and-forth with the agency was literally a week. Agreeing on a protocol, agreeing on endpoints, agreeing on trial design, et cetera—that was all done very expediently, in fact, at a pace that it takes a global pandemic to get done. And that also enabled the investment in manufacturing and the scale-up in manufacturing that are usually done later and in a much more circumspect environment.

By the time we had to scale up the investment in manufacturing, there was a period of several months when it was challenging, and the company actually had to foot the bill and go raise money in the capital markets. People think the government funded it. No. It was private capital, if you go back and look at the history, that actually funded the commercial expansion.

What the government did at a certain point was backstop the investment and say, “Okay, we'll pay for these doses.” But the capital actually came from the private sector. And then the final factor was actually, well, to test a vaccine or any medical intervention in a phase 3 trial, you need a certain amount of events to happen.

So you give half a population your vaccine and half don't, and then you see, well, what are the event rates? And this is where modern technology actually helped us. We were looking at the spread of the pandemic as we were rolling out the phase 3 trial, and we could predict weeks ahead where hotspots would be. We used those predictions to go and open clinical trial sites in places where we knew that event rates would be high.

And if you look at the original design of the trial, the number of events that was expected to take up to a year actually occurred within 3 months. So because we were able to target the trial to where the events were happening, we could get the answer quicker and be ready with a vaccine. All those elements, I think, are what enabled that unusual success.

Now, I do have to mention one word about safety here, and we'll come back to the public-perception question. These mRNA vaccines have been the most well-studied medical intervention in the history of mankind.

Hmm.

Full stop. I can tell you, as the one responsible for setting up the collection and analysis of the adverse-event data, that we had thousands of people poring over tens of thousands of adverse-event reports. We had given this vaccine to probably, at the time, millions and counting as we were rolling it out. It ended up being billions.

We collected the safety data with a rigor and breadth that had never been done before in human history, using tools that had never been available before in human history. The event rates in terms of side effects that people were reporting were about 10 to 20 times higher than had ever been reported before with even prior vaccines. The reason was, A, the public was fearful—it was a pandemic. B, this was all quick; nobody had heard of mRNA. And C, all the government entities were pushing out tools to report safety.

So this was a real concern, and we had to stand up systems to capture all of that. Now, the proof is in the pudding. If you remember some of the other vaccines that ended up getting on the market and then getting pulled, and people don't use them anymore—the adenovector vaccines, whether it was the Johnson & Johnson vaccine or the AstraZeneca one—we discovered event rates that happened at 1 in 150,000 cases, and they were discovered within weeks to months. They were assessed as true and relevant, and they were quickly added to the label.

We even discovered this rare myocarditis finding in young adults. It was reported and appropriately put on the label. Still, by the way, the risk of getting the virus is much higher, even just for that side effect of myocarditis, but it appeared to be related to the vaccine, and it went on the label.

And so I've never been as sure of anything in my life as I am of the safety profile of these vaccines because of the data and the way it's been collected and analyzed. And look, when you're the executive on the other side of it, A, this is your life's mission to get it right, but B, if you don't, you go to jail.

I mean, this is as important as it gets in terms of getting it right. And believe me, all of us sitting on that side of the table were super conscious of the responsibility to make sure we got that safety profile right. Now, where did we miss a beat? It was probably in terms of the public backlash.

And as I stand back and look at that success, again, both on the efficacy side—I mean, people have done the math. These vaccines have saved millions of lives, full stop. Anybody who believes in reason and science will agree that that's incontroversial.

But where did we fail? Because obviously, the sentiments I'm expressing here are not uniformly shared by the public. I even have people in my own family who refuse to take the vaccine. So I think it's a question of trust in institutions, and I think it's also a question of probably scientists, myself included, overstepping our boundary in a sense.

And it goes back to the point I started with of the misguided philosophical framework that we operate under, which assumes that because it is science, it is good. Science per se is not an ethical framework or a force for good. People who deploy it for good have the obligation to explain why their moral framework is what it is, and that abuts against other ethical and moral considerations having to do with people's autonomy, freedom, and other aspects that also have ethical and moral perspectives to them.

Our society is a balance of these different forces. And as a democracy—and I'm very proud of our democracy, and I'm a proud immigrant into this country for both the opportunity and the democratic processes we have—I also have to recognize that that means the public gets to say and vote on the things that have ethical and moral frameworks.

Forcing people to take vaccines is an ethical and moral consideration that the public has to buy into, and it's legitimate that they don't. And that's a hard and messy thing in a democracy. In a dictatorship, it's easy. And yes, on average, in a dictatorship everybody gets a vaccine. No question, they'll be healthier.

That being said, you now have a balance of 2 opposing moral forces, and that's a reckoning that has to happen in the public square. It is not the scientists' or the physicians' purview to dictate that. And I think that's where we probably overstepped, and I think my biggest learning has been the importance of open and honest dialogue with the public on what it is that we do.

It's an incredible recounting of a crazy episode in our history. I like your balanced take a lot. Obviously, it's above my pay grade to have some long, drawn-out conversation on the pros and cons of these things, but I do think it highlights how complicated and nuanced all of these issues are when you really start digging into them.

And it makes me wonder a little bit about the future of applying these platforms to other things. Maybe we won't have something on the scale of billions again, maybe ever in our lifetimes, or maybe we will. But as you think about something like an mRNA platform, I'm curious what other—I’ll call them technologies or platforms like mRNA—you think are the most important to address disease in the future. What do these things now enable for us?

You were 8 for 8. That's so cool, but 8 is small. What do you think the future is of mRNA specifically, and what will it change about the way that we deal with our health in the future?

7. Nucleic Acids Expand Medicine

I will broaden it out to nucleic-acid medicines, if you will, of which mRNA is an important part. There are other types of RNA, most notably siRNA, and I think these medicines are enabling new types of pharmacology that we haven't had before.

They need a lot of investment in technology and delivery and getting them to the right tissue, but they all benefit from the same thing we benefited from in the vaccines as a platform, which is that once you get the first one right, the marginal cost for the next one is a fraction. Nucleic acids have that unique property of being almost software-like. We used to call it, in the early days of Moderna, a software-like platform because basically it's the same physical construct of a drug, but you change the information in the nucleic acid, and you get out a different drug or a different vaccine.

But actually, under a microscope, it looks exactly the same. In fact, it's the same components. It's the information they encode that makes the drug. So these drugs as information—that's a new concept.

And then there's a whole world of opportunity in what's called gene therapy. I started by mentioning rare genetic diseases, where somebody is born missing an enzyme or a protein, and if you can actually just put in the information that encodes for that protein, you fix the problem. And so now physiology should reassert itself as the normal physiology of any other person.

You can do that in mRNA with the traditional approach. Moderna is doing that for rare diseases, but then that mRNA is transient, and you need to redose every time.

But if you could actually use that mRNA as an intermediary to get in and fix the genome once and for all, then you would potentially have a long-term beneficial effect without having to redose for the entire lifetime. The company where I've been spending the last year as acting CEO, Exilio, is doing exactly that. It's taking this mRNA in a lipid nanoparticle, which is conceptually the same as Moderna's vaccine—not the same chemistry—but the information encoded in it is actually allowing that mRNA to insert itself into the nucleus and make a change forever.

Now, that is going to bring a whole slew of new challenges. One of the criticisms and the public backlash against mRNA was the mistaken conception that it would somehow change the DNA. The truth is, mRNA does not. It can't based on first principles, and there's no empirical evidence that anybody's ever been able to show, because it can't.

So mRNA doesn't do that. But there are ways to encode different information with mRNA where it will do that. Now we're very intentionally changing somebody's DNA to have an effect that will last for their lifetime. There are some people who are going to find that challenging based on their ethical and moral framework. That's legitimate.

There's a concern about what this means for future generations and making sure that we're doing it for the individual, not with unintended consequences for their offspring—or maybe sometimes with intended consequences. That will raise a whole slew of new moral questions and ethical debates that we will have to resolve before this is widely acceptable. But at the end of that, if you can have a medicine that you take once, twice, or 3 times and then you're cured for the rest of your life, as opposed to having to take an injection every 2 weeks, I think that's huge progress for those patients.

I wonder if you could—this is meant to be a little bit fun, and no one will hold you to these predictions—but if we think 5, 10, and 20 years hence, to pick 3 time periods, what sorts of things do you think may be happening or possible in terms of the way a single person manages their health?

As an example, I've always been interested in this idea of fairly constant, ongoing data collection for myself—on blood work, scans, genetic information, or whatever input data might help identify things early and suggest ways that I live a healthier, better lifestyle. I think basically everyone wants to live longer and better. Today, that's incredibly crude. Maybe the best-in-class people get their blood work once a year, pay attention to it, and know what it means, but almost nobody does. How do you think that general state of how people approach this problem will change in those 5-, 10-, and 20-year time frames?

8. Health Care Becomes Preventive

There's an old saying in the Talmud, which is my ethical framework given my religion: Since the time of the destruction of the Temple, prophecy has been given to fools. To the degree that I'm going to give you any prophecy, that makes me a fool, but to the degree you listen, well, that's on you.

5 years is probably the easiest. I'm of a certain age and generation and fairly conservative in my views, so up until now I looked at all this data collection and thought to myself, “Everybody knows what it means to eat healthy and exercise more, so getting more data to tell you what you already know isn't actually going to change much.” But I'm actually coming around, because I do agree that I think there are a lot of differences between people in terms of what the right intervention is for them. I think my generation of medicine is guilty of peanut-buttering the effect across a population without really understanding the chunkiness of it.

I actually told my wife the other day that this is the year for us to start collecting data in a more thoughtful manner and at a much deeper level than we've done in the past. Like most people, I go once every year or two, get the usual blood work, take my statin for a bit of high cholesterol, and think I'm done. But I think there's a whole lot more that we're going to learn that we can do.

It is amazing to me. One of the triggers is the recognition that when you go and get blood work in a hospital, the thing that I find very depressing is the fact that if you go into an emergency room today and get blood work, you're going to get the same panel that you would have gotten when I was a medical student 40 years ago, maybe with 2 additional analytes that have been added in the last 40 years. I find that so depressing. With all the progress we've had, it's the same blood work.

Now, here's an interesting fact. There's a brilliant AI scientist at the Weizmann Institute called Amos Tanay. He spent most of his career designing better proteins to make drugs and things of that nature that you would associate with AI. But he was curious to understand how much of that regular blood work variability can be trimmed if we understand you as a person better.

He teamed up with one of the Israeli healthcare networks to collect information on 250,000 people over a decade. He took that same blood work and asked himself, “If I correct somebody's blood work for other parameters that the machine measures but I don't even know what they are, can I reduce the variability? How much of the variability in what's considered normal is actually true variance versus things we could correct for that are based on you?”

It turns out that about half the variance in the distribution of what's considered normal is actually false variance. If we could correct it relative to your factors, it would make that distribution much narrower, which means that we could now pick up things that otherwise look normal. Based on all those other things that the machine can figure out, this is a perfect application of AI, because it can start to pick up abnormalities much sooner.

Now, the question is, does that lead to interventions that we know will be of benefit sooner? That's where a whole world of drug development has to come in. But that's where I agree with the folks who have been saying that we've got a sick-care system as opposed to a health-care system.

Look, my greatest learning as a physician from the COVID vaccine—I started this journey as a medical oncologist. Now, what's a medical oncologist? Somebody who deals with somebody who's got cancer. What's my vision for a patient who comes into my office? We started talking about vision. My vision for that patient is their past. Think of that. The best I can envision for somebody is to be as healthy as they were before they had cancer.

Now, think of a vaccine. What does a vaccine do? It actually takes a healthy person and makes them healthier in the future. A serious medical doctor like myself in the previous era, if you told me that I would be interested in spending time making healthy people healthier, I would have looked at you strangely and said, “Yeah, no, the people who do that are the yoga instructors, the quacks, and the nutritionists. I'm a serious doctor. I don't make healthy people healthier. If you're healthy, get out of my office. I've got sick people lining up.”

I don't do that. But actually, that's what vaccines do, and that's what this new age of technology is going to enable us to do. Now, how do we deploy that information? I think in the 10-year horizon, you're going to see some of those insights now becoming institutionalized.

I think it's going to take a long time, because it will bring a profound change in what it is to be a physician. I think that change is already occurring. If you look at the last 10 or 20 years in the US, we talk about physician burnout. Physicians have gone from having agency, making decisions, spending time with the patient, and earning a return on their time to now being employees of large health-care systems, whereby the treatment they mete out is a function of algorithms, the quality control is a function of algorithms, and the reimbursement—what they can actually prescribe—is a function of what the health-care system will tell them.

By the way, they're all employees of the health-care system to begin with. All of the agency has been taken out. The time that they're able to spend with people has been trimmed and trimmed and trimmed. The role of what it is to be an internal medicine doctor has changed.

I think that with the system of knowledge that AI is gearing up to be, it will continue to evolve, and the role of the physician will change into being the translator for people of some of this wisdom, as opposed to the person who is entrusted with having all that wisdom in their head.

My kids are 10 and 8, so they're young, and they're going to be the beneficiaries even more than you and I are of so much of this science, discovery, research, and product building. Do you think that, in their adult lives, they'll look back on this time and think that it was almost barbaric how we discovered, identified, and dealt with disease?

Is that degree of radical change coming? Might we look back on the time before vaccines or something and just think, “Can you believe that people would die from this stuff? How crazy is that?” Infectious disease and otherwise, which no one dies from anymore relative to the past. I'm thinking about the 20- to 30-year horizon now.

I think in the 20- to 30-year horizon, they will experience health care very differently than we are experiencing health care. I think what it means to go into the hospital, what it means to sit with a physician, and what it means to get diagnosed and treated will be different. I think it will have more interventions far earlier. I hope that there will be an emphasis on maintaining health and promoting health, as opposed to treating illness.

Illness is always gonna be with us, and we will need to evolve our tools for treating it. But yes, I do believe it’s gonna be different. That being said, we are also living in a time where—and this is what makes my unique job so fascinating—technology today can do more than our ecosystem has figured out how to integrate. I say this both as an investor and a consumer. There are programs that can scan a pathology slide and give you a more accurate reading than a pathologist, hands down.

They’re FDA-approved. But the problem is that less than one slides of any tissue micro section ever gets read by a computer. We still have microscopes and people looking down the microscope. Who’s gonna pay for changing all those microscopes to be digital, and who’s gonna get a return on that investment? So you can see, everywhere you look, bottlenecks in the ecosystem of applying a technology that’s already here today and continues to improve. It’s also true of drug development.

One of the examples I gave at that National Academy meeting, and I think is true and is a warning sign for me, is the failure to develop antibiotics. So, back in my dad’s era, he benefited from drugs that are today generics, but when he went into the hospital with an infection, doctors had an antibiotic off the shelf that worked against the bugs in the hospital. We don’t have those today. This is a growing problem. The likelihood of dying from significant infections continues to go up, and it’s not a lack of scientific tools to understand bacteria and how to make better drugs.

There’s just a complete lack of commercial incentive to do so. I’ve got colleagues—I’m working on some rare diseases—and I’ve got colleagues who ask me, “What do you think? Pharma companies have pulled back a little bit from rare disease. Is that gonna be a challenge?” I use this example to say that it’s not a given that the economic incentives will always be there, and we need to be super careful and super mindful of the policy implications that we have at the societal level—of what we think is important and how we reimburse these efforts—because it is the return-on-investment engine that ultimately drives the innovation that benefits not just us today, but our kids and our grandkids in the future.

It reminds me to ask another investing question: What do investors of your type most consistently get wrong when underwriting therapeutics investments specifically? What risks do they consistently care too little about, and what risks do they care too much about? As we said before, it’s a hard area to earn a return. There are huge potential returns available because the market is all of humanity. What do you think investors like you get wrong most consistently?

9. Therapeutics Investors Miss Clinical Benefit

As I joined this field, as I became an investor, one of the more interesting books I read is Scott Kupor’s The Secret of Sand Hill Road, where Andreessen Horowitz talks about what it takes to be a VC investor in tech. They have a mindset where their returns can be 1,000-to-1, and so they’re okay to win only 1 in 100. The math still works. In our space, that actually doesn’t work like that.

Our returns are not gigantic. To get a 10x return for us is wonderful, not something you see commonly, and to get a larger than 10x return is rare. So we shoot for 3x–4x returns. A 5x return is a great outcome for us, which means that we can’t afford to get as much wrong as the tech folks can, because the upside is just not as much. It means that when you look at your investment portfolio, you have to take a very careful view of the balance of risk.

Now, each individual company, you gotta, as Carl Gordon, my boss, says, “Somebody’s gotta come and pound the table, believing that this thing’s gonna work. If there’s nobody around to pound the table, we’re not making the investment.” So there’s gotta be that sense of belief in the concept and the people and the potential. We try to take a very hard-nosed look at what the probability of success is across the myriad of dimensions, and there are, as I said, many.

Now, what do we get wrong? We probably get each and every one of the dimensions wrong. Sometimes we get the team wrong, sometimes we get the science wrong, sometimes we get the clinical application wrong, and sometimes we get the commercial opportunity wrong. I don’t know that there’s any one that stands out, because we try to look at the portfolio, but I think the one that we get probably the most wrong is the one that is still the hardest to predict: What is the magnitude of clinical benefit that this will bring?

Clinical benefit is not black-and-white. We typically invest in things where we understand the biology and think we can de-risk it along the way. Okay, that’s gonna work. Then we have a belief in the drug because we understand the preclinical pharmacology, and that’s all good. So it should work to some extent in the clinic, but how good is it gonna be? In a world which is so competitive, it can’t just be something that somebody’s already done before or even slightly worse. That’s gonna be dead. So it’s gotta somehow differentiate. It’s gotta somehow be better. It’s gotta offer something. That’s probably the part that’s the most challenging to predict.

There’s a wonderful obituary that Malcolm Gladwell wrote years ago of Albert Hirschman, who was an economist in the prior century. His whole economic thesis was that what drives economic progress is the naiveté of thinking how easy it’s gonna be. You get excited about an idea, and the example he gives is back when they had to dig a tunnel through the Hoosac Mountain to connect the Boston area with the Hudson Valley. They said, “This is critical for economic development, so what’s it gonna take?” So they brought in a bunch of geologists, and they started picking at the stone, and they came to the conclusion that, “Look, it’s a big mountain. We gotta tunnel through it, but here’s the thing: The shell of the mountain is really hard, but the core is soft, and so we’ll get through the hard part, and then it’ll be soft. It’ll be easy digging. Then we get to the other side, a little bit of hard, and we’re done.”

“So give us a couple million dollars in those terms and 2 years, and we’ll be through it.” And 5 years in, it’s still hard. There was never any soft part in the middle of that mountain. But what are you gonna do? You’re halfway in. You’re gonna stop now? So they mustered up, and they dug their way through it. The moral of the story is it’s always gonna be harder than what we anticipate when we make the investment. That’s okay as long as we’re clear-eyed about what it is, you’ve got the right people around the table, and you’re in a position where you can take the portfolio approach to those investments.

Now, it is interesting to me, coming over to this side, the difference between being an investor and being an executive in one of these companies. I can take a very hard-nosed look at the risk, and I know I owe my LPs a certain return over time, and it’s gonna be met by this portfolio. An executive in the company, oh no, they’re gonna live and die by that being successful.

I remember when I joined Moderna, my wife, who’s the smart one in the family and has a PhD in biochemistry, looked at me and said, “Tal, come on. This thing… I mean, mRNA’s never gonna work. Really?” And I looked at her and I said, “Honey, not only is it unlikely to work, we’re also never gonna make any money,” because the book value was yay high when I joined, and who thought? But I said, “I’d rather fail on something big than succeed in something small. I can afford to take the risk where we are in life, and if this were to work, this would be big.” So she said, “Okay, fine. Go have fun. Do whatever.”

We agreed that I’d do this for 3 to 4 years, and then it would flame out. I’d have an interesting experience, and I would still be employable, probably, and it’d be okay. I remember when we went public, and investors started asking me, “Well, Tal, what do you think about the stock price?” And the only answer I could come up with was, “I’m investing in this company with something far more valuable to me than capital. It is my time on Earth.”

So you do with your money what you see fit, but I’m telling you, I’m putting my time on Earth into this. That’s how I believe in this company. And so it taught me the primacy of sweat equity, if you will. I wanna make sure that my capital equity follows the sweat equity of people I believe in because, at the end of the day, that is the driver. Yes, the science and the technology and the medicine, they all have to work, but it’s the people who make it work, and it’s the people who believe, who are putting their time on Earth and their talents to drive it, that are gonna make the difference and are gonna make the returns.

Do you think it’s possible for an investor who’s incredibly talented in lots of traditional ways but not formally trained like you are as a physician, as a scientist, as a medical professional, to earn great returns investing in this space? Or should it just be the domain of people like you who have deep domain expertise?

No, I think it is possible. I think it’s been proven possible. I think those people who are successful in it are successful because they’re smart in recognizing the multidisciplinary nature and knowing how to ask the questions. I’ve seen this firsthand. In fact, some of my colleagues even at OrbiMed are not people with operational experience at all, and one of my realizations early on was, “Huh, I thought I needed all this experience to be good at it.” Well, in fact, no. It turns out that people can be much smarter than me and can get there without the experience just by virtue of their wisdom.

That being said, I do believe that you can replace experience with wisdom up to a point. Beyond that, you better go and ask somebody. And so I think those investors who can and have earned great returns without the deep domain expertise have done so because they know how to find the right expertise and how to understand the question.

And again, given that it's such a multidisciplinary challenge, even I don't have the expertise. I've got a narrow band of it. All I've learned, if anything, is to ask questions about the areas that I don't understand.

If that's true, then of course somebody doesn't even need my expertise to ask those questions. They can come from wherever, and you've seen people do that. It's that ability to ask the right questions and find the people whose answers you trust, and the understanding of why you trust their answers, that I think makes a great investor.

It's also the traits that make a great general manager. I was fortunate to work with Stéphane Bancel and his executive teams. His brilliance is being able to go function by function and just ask the five whys: Why is this? Why is this? Why is this?

The second thing that made him so effective is that when you gave him the answer, it had to be in plain English that he understood. If he didn't understand, he said, "I'm sorry. I don't understand what you're telling me. Can you please—"

Say it again.

Yeah, dumb it down for me. Why is it so? My PhD mentor taught me early on that if you can't explain what you're doing to a kindergarten student, then you don't understand it.

My penultimate question for you, given that it's the one on everyone's mind—I'll try to think of a unique way of structuring it—is the impact that AI is going to have, or might have, on this field, just writ large. Maybe the way to think about it would be your bear case, your base case, and your bull case for the ways in which AI models may affect the world of medicine from this point forward.

10. AI Reshapes Medical Care

The bear case is that they make slow inroads, one thin vertical domain at a time. That's already starting to happen, but it's slow. It gets encumbered by finding a return on investment for each one of those thin verticals.

What does that mean? What are inroads in a thin vertical? Just bring that to life.

It means that I've got a company that's figured out how to do better transcription for nurses, but not physicians, so they're going around and making a new AI tool just for nurses and nursing homes. Then I get somebody else who says, "Hey, I've got a great system that can read the charts and figure out who's a good patient for a clinical trial. I'm going to go and deploy that in more institutions."

Those are all such thin, piecemeal applications—

Got it.

—that it's going to be super challenging to get them integrated. The bullish case is that there's somehow an integration of these systems and a realignment of incentives that allows people to leverage productivity.

I think it was one of your prior guests who made the point, when they were looking at the early days of EMRs, electronic medical records—now we call them EHRs, electronic health records—and realized that the implementation of those systems into hospital systems actually hurt productivity, which is an abomination. The whole point of technology is to improve productivity.

I remember this as an intern. People started putting PCs on nurses' desks and having the nurses start to spend time typing things in. That just took time away from what they were doing. It didn't help anybody.

If we figure out how to turn this on its head and actually find ways to put in systems that will improve the productivity of the healthcare system, that's where I think the bull case is. Once we figure out how to do that and align it with an economic return on that investment, then I think you'll see an acceleration of change, and the pace of change can actually be very rapid.

Do you think there's a world in the bull-case category where biology becomes engineering, where we're able to run simulations in digital twins of humans, not in real-life clinical trials, and therefore we're able to iterate at a speed that's impossible in the real world and solve all the problems? This is the utopian take on how AI will affect drug discovery, therapeutics, medicine, and health. Is that possible? It feels like if it's possible, we're going to get there on some time horizon because of how fast technology is evolving, but I'm curious if you even think it's possible.

That, for me, is still a little bit in the realm of science fiction. You can get digital twins up to a point. They're going to be valuable, again, up to a point.

The place where you need to be careful is that we have to be able to balance this with the ethical obligations to patient autonomy and the other things that we hold dear and near to our hearts. It can't be utilitarian, because the utilitarian approach will be dictatorial, and that's not going to work. I think we're already seeing a public backlash against science trying to assert itself in the name of good.

Science is only in the name of good in how we use it and deploy it. You could probably get your best healthcare system up and running in a place where patients had no autonomy if you measured outcomes in a certain way, but of course, that's not a way that's acceptable to us to measure outcomes.

I think the world's going to stay messy, purposefully so. I hope so. I think it's what makes us human beings human. In that regard, I think I'm a little bit more circumspect about how rapidly some of these technologies will actually be able to be deployed. We're going to have to be careful to balance, and that's true of AI in general in every application sphere, no less healthcare.

Tal, this has been such a total blast to learn from you today. I ask the same traditional closing question of everyone: What's the kindest thing that anyone's ever done for you?

I'll tell you, the kindest thing anybody's ever done for me was probably Steve Rosenberg when he took me into his lab at the National Cancer Institute back in the late '90s. The kindest thing he did was that, as I probably wasn't one of his best postdocs—I kept proving why things didn't work as opposed to showing things that did work—as I was leaving, he said, "Look, Tal, if there's anything I can do to support you in your future, don't hesitate."

I didn't feel like I was that good. I wasn't sure: Did he really mean it? But it took me almost a quarter of a century to realize how deeply he meant it, because not only did he come and support me over time, but I had tried in his lab to make an mRNA vaccine for cancer. We did it in mice, and we published it back in 1998 or so.

When I was at Moderna—and now this is 2016 or 2017, 20 years later, and we've come up with this personalized cancer vaccine—I called him up and I said, "Hey, Steve." He's the best example of a public-servant scientist that one can ever hope to meet, and he's one of these people who've not just moved the whole field. The whole field of immuno-oncology owes a great debt of gratitude, because not only has he uncovered some of the key themes himself, but he's actually trained an entire generation of us, and I'm proud to be one of his fellows.

I called him up and I said, "Hey, we've got this personalized cancer vaccine. I think it's got a shot at working. Would you be interested in maybe collaborating from the National Cancer Institute?" Of course, I'm at this mRNA company back in 2016. Nobody thought this would work, my wife included.

Yet he invited me to come down to Bethesda and give a seminar and talk to his team. They partnered with us, ran a small clinical trial, and he actually listened and continued to support us. He bought into it. I have to say that that belief in me—that true mentorship that has really lasted, I feel, a lifetime—is certainly the kindest thing that anybody has ever done for me.

In a small way, when people call me up—whenever somebody's ever worked with me—I always have that in the back of my mind. I feel a debt of paying it forward, if you will.

If we could all do that for one other person, it feels like the world would be a lot better place. Tal, great story. Thank you so much for your time.

Patrick, it's been a real pleasure. Thank you.

If you enjoyed this episode, check out joincolossus.com. There you'll find every episode of this podcast, complete with transcripts, show notes, and resources to keep learning. You can also sign up for our newsletter, Colossus Weekly, where we condense episodes to the big ideas, quotations, and more, as well as share the best content we find on the internet every week.