Panna Sharma: CEO of Lantern Therapeutics

Panna Sharma, CEO of Lantern Pharma, headquartered in Dallas, Texas, shares his thoughts with BioBoss host John Simboli about leadership and how Lantern is working at the intersection of artificial intelligence, genomics and machine learning to advance precision oncology therapies.

Transcript

Copy Transcript Copied

John Simboli  

Today I'm speaking with Panna Sharma, CEO of Lantern Pharma, headquartered in Dallas, Texas. Welcome to BioBoss, Panna.

Panna Sharma  

John, thank you very much. I'm really looking forward to talking with you today and I love your show and listening to it. So thank you for having me.

John Simboli  

Panna, what led you to your role as CEO at Lantern Pharma?

Panna Sharma  

I was the CEO prior to Lantern Pharma at a company where we provided a lot of genomic and biomarker services for patients, but also for big biotech and big pharma--I think 10 of the top 15 and 18 out of the top 20. And were doing millions of dollars of genomic work trying to understand and assess how drugs should proceed or not, how tumors were evolving or not, how they were responding. And I really loved it, I enjoyed every aspect of doing that work and learning about it and learning, what could eventually impact patients. But we didn't own the therapy. At the end of the day, we were yet another service provider that was a checkmark somewhere. And I thought a lot about the evolution of the company. And in my heart, I really wanted to take a shot at the bigger prize of really creating therapies that were life-altering. And I saw so many wonderful therapies. And I also saw a lot of wonderful insights about how to diagnose patients and how to understand the tumor. But so rarely were they brought together. And there was always some kind of disconnect. Not that I could do it better or differently, but I definitely felt the next wave of cancer drug development was to integrate that type of effort, to integrate the data that was coming real world from patients, and integrate the biological findings that were coming from large scale discovery and biomarker efforts. And so we tried to transform my last biotech, publicly traded biotech, I left at the end of January of 2018, after spending probably a good 6,7,8 months developing that plan of how we transform ourselves. I always feel like the ultimate question is, what do we do best? What is it that we're really doing? We're growing top-line revenue by 20x percent a year, we're sucking up this kind of capital, we're launching these essays. All that is stuff, but what is it that we really are doing? And so I went to the board, I said, Look, this is a tough question. But we're doing a lot of stuff but we're not generating enough value. In our space, around genomic services, is totally changed. This is a commodity item and so we're really not being valued for the kind of insights and efforts that we have. But what we're really doing is we're de-risking and accelerating development, in many cases. And also, in many cases, what we're really doing is helping match the right therapy options for clinicians on behalf of their patients. I said these are hallmarks of really precision oncology development. So like any good board, they were split and some thought it was a great idea. And some thought it was absolutely horrible and that it was a totally different business. 

Panna Sharma  

And I recognize that and I said, there are a lot of companies in the history of biotech, and life sciences that have made that leap. Genzyme is a great example. Genzyme did not start as an enzyme therapy company, they really started developing as an assay and reagent provider for ultra-rare enzyme therapy, where they were providing the diagnostic kit, and then the assays. And there's examples like that throughout the history. Evotec is a wonderful example. They were a client of mine when I was a consultant and banker, and I saw firsthand the painful but very, very disciplined execution of going from being a large-scale services and equipment company, which was the old Evotec, to a biotech. It took two CEOs and it took difficult decisions about what to shed off. And it took difficult decisions about what is that core competency of ours that no one else can do? In terms of finding those hits in neurological indications? Let's get someone else to pay us for that. Can we take a piece of that hit? Can we take a piece of the upside? And now look at them--multibillion-dollar, fantastic biotech, really. And there are many examples like this. Albany Molecular is one. And so I was really inspired by that. And I said, we're living now in an era where the kind of information that we're sitting on has the ability to change drug development, and it has the ability to potentially get therapies out there faster and with less risk. I'd like to have a shot at doing that. 

Panna Sharma  

I did not get that shot. And, you know, things happen for a reason. I left at the end of January. I think maybe the announcement was in early February, it was the first or second day of February. And then I really started talking with companies. Public biotechs can be in high demand, public CEO, biotech. So you know, people are saying, well, we have this company and we got this one and this one we want to do IPO. But I said, well, what do I really know now and love it. And I really knew that I wanted to continually focus in on oncology, and I had a long path to get there. And that I really understood it. And I was able to take a seat at the table with lots of really intelligent, wonderful people, globally, in cancer. And I wanted to leverage that. I sat on the joint venture board with Mayo Clinic. I did a collaboration with places like Sloan Kettering and Moffitt Cancer Center and Johns Hopkins. And so I thought, I really want to continue in this community of people who are focused on cancer. As I spoke to a couple of the startups, and part of my analysis is who is using these kinds of data and machine learning type approaches to change drug development or try to change drug development? And one of the companies is Lantern. And so, the CEO, I knew pretty well. But more importantly, I knew the board members, from being out there, and being in the investor community, and so the VCs, and I actually had had discussions about that space and about genomic data. And as you're often do over a cup of coffee, or, or a cocktail before the end of conferences with folks, and one of the key guys at Bios Partners and I got along very well. And he was the primary reason I'm at Lantern. And so he trafficked me in with the rest of his partners and colleagues and the other co-investors. And a key decision came up at Lantern about where was the existing founding team's natural place. They had some very strong opinions on that. And so I got involved with the company in August. And I think it took over a CEO a few months later, in Q4, and was really thrilled to be given the ability to kind of take it to the next phase. And so kind of as a culmination, as I explained to some of the colleagues who knew me, but also my family, it seemed like a wonderful culmination of getting to take a company public, again, which I'd done twice in the past, but leveraging the cancer knowledge that I had put together over the last seven or eight years, and also my passion for AI and data, which is actually what I studied in college. And so, I thought wow, I'm able to harness all these different pieces. And so it just seemed like a natural, it seemed like, I would have to say, from the day the process started, it just seemed like a perfect fit.

John Simboli  

When you were going through that process of thinking about integrating these different pieces that weren't quite falling together, the way you wanted at your previous company when you're looking for that opportunity is one of the things you considered, I'll find a big pharma company where the infrastructure is there, and I'll convince them of the efficacy of this idea, they'll bring me in?

Panna Sharma  

It takes an hour, sometimes, to start a meeting at a big pharma. What we do at Lantern is we have stand-up Scrum meetings; they last 15 minutes. In a big pharma, I used to go to some of these big ones that you probably have heard of, outside Boston or in New Jersey, people are 15 minutes late. We're done. We are done with our stand-up meeting about our AI platform in 15 minutes. It starts at 1045; it's done by 11. Everyone else has other meetings by 11. And everyone has simple notes. I mean, the pace at which we're doing, I mean, meetings are done differently. The culture is different. Just the reports are different. We don't need  12-page, things that you put in binders, and people are stapling together after a meeting. Why sit in those meetings that I'm horrible or ill-equipped for? It's a waste of shareholder resources, and it could be frustrating to the team also. 

Panna Sharma  

Part of that is, "Don't be afraid to say move ahead without me." And that's one of the biggest challenges founders and CEOs have. And so I say that now all the time, all the time. So we're having a meeting, for example about, not that is not important; it doesn't reflect the values of the company or myself, it reflects where can I be of best assistance and help? We're trying to determine for an upcoming trial, some policies and communications plan around compassionate use, and making the drug available, and how do we communicate that, and where do people go, and the forms involved and the language and you know, what do we want to say? And I said, Well, there are so many people that are so good at that. And so many have done it before, why don't you just tell me; I don't need to be involved in these meetings tonight. I'm going to probably frustrate you and make some suggestions that you say, well, we  can't do that, or we have to do this or we have to do a forum I said, you guys already know the answer, so move ahead without me. If this stuff looks good, you guys know our brand guidelines, you know where we're going to go. And I'll save myself three, four or five hours over the course the next . . .  and I don't need to be there, I trust, the team will be really good. And I don't need to sit there and just take notes or think about people. I think it allows leadership to pop up, and allows me to focus on some of the other aspects that I'm really good at, for example, trying to make a decision on the data link architecture, and who's going to do that, and why and grilling the architect. I know our radar team could use my help there. 

Panna Sharma  

And also, the other thing I learned a lot, is that there are gifts that each one of your people that you really like has. And you've got to remind yourself what those gifts are so that you can give them adequate time to shine. And so, you know, I'm really good at XYZ things; if I don't get to spend time on those things, the end result will be different or could be different. It's the same thing in any emerging company, each one of those people is there because they believe in the mission, they're uniquely gifted, they have interdisciplinary skills, they want to feel a sense of accomplishment, which is why they're in a smaller emerging company. And so you have to constantly recognize, hey, you know, Jocelyn is really good at this. So Nicole is fantastic at that, or Peter can be highly frustrating in this 80% of the time, but this 20% there's no one else out there. So you have to really, always make sure that you take note of that and allow those people to run with those efforts. And so the biggest challenge, oftentimes I see with CEOs is they sit in too many meetings. And I think that is something that maybe younger CEOs, I think are . . .   I hate to say it's an age thing, but I do think—I'm not in that group, by age, I am older—but I'm kind of sitting in between, but I definitely think that guys who are 50, or over, maybe 47, there's an age cutoff, they want to be the serial kind of manager in all the meetings. They want to be seen, they want to have people report in and, you know, make some kind of decision. Whereas I feel like the younger CEOs, today, do focus some of their time more narrowly, on the one or two things that they happen to be really great at, like either go to market or supply chain stuff, or communication, or the three or four things they pick, you know, there's ultra-rare guys that are gifted at all those things. I think the younger guys will pick the three or four things and leave everything else up to the team. And I don't necessarily see that with guys at a certain age or above. We're now actually getting into where even the biotech companies are very, very different culturally than the new breed of, I would say, more data and AI-driven companies in biotech. Because, you know, there are age differences. Obviously, the average age in the data science and AI team is probably, you're lucky if you're hitting 35. And for the folks that are drug development people, or CMC people, or medicinal chemists, or molecular biologists, their average age is anywhere from 47 to 70. When you talk about a company like ours, and other great companies, Recursion, Atomwise, Benevolent, you have whole big squads that are probably lucky to be pushing 30. And that changes a lot in the pace and expectation and work style and all the other stuff, as well.

John Simboli  

It's moving further in the direction of being a software company or being an agile company.

Panna Sharma  

Yeah, I think more agile. Drugs still have to be made; you've still got to manufacture them, you still have to test them in animals and humans, and software is not going to replace all that. But if you can make decisions more rapidly as a result of software-driven methodology, instead of doing a screen against 60 cell lines, you can pick a screen against the eight or nine that are most likely and you already have engineered the next eight or nine. And you didn't take two months to do that, then that just changes the pace. So the whole preclinical ramp up, instead of taking one year, if it can take you three months or four months, that's hundreds of 1000s of dollars in savings every month in a biotech. All that stuff adds up and all that is software-driven. And so if you can get the right cancer biology people and the right software people together and get them to start mashing properly. That's like magic. That's how you can create great companies. Because that's tough to do. Even in big pharma today, it's very rare to find that, with the exception of a few, but they typically have their data people in a silo, they have their bioinformatics people as an internal service provider to maybe the biology team. And the biology team doesn't really listen to bioinformatics people because they don't speak the language and the bioinformatics people don't understand really everything about drug development or clinical trial design. Because they feel like well, we told you the markers, why aren't you using them? And so, the sense of silos is much more powerful than the sense of team.

John Simboli  

As I've understood what you said, your gifts, one of your gifts, is to know when to say "move ahead without me." Implicit in that, it sounds like, is the ability to say, "I know enough about this" to put this person and this person in touch and keep an eye on it, but not to feel like I have to be the connector. So how do find that balance? I mean, that must be a hard thing to know. 

Panna Sharma  

Sometimes things don't happen at the pace you want it and you kick yourself for not being more involved. But sometimes things happen that you know would not have happened if you're sitting with your thumb on it, or they went in a totally different direction. You also have to try new things; sometimes you try new things. And you think there's a 10% chance of it working, but you're going to learn from it.

John Simboli  

So can you remember when you were eight or nine or 10? And you're thinking, well, I want to be this kind of person, when I grow up? For most of us, that was undoubtedly, "I wonder what my parents want me to be when I get to be of that age." So can you remember what that was? Does it have anything to do with where you are in life now?

Panna Sharma  

I wanted to be a brain scientist. I really thought we could create artificial brains. And, you know, this was the rise of the computer. So when I was 10, it was 1981. I was born in 71. Those were early PC years if you remember, right? That world of Atari was coming on and Commodores. And so there was this great idea that one day we could one day have these banks of computers that could think like a human and I thought maybe one day we could reduce the neural circuitry to stuff that could be replicated on chips or boards. I remember taking a programming class, I was really fortunate to grow up in California at the time I did. I took a class, I think I was like 10 years old, maybe nine, I don't know, I was pretty young, at the local community college—programming. I remember programming on stacks of cards. So that's how old, I mean, no one even my age actually knows that. I just happen to have that experience. I have no idea how I did in the class. But I remember, I took the class, or maybe I audited, I don't know what I did. But I remember having to carry the stacks of cards, and they would every 10 or 50 cards of color would change. So in case they fell, you could put them back together. Remember doing like simple things like getting into print and doing a raise of, you know adding things and subtracting things and simple math and just thinking that's really cool. Which really helped me later because then the whole era of breaking apart basic programming at home came up, where you can hit control, break, and break into the program. And then find different stuff and tweak and see how it worked. So, at that age, I thought maybe . . . but I really like biology more. I remember I wanted to try to make a brain. And so I really thought I would be a professor and one day make brain stuff. And the other part of me I really wanted to be in planning cities the future. I don't know why. Like every young kid probably, are we going to have flying cars in the future? Are we going to have homes that live up in the sky, because we're going to pollute ourselves so much. Because remember, at the time LA had smog, bad, bad smog. And I think my dad had some allergies to smog and my younger brother definitely did. So, I was always aware of it. And so I remember thinking, well, what if we lived in the mountains or above, and there was a lot of like, when you go to Disney, there are these vertical cities, right? And so I just thought, well, we could live above the smog layer;, it's not that high. And so, I envisioned these cities of the future that would have those monorails at 1000 feet and like a whole different plateau. There'll be an artificial plateau above the smog layer because we polluted everything. So, I was fascinated by that concept, and I thought we'd have flying cars like the Jetsons. So I remember I would sit there and draw what these artificial cities would look like. And I definitely thought in the late 70s 80s, how foolish I was, I thought definitely, when I grow up, we're gonna have flying cars.

John Simboli  

Well, I think it sounds to me like you are "making brain;" you're just making brain that . . . 

Panna Sharma  

Does something very, very narrow, very, very narrow, which are called narrow AI. And I think that's one of the things that I, when I talk to other people doing AI, is really like to understand kind of what is the problem area that you're trying to initially solve for, or that you feel like you've already solved because you can't go tell a set of code or build a set of code that says, you know, drive a car; you start with something specific, follow a yellow line, make the camera understand that the wheels need to go left or right of this yellow line. If you can do that really well, now you can move on to the next task. Now if you can teach the computer how to do that task via line now tell it to teach itself regardless of color. Now tell it to teach itself, regardless of the type of line. But what is the one thing that you really solve for? When I came to Lantern, we said, what is the one thing that we feel we really have solved for. We made a set of algorithms, three algorithms at the time; now we have three dozen-plus. And one of them was specifically to reduce the transcriptome complexity to something that was much more manageable in some kind of companion diagnostic signature. And those signatures were somewhat replicable. They somewhat give us some insight into what the biology was of the disease. They had genes that were significant, not only statistically but genes that were significant in terms of biological relevance. And so I said, OK, so that's kind of what we do. Now, it's how do we do that on a scale that no one else is doing it? And so we did it well, at the time and in prostate cancer, and certain solid tumors, mostly GU cancers. And I said, well, we've got to do it an all cancers. So the company had 20 million data points at the time, mostly from cell lines and some PDFs models. And I said, well, this is not going to cut it. We need real-world data, going back to my own background in that. I said we need to go from 20 million to a billion. How do we get to a billion?

Panna Sharma  

And so when we went public, we were at 275 million, we're today, probably close to 6 billion. So even in one year, we went from 275 million to now over 6 billion cancer data points, over 85% of which are real-world tumor, drug interaction data, or tumor drug sensitivity data, or from tumors pre and post exposed to various drugs or drug classes, or from diagnostics. And so now, as the number of data points grows, so does your ability to go across more tumors. So that'll make our algorithms more powerful; it allows us to ask a wider scope of questions, it allows us to potentially come up and predict combinations of our drug and other drugs both approved and in late stage. When we first started, we didn't have late-stage drugs, we only had approved drugs. So now we have late-stage drugs. Before it was manually entered. Now robots can read that data. Before it was human tagging, now we can do most of it, again, a robot AI machine does the tagging cross 28 measures. Again, going back to what does the AI solve for? You have to start with something because unless you do that one or two things really, really well, you can't go on to the next. Any problem is really, really complex. It could be like, what are we going to have for dinner? Someone could probably write an algorithm for that and be pretty close. What does the dad feel like? What does the mom feel like? What time is it? What day is it? What's their social class? What restaurants are near them? How many daughters do they have? How many sons? Someone can probably watch a household and say, okay, I've watched all households with three children, here's what they have for dinner; I've watched all households with one child, here's what they have for dinner; I've watched households that make between $50,000 and $70,000, between $70,000 and $140,000. They can put all the variables to figure it out and say, with about 70% accuracy, this is what they're going to have. And if you can't solve that, then you can't go solve other problems. So that's the same thing. Let's solve one thing really, really well. 

Panna Sharma  

And the thing that we're solving has a very important impact because we can save 1000s of lives faster and cheaper. And that's one of the things they care about most. You can hire all these data scientists on these data wonks, but you have to make sure they understand the mission. Our mission isn't to create great data. Our great data supports one fundamental thing, which is cancer drugs cost too much. Cancer drugs take too long. It should not take 10 to 12 years and close to $2 billion. There's just no way. We have 89 approved therapeutics that are targeted. Forget the bendamustine and the cytotoxic agents—we have 89 approved drugs for cancer. We have at least 100 different types of leukemias and lymphomas, probably 30 or 40, multiple myelomas, probably 600-700 different types of solid tumors that we can classify genomically. So how long will it take us, if we really want to have precision medicine, we should have drugs for all those. Let's say we have drugs for half of those cancers, we need to go from 89, to at least 400-500, approved therapies. It should not take us another 40-50 years. And it shouldn't take us a billion dollars per shot; no one's got that kind of money or time. 

Panna Sharma  

So at the core of a lot of the essential problems in cancer, is data, is capturing that data, and then being able to generate large scale algorithms that can learn and teach themselves and learn because you don't have enough smart people in the world, you just don't. And so one of the things that we do, and it was one of the big lessons, is that if you can't do something by the third or fourth or fifth time, kind of automated, then you're not filling our mission. If everything is going to be manual, in the data world, that's not going to work. Your ability to replicate it is very low, your ability to automate it is obviously low, and your ability to then scale it is going to be very low. And so that's not machine learning, that's just you being a better data scientist. So, we're solving a narrow enough problem, you know, we're not trying to solve Google Earth and we're not trying to solve all kinds of really much more, potentially, complicated problems; we're solving a very narrow set of problems of where can we point this drug? And if we don't have a drug for it, what does that drug look like? And then what drugs can go together? And every time we do these questions, we've got to get better and better at it. Well, we know what drugs go together, but we can't do it for these 12 classes of drugs. OK, let's put those aside. Let's do it for the ones that we can do it for, and be really good at it. Then let's add more classes of drugs, then let's add more classes. Why can't we do it really well, for these? We don't have enough data, let's go get that data. So everything that you do really well allows you to then think about where you expand it to. And that, I would say, is one of the important things; you have so many smart people, and they just want to take this huge bite at everything. Right? They just want to like, oh, this. And you 've got to say, OK, well, what is it that we're actually going to do? And it's  OK to do something very small and specific because you'd be surprised how quickly you can scale that in today's world. Because the data is all there. And it's available in oncology. And the computing power is there. It wasn't there three, four, five years ago. And so that's why probably the other guests you've had, you know, that's why we're able to have these aha moments and be like, wow, we can go back and do it. I mean, you had these brilliant scientists, totally brilliant clinicians who said, Oh, I'm going to use this leprosy drug in multiple myeloma. What AI is going to tell you is that? There's no way, right? So it's just amazing. But it shouldn't be accidental, it should be that we find 25 of those insights, and maybe three or four of them work. And we can't rely on the brilliance of you know, one guy in one generation to do that.

John Simboli  

What's new at Lantern pharma?

Panna Sharma  

So our platform is, we think it'll approach close to 10 billion data points, by the end of the year. A lot of that growth is going to be in hematologic cancers. So you can expect a lot in hematologic. To date, all of our programs have been in solid tumors. And they've mostly been in monotherapy. So you can expect some things in combination therapies that we've identified to start emerging now, as well. We have developed a new molecule called LP 284, which is a really interesting molecule that has been pretty much shepherded from day one with AI. So there's a lot of programs like that. So on the data side, I think there's a lot of milestones that we'll be talking about. And you'll see the data manifest itself in terms of new drug programs that are going into humans next year. We have two phase 2 programs now that are active, again, in very targeted indications, one in metastatic castration resistant prostate cancer, and one in never-smokers that come down with non-small cell lung cancer, which is a very different disease, genomically very different, drugs work very differently. And again, you know, we use data to help solve the problem of why and how does this particular drug work or not work? And can we really rescue it? Which, at the core of it, really was a data problem.

Panna Sharma  

A lot of new stuff that you'll see from us is really focused on getting our insights translated into dosing of humans that improve outcome, that, hopefully, we can get these drugs to market or sold off to big biotech and big pharma partners, and repeat. We're a fairly small team, you know, we're only 17 people. We went public, we were six people. And so we've been able to raise a big chunk of money, over $96 million in public capital, and fairly disciplined about our burn rate. And so I think, as we continue that, we're going to look more and more to our external partners. So we'll announce a lot more partnerships. So we've got partnerships with Johns Hopkins, Fox Chase, Sloan Kettering, C-TRIC in the UK, partners in Denmark. So we've got a great who's who list in cancer research. And so that'll continue. So I think partnerships will continue growing our march towards getting these drugs into later stage human trials, and the growth of the platform to do more indications, more combinations, and also opening up the platform to other companies.

John Simboli  

In my experience, when a CEO or founder goes to, say, an investor conference and makes that presentation, there's a certain number of people come afterward and say, yes, I understand; I want to talk further. A certain number just walk away because they've understood it, and they don't have a match. And then there's a sizable percentage who say, oh, I've heard it, and then you're thinking yourself or I, as the branding person, I'm thinking myself, they didn't get it, they heard something different than what I intended. I'm sure that has happened once or twice in your life. So when people misunderstand who Lantern Pharma is, what do they say? Is there a pattern? And then how do you help to get them back on track?

Panna Sharma  

Yeah, it's a wonderful question. Because it happens, you know, especially with small companies, people hear the pieces they want to hear, or that they're predisposed to. So for Lantern, I feel like a lot of people walk away, or sometimes think that we're a service provider, that we're using our AI to help other companies or just to find new uses for a drug. And that is what we do. But we actually develop the drug; we're a drug developer, we're just using AI to develop our portfolio. So instead of using traditional high throughput screening, and combinatorial chemistry, and high-density biology, or whatever you want to call it today, and going into animals, we're doing that using computers. Instead of teams of 20 biostats people, we have AI and deep science people. So I think most people walk away thinking that we're a service company or service provider that tells a company what to do with our drug, as opposed to we own the drug. Which is a, you know, a totally different outcome for investors.

John Simboli  

Do you still get the time to, maybe on a weekend or something, when you're with your kids to say, you know, if this crazy thing works out, I'm actually going to help this person, this human being that I have seen in a hospital is going to maybe have a better life someday? Do you get that chance or does that come later, to think about stuff like that,

Panna Sharma  

I was able to spend a little bit of time and get some unique insights, was actually spending time in hospitals when I was at CGI with clinicians and with patients, and actually with members of my own family that got cancer and members of close colleagues that got cancer. And I think that is something I try to instill for everyone, to try to spend time in the real patient world, or really, with the tough decisions that clinicians have to make. The number one thing that I hear over and over from adult cancer patients in the US that they're worried about is cost. That's horrible. They don't even have the brain space to think about Should I go on this protocol, or this regimen or not? Because their brain is so consumed by the cost and the time away from family question. And so, you know, a lot of times a clinician has to make tough decisions, but they're oftentimes not guided because the patient themselves isn't spending adequate time thinking about, am I a better candidate for this regimen because they don't have the information or brain space, and so I feel passionate about you know, we have a unique opportunity in our space right now over the next 10-20 years to totally smash the product development cycle. And every time I see silos, you know, it's every other industry has done it. You know, shirts today cost the same amount of money as they did 30 years ago and it's probably gotten cheaper shoes. The only things that have increased in the United States is higher education and pharmaceuticals. Those are areas that technology and AI have the potential to change, not because they just should be cheaper, but because they can. It makes for a sustainable, more innovative future. If we have 500 cancer drugs, just think how amazing things would be. Right? It's not just because, hey, you know, we're maybe giving too much money to pharma, just, you know, is our output strong enough? Do we have enough output? And, I think those people, who ask those tough questions, can change the future. I mean, look at companies like Uber, they took a ride that costs 100 bucks from JFK—I mean, now that brought it back up— but, you know, they changed it. They changed people's thinking. They changed the experience, they changed the amount of data, and we have a similar opportunity. And so I think, the COVID-19 lesson of the vaccine development, we can learn a lot from that.

John Simboli  

Thanks for speaking with me today, Panna,

Panna Sharma  

John, thank you for having me. I really enjoyed talking with you.