Alex Telford - Unlocking Innovation in Pharma - [Invest Like the Best, EP.360]

25:38And then the management thinks they'll say that we knew it all along or something in retrospect. But you can't prospectively figure out what drugs are going to be really impactful a lot of the time. So I think it's quite dangerous in an industry like pharma, which is so innovation driven and science driven to impose these top-down pre-described notions of what people should invest in. You're very likely to be wrong. Until a few years ago, I think Novo Nordisk was regarded as a pretty boring European pharma company, not very innovative, just doing insulin. But now they're, it's like the 15th biggest company in the world or something.

26:16And everyone's excited about the GLPs. You would have necessarily predicted that would come out of Novo Nordisk. But Novo Nordisk was able to produce GLPs because they had this really long standing interest in a specific area, metabolic diseases. They support the scientists there. And they've just been tinkering in this area for a long time. So to come back to the Super Mario idea, that post is about this idea that one of the ways you might productively use technologies like artificial intelligence and simulation tooling is to have an intuition about what areas of science might be promising and then use automation to do scalable tinkering around that nucleus of an idea.

26:57so the example in super mario was that there is this idea that you could do a specific type of jump to get through a level in super mario much faster than people have been able to do so far but no one knew it was possible so until a software developer developed this tool called scattershot which simulates millions and millions of marios jumping up this level in every possible configuration of jumps and starting positions he managed to figure out there actually was a way to get up to the top of the castle they were trying to jump up and finish a level much faster than previously possible. It's changed this whole way that speedrunning is approached for that game.

27:29Speedrunning itself is probably out of interest to that many people. But the general principle, I think, is interesting that you have this intuition that something is possible. But Janssen, when he was developing his drugs, he had an intuition that you could improve this compound, pethidine, to make a better drug. But he didn't exactly know how to modulate it to make it the best drug it could be. He was just tinkering around. So if you could use technologies to automate a scientist to feed the process of tinkering trying out solutions testing them then you can tighten the feedback loop around the industry and maybe develop drugs a lot faster than you might otherwise be so one of the big problems with pharmaceuticals and industry is that the iteration speed is super low compared to something like tech like in tech if you have an idea for a product, you can spin up a demo in a few days and you can show it to some customers.

28:24You can get some feedback on that. You can then iterate it, layer instantly in code effectively, and then show the next demo to customers, get more feedback. You can do these A, B tests, massive scale. You can move super fast and learn very quickly. But pharma is a very hard domain to learn quickly because everything takes so long. Clinical trials will take around 10 years to get everything through clinical trials. Even to do animal experiments, it takes weeks or months to do animal experiments in certain cases. Making all these chemicals takes a long time. You're testing even in the very earliest stages.

28:56So you learn at a very slow rate about what works. And if you could just find ways to speed that up across the whole process, then you'll do a lot to, I think, address some of these problems with the pharma industry, like drug prices, is just a huge healthcare spend, things like that. How does AI figure into all this, obviously alpha fold within the mainstream consciousness? I think Jensen Huang, the CEO of NVIDIA, was recently talking about how much explosion of activity there is using AI in the world of medicine and biology. So talk about the role that this new set of tools might play in both discovery and speed, in tinkering and all these different ideas.

29:37I think I'm short-term pessimistic, long-term optimistic. when it comes to AI for drug discovery. And I'm much more optimistic in the short term about using AI to automate some of the processes along with drug development. Not necessarily the designing of drugs itself, but all the steps to bring a drug to market. So if we start with drug discovery, like finding a drug, I think it's just very difficult to see how AI is anything more than just another tool in the expanding toolbox that drug developers use to produce and test molecules. Alpha Fold's an example where now you can predict the protein structure from a protein sequence, which is a problem that we didn't know how to do to a high degree of accuracy until before Alpha Fold.

30:20And a lot of people are saying, oh, Alpha Fold is going to revolutionize drug discovery. But in reality, it's just one little piece of the hugely complex puzzle where sometimes you'll have situations where you don't know the structure of a protein you're trying to make a drug against. and AlphaFold can be a helpful starting point for that specific circumstance. But actually, AlphaFold has problems with the very granular predictions of the configuration of sites where drugs might bind. So it's not completely accurate. So we need more accurate models for it to be really useful for drug discovery.

30:53And to improve AlphaFold, we need more data. It's just takes a long time to collect all this data. So AlphaFold is trained on decades of protein structures that have been collected painstakingly by researchers growing crystals in dark rooms. It's definitely speeded up parts of that process, but a very small part. And I think the way AI is going to impact the pharma industry looks similar to that. A lot of little tools that are going to be very useful and accurate, but they're going to take a long time to deploy and figure out how best to use them and work them into the process. One thing I should also mention is that there's this idea that drug discovery is rate limited by our ability to design molecules and design new drugs.

31:33And that's not really true. Our ability is much more rate limited by the downstream clinical development and testing these drugs and gathering information. We need to actually justify approving these drugs for sale and marketing them. While I'm optimistic about over the long term using things like AI to predict what drugs might be effective to do drug discovery and drug candidate selection. I'm much more optimistic about using AI in the short term to help drugs get to market faster than they otherwise might have been. So things like, can we use AI to help prepare regulatory documents quicker? Can we use AI to do things like help companies figure out what the markets for their drugs are and prioritize what opportunities they should invest in?

32:21How can we use AI to help identify patients who might be a good fit to enroll in a certain trial? So things like trial enrollment just takes way too much time. So just better tooling to identify patients who might be a good fit for a trial and enrolling them seems pretty impactful. If you were just the czar of this entire universe and godlike powers, what two or three changes would you make regulatory, technology-wise, industry structure? Anything that you could change, what would you change that you think would most benefit patient outcomes or the overall system? I would make friculatory changes.

32:57I think I would want to formalize the idea that different diseases require different standards of evidence and then pre-publishing the degree of evidence that we required for each different disease. So something like diabetes should have quite a high standard of evidence. So you should have to enroll very large numbers of patients and test mortality because we have extremely good drugs, different diseases where our standard of care is not as effective. We should formalize this idea of smaller trials, maybe approved drugs with a weaker standard of evidence. And we are seeing some moves towards that with rare diseases.

33:32So you have a lot of these conditions that are too small to really be able to tell. So they maybe have too few patients, or they may have too few patients plus the disease evolves over too long of a time for you to reasonably able to figure out in a standard clinical trial whether or not these drugs are really working to any high level of confidence. So I think in those conditions, you'll have to be able to say, look, we're just not going to be able to determine this in a trial. We have to, based on some principled, reasoned evidence we've collected about how the drug works in animals and maybe some initial human testing and some preliminary biomarker evidence, we can then approve this drug and roll it out and then collect information more rigorously once it's on the market because you're just never going to get drugs for certain diseases.

34:16It's just uneconomical to achieve this vision of hyper-personalized medicine where you're developing a drug for everyone's individual condition and biology without some ability to make inferences about what is likely to work based on early evidence and genetics and reasoning about the mechanism of action. So I think the second thing I would do is just be much more rigorous about data collection post-marketing. So you have a lot of drugs that go onto the market and then they're approved and then we're not really following up sufficiently to be able to tell if they really work. So there's been some instances of drugs that have got this accelerator's approval on some fairly flimsy evidence, but they seem promising.

34:54But then the company's meant to do a confirmatory trial to determine, okay, actually, we have a good reason to believe this drug is going to work, but we don't actually know that until we've collected this evidence over five, 10 years or whatever. The company's dragged their heels on doing the confirmatory study. I think the regulators should be much more stringent in forcing companies to collect that data and report it and takes the market. So this combination of being more open to approving drugs based on a lower standard evidence when there's no good treatments and when it's just unfeasible to do a standard clinical trial, coupled with more rigorous data collection and then taking things off the market when they're not working.

35:30I think that'll help increase the sort of rate of learning and the regulatory process. Surrogates as well, I think are really important. So surrogates are measures like biomarkers that you can use as a proxy to tell whether a disease is being treated effectively or not. So for cancer, what you really care about when someone has cancer is stopping them from dying. But you can get a proxy for that by looking at the size of the tumor. And it's not always a perfect correlation. You shrink the tumor. Most of the time, that's a good thing. Sometimes you shrink the tumor, but it doesn't actually help the patient survive longer in the end.

36:03So you're exposing them to toxicity and you're not helping them survive. So it's actually you're doing net harm with that drug. The idea of a surrogate is that you have a measure like tumor shrinkage that predicts benefit down the line, and that helps you develop a drug much faster. A lot of drugs have been brought to market much faster in cancer than otherwise would have been many really effective drugs because they've used these surrogate markers of efficacy like tumor shrinkage. But it's harder to apply this principle in other conditions, like more complex chronic conditions, where you don't have an objective measure that's as simple as does a tumor get smaller or not.

36:36If you're treating a disease that's eventually fatal, if you see that this biological response happens early on, then that's very highly predictive of improved survival down the line. And just finding more of these surrogates is a really, I think, valuable activity that something like the NIH and government agencies should invest a lot more money in developing. because as soon as you have an established surrogate that the regulators can use, you really speed up how quickly you can iterate on developing drugs. You can take it to the clinic, you can see if it works in the surrogate, and then you can decide whether or not to continue or discontinue based on that.

37:09You don't have to wait 5-10 years to see if someone survives or not. So something like aging, we're never going to get drugs to prolong people's lives until we figure out a surrogate for aging. Can you talk about RCTs, randomized controlled trials, and something like vitamin D? I love the idea that this tool, RCTs, it's an incredibly important process that's helped us learn a lot. What, if any, limitations are there in your mind to RCTs and learning? So the big limitation to RCTs is they take time, they're complex to run, and then people don't want to be in the control group, essentially. So you have to expose patients to something that they don't want to take, really, because you're entering an RCT for a drug.

37:50You want the active drug. But maybe take a step back. So the idea behind an RCT is that you have patients or people who are going to take some intervention and you want to figure out, does this intervention really work or not? And you have to find some way of assigning the intervention randomly, because if you rely on anything other than randomness, there's going to be some bias that creeps into the study. Let's say you have this new drug and you take it to doctors and you say, I want to test this new drug. then the doctors may be more inclined to give the drug to patients who are sicker who really need it and they're less inclined to give experimental drug to patients who seem like they're going to recover anyway so if you just do that non-random allocation you're going to see that people who take the drug die at a much higher rate or have a much worse illness than people who don't take the drug because you have this selection bias of doctors giving the drug to people who are less likely to have a good outcome so you need to find a way to eliminate those biases and you can do that with just total randomization.

38:50And so coming back to the vitamin D example, one thing in data and retrospective data that looks at vitamin D levels and outcomes is that you see people with low vitamin D levels have generally worse health outcomes. So they have higher mortality, they have cardiovascular disease. I think it's probably been correlated with every bad thing you can probably get. And there's a natural inference there that we should just supplement vitamin D to the people who are low levels and then they'll have less rates of all these bad diseases so then people do the trials and they do a randomized control trial when they assign people randomly to groups and they give them supplementation of vitamin d and then they measure outcomes on all these things like mortality and health and then what they find is that actually vitamin d has very minimal or no effect on most of these outcomes even though some of the data from retrospective analysis looks pretty strong.

39:45So how did that happen? What you find out is something like vitamin D is actually a marker for poor health in general. So vitamin D is produced by your body in response to sunlight. And if you are ill or poor health generally, you're less likely to go outside or you may be less mobile in general. So you may get less exposure to sun. And then you'll see correlation between just general poor health and vitamin D. It's like with push-ups. They say, if you can do less than 10 push-ups, then you're much more likely to die. Not literally the ability to do push-ups that determines how likely you are to die.

40:20It's just that being able to do push-ups is a decent proxy marker for general health. So RCTs are just the most effective way we know of eliminating all these biases that you don't know exist from the start. Because otherwise, it's just everything you try gets really confounded. You don't even know how it's confounded. Can you imagine an alternative or do you think this will be and needs to be the method by which we learn truth and efficacy across medical interventions? So I think RCTs are always going to be the gold standard. And I think when we can do RCTs reasonably, we should do them because they just eliminate so many of these bias problems that you just aren't aware of.

41:02And that's because even if you control for the sources of bias that you think are likely to exist, there'll be hidden sources of bias that you just can't control for. So I think we want to keep RCTs as a confirmatory method, but there are other tools in these instances where RCTs aren't feasible that have a lot of promise. So things like if you do a single arm study, which is just one treatment group, and you compare the progression of a disease with a matched digital twin of these patients, you can train with information from natural history data. We can make a good assumption about the rate at which these patients would have progressed had they not had the drug.

41:36So there's been a few approvals that have used some of these matched natural history data as an alternative cohort. And those have been in cases where there's been too few patients to really do large-scale randomized trials. And then you have these instances of super rare diseases where you may have just a handful of patients in the US, for instance. It's literally impossible to run an RCT and have any reasonable statistical outcome. You need to use things like digital twins or surrogates, single arm trials. You can't use RCTs. There's a lot of innovation in this digital twin idea and natural history cohorts.

42:10obviously a big driver of all this is profit. So we talked about how people don't like pharma in part because it charges so much for really valuable drugs and they make big profits and that bothers people as it relates to their health. But profits and revenue drives the motive for discovery and the U.S. has always been a leader here. Can you talk about what you've learned about blockbuster drugs specifically? You can define what you mean by a blockbuster drug, but it seems like the biggest drugs that represent, there's sort of a mini power law here that the top handful of drugs represent a huge percent of the entire industry's revenue.

42:45So talk about the business side here, the distribution of revenue, the role of blockbuster drugs, the good, the bad, the ugly. Love to hear what you've learned there. Yeah. So a blockbuster drug is a drug that makes a billion dollars in annual revenue. So blockbusters are really having outside imports in the pharmaceutical industry because they account for a huge proportion of the overall revenue of the industry. The way the economics operate is pretty analogous to venture capital and these long tail models where you have many losers who don't make much money or don't recoup the money that's invested in them.

43:19And a few really huge winners that will go off and generate billions and billions and billions of dollars and pay for all the failures many times over. So the industry is in many ways worth investing in for investors and for companies because there's always this potential of hitting it really big and having a mega blockbuster like Humira or Keytruda or the COVID vaccines that generate tens of billions of dollars in revenue a year. because if you look at all the drugs that get launched it's sad that you have all this work that goes into developing a drug and it takes me 10 years to get it to market or 12 years to get it to market and it launches and then it flops 55 percent of drugs that launch that get through that process they don't even recoup enough money to pay the average development cost of a drug and then you have a very small number of drugs blockbusters that 30 to 40 percent of the revenue of the whole industry is made by these blockbuster drugs, which are quite a small fraction of all drugs.

44:16But when I look at the numbers, something like 170 blockbuster drugs that were actively generating revenue, and then the whole long tail of other drugs is making very small sums relative to blockbusters. So because the industry operates like a lottery type model, where you just want to really hit it big, you have a venture capital, it distorts a lot of the incentives. So you see a similar thing in biotech investments that you do with VCs where big pharma is only really interested in developing a drug if it can potentially become one of these blockbusters. Because anything less than that is just unlikely to recoup the investment.

44:53So you get a lot of these small markets that patients really would like to have drugs for. They could be really beneficial and we know that we could potentially develop a drug against this disease. We have a good understanding of the mechanism, like many small genetic diseases. But it's just not worth big pharma's time or even biotech's time to invest in developing these drugs because the economics don't work out. Even if you get through the whole process of developing this drug and it works and you get to clinical trials, you maybe make a few tens of millions a year and that's just absolutely not worth it because it will cost you 100, 200, 300 billion just to get the drug through the whole development process.

45:26So there's a lot of, I think, alpha that could be unlocked in making the process much cheaper in how drugs are tested and validated, because it would mean that it's actually worth developing all these drugs that are not worth developing. What do you think about drug pricing as a key variable in all of this big equation? Blockbuster drugs are driven by a price times an amount. What drug pricing is something that comes up all the time as predatory or strange or US system subsidized, seems to subsidize a lot of the rest world, we make an outsized percent of the discoveries and you want to be compensated for that.

46:02It seems like a very complicated equation. So what, if anything, have you learned about drug pricing that you feel is different than the norm or how would you change things? So I think drug pricing is really difficult. One thing I guess to say about the U.S. is the U.S. has by far the highest prices in the world. I think people recognize that it's about twice as high as Europe on a net price basis. And the U.S. market for drugs is something like 40 % of the global market. And it's something like 60 % by revenue of newer drugs. So drugs launched in the past 10 years or so. So the US accounts for an outside share of the revenue that drug companies make.

46:35And part of that is because the US system is, for the most part, some of this is changing, but it is mostly a free pricing system where you can charge whatever you want. So when you're pricing a drug for the US market, you'll often, as a company, just try to charge whatever you think you can get away with, whatever you think the market will bear. And in a reasonable world where all the incentives are aligned, what the market will bear is close to the actual value of that drug. In healthcare, you have all these strange incentives in markets that don't function properly. But in the rest of the world, you have systems where you have governments doing the negotiation on behalf of the population.

47:10So they'll make assessments of the value of a drug based on metrics like how many quality adjusted life years does this drug give us? What's the benefit of this drug over the existing standard of care? And how much better is it? and how much of a price premium can we give it on the existing drugs. So both systems have flaws. I don't think it's possible to find a perfect system that satisfied everyone. If you look at the profit margins of drug makers, they make something like 10 to 20 % profit after you take out their costs and the gross margins are something like 80%. But really, that's a distorted picture of the actual profitability of drug companies because there's a distortion between when R &D expenses are paid and when the drug revenue comes in.

47:51So it's not necessarily a good picture of how much a drug company is getting for their investment. And if you look at how much drug companies are getting for their investment, actually, most of them are pretty near zero or very low return on invested capital. The industry as a whole is not that great of a business. It's very much like a lottery model that is just type of lottery that's attractive to people with biomedical PhDs, where most people are not making that much money, but a small number of drugs and companies are making super normal profits. and you can point to the examples of people who have very high prices and making supernormal profits like humira 20 billion seems like an incredible amount of money for a drug to make and i think if you look at that isolated example you can say just seems unreasonable that anyone is making 20 billion off this drug per year it's way more than what someone should reasonably get for producing this drug but you have to think more about the system of incentives you want to worth investing in developing drugs.

48:49And so the incentive needs to be very strong. And because such a small number of drugs produce so much of the revenue, it's like a real Pareto distribution in revenue, you need these lottery winners to make it worthwhile investing in the system, which is an unfortunate reality that because drug development is so inefficient, you need high prices and you need these super blockbusters to make the economics work out. When you're doing R &D, you're paying money that you have now for revenues in 10 years. So from a temporally discounted point of view, you need to think you're going to get a really huge amount of money in the future to be worthwhile you paying hundreds of millions of dollars in the next few years.

49:30I think it's a really difficult problem. If you put pressure too much on drug prices, you can very easily remove the incentives people to invest in the industry. Quite a fragile ecosystem that's built up. Once you start cutting down the tall poppies, you actually are removing a lot of the incentive to even develop any drugs. And it's just that you can do outsized damage by cutting down even just a small number of these huge lottery winners, if you like. You have to accept a bit of a trade-off where you are trading off innovation for how much you're spending on drugs. What about prevention? We've talked entirely about interventions where something bad happens in the body and we've developed ways of treating that thing or making it better.

50:16What about similar research that could go into whether it's lifestyle or other things that we do before these bad things happen to us that would prevent them from happening in the first place? Any thoughts on that side of the ledger? Yeah, I think prevention is difficult. Well, people in the US often say that the reason why the US doesn't invest in prevention is because a lot of people are on these employee insurance plans and there's a lot of insurance plan turnover. So any insurance company that invests in prevention for its members is unlikely to reap the benefits of those investments when they switch to another employer and get a different plan.

50:54But the problem is that if you look at other countries that do have these nationalized healthcare systems like the NHS, other European countries, they actually don't invest in that much in prevention either, even though they should, in theory, because it's a whole nationalized health system. They should be wanting to invest in those things. So is the problem really that there's a lack of supply of preventative treatments, or is it really there's a lack of demand for preventative treatments? And I feel like the issue is probably more on the demand side where people, unfortunately, don't have a strong demand for preventative medicine, and they don't take up preventative medicine in many cases when it's offered to them.

51:31So it's the whole thing with the argument about a Zempic and why don't people just practice diet and exercise? That's a more sustainable way of dealing with obesity. The reality is that people just don't do that. So they would rather just take Zempic when they are obese and have that treat their obesity. So I think there's a whole consumer attitude shift that needs to happen before preventative medicine ever becomes really established. It needs to be demand-driven. So you think about the healthcare system. Who is really the end consumer of pharmaceuticals and healthcare system? I think you can make the argument that it's actually healthy people who are enrolling in the plans and are paying the majority of the money that goes into the system.

52:07And they want that system to represent value for money for when they do get sick. So you have to have the people who are healthy have a greater demand for products of prevention that then incentivize companies to invest in them, insurance plans and health systems to actually provide these services. I'd love you to talk about two extremes. the things that have you in this entire world the most excited for the future and the things that have you the most worry? I think what I'm most excited about is not a specific technology, but more this Cambrian explosion, if you like, of modalities in biotech.

52:44So for a while, I think back maybe 10, 20 years ago, there weren't that many different types of different treatment classes. And since the rise of biotech, you have all these different ways of treating disease that are becoming established and have a lot of potential. So you have CRISPR, RNA interference, these CAR T cells we talked about, base editing, different variations of CRISPR. Our ability to control our biology is getting much better than it was even just a few decades ago. So this increase in our ability to control our own biology and intervene in biological systems and very precisely uses molecular scaffolds to push a system into a desirable state is really interesting.

53:27And just this flywheel of improving on the modalities that we do have. So you look at something like CRISPR, we just got our first CRISPR-approved therapy in the UK and the US, but CRISPR is already in some ways becoming a bit of an outdated technology. You have a lot of investment in what's next. So instead of making cuts, you're going to have these base editors that are changing specific letters in the DNA sequence. And that's a more effective way of precisely treating certain genetic diseases than making cuts, the CRISPR cuts. And then maybe after these base editors, you have prime editing.

53:58It's even more versatile in the type of edits it can make. And you have different types of editing. So just a lot of improvements in tooling for how we intervene in biological systems to treat diseases and seemingly like an increase in the rate at which we're developing these tools and applying them to the clinic to solve biological problems and to treat diseases. It's like general explosion in technology, our ability to control systems is really interesting. I'm interested in just general nearly new business models of the industry. So one trend you've seen over the past, I think it started probably in the 2000s, is this greater externalization research.

54:35So pharma companies, big pharma companies used to do a lot of research in-house. They've mostly externalized that to smaller biotechs that are venture capital funded. so the big pharma companies are really just like commercialization machines all the time and they do maintain some research labs but the purpose of the research labs in many cases just to validate external opportunities and test them in-house and have the expertise to actually meaningfully evaluate these external opportunities they're going to bring in maybe you'll see greater and greater externalization of more and more functions now we have innovation r &d externalized to little biotechs we have running clinical trials being externalized to these clinical research organizations that specialize in that.

55:13You can argue whether that's been a good thing or not, but it's just part of the externalization trend. And then you have more and more just pieces being externalized. So commercial analysis will be externalized in the future as well. So you may end up with other companies just being very specific IP holding companies, plus just managing the finances of actually selling and distributing these drugs. Every little piece of the ecosystem is focused on some specific component of the process, and maybe that will help it be more efficient. So I'm interested in that. I think you can argue that it may not be helpful.

55:42Externalization hasn't been completely helpful. There's been some trade-offs there. But there's a lot of redundancy in the industry. And if you're in a biotech who's getting ready to launch, people having to build up redundant capacity to run and execute trials in an efficient way. And then what I'm worried about is there's a genuine worry that it's not going to be worthwhile for a lot of companies to continue to invest in developing new drugs. So I'm worried that if we don't get the balance, rewards, innovation, and incentives to innovation, we'll end up with a system where companies will decide, okay, I'm not actually incentivized to really invest in doing meaningful fundamental research and spending 30 years tinkering on some opportunity that may eventually bear fruit.

56:24I'm actually just going to market the drugs I already have that are old. I'm going to invest in drugs that are very hard to copy. so even though once they do go generic no one's gonna be able to copy me so things like these radio radio therapies talk about car t's they have a whole process associated with them so they're quite hard to copy and then you're just going to sit on existing treatments and try to milk them as much as possible and use tricks and techniques to extend the patent life and make it hard as a copy and try and extract as much value for as long as possible and maybe not invest in this fundamental research that really drives meaningful products that has this element of unpredictability to it.

57:02One thing I think is quite dangerous in the pharma industry is just this over-reliance on an accounting type mindset rather than a recognition that there's a science driven industry, there's an innovation driven industry. The commercial parts of big pharma should be in service to the R &D portion of the organizations or the ecosystem. And you should really be like hiring good people with good intuitions about what's worth developing and letting those scientists tinker for as long as they need to on some of these really challenging ideas at the forefront of what is possible and give them enough time 10 or 20 years for these ideas to mature into actual products and then you can commercialize them and if the returns innovation are down too much and maybe you get too many for the managerial positions who are focused on the accounting aspect of it they just want to milk existing products and copy the big ones that we know have an existing market then you're not going to get the kind of meaningful breakthroughs like Keytruda, these immune oncology drugs, the GLP-1 agonists that we're seeing now.

57:58So yeah, that's one thing I worry about. Just a totally fascinating overview of one of the most interesting parts of the world and of the business world of just the technology and innovation world. I've so appreciated everything you've written and sharing so much with us today. In these interviews, I always ask the same traditional closing question. What is the kindest thing that anyone's ever done for you? I think this has a pretty obvious answer in my case. Many people have done many kind things for me, right? But I think I just have to go with probably what is the most common answer, which is my parents have always been an unending well of support.

58:30And they bailed me out of many low points in my life. So I'm very thankful for that. And now while other people have done me many kindnesses, nothing really compares to what my parents have done. Alex, thank you so much for your time. Thanks. 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.

59:22Thank you.