EP 143: Debunking Healthcare's Biggest Myths with Zach Weinberg and Derek Thompson

28:31You can say the word abundance. It's fine. I'm not going to stop you. Yeah. Yeah. Are you saying abundance Democrats or are we? Just abundance people. Abundance people. It doesn't have to be Democrats. Yeah, yeah, that's fair. I guess, I mean, the criticism seemed to be that it's been super wasteful. And so they're reducing the budget by$18 billion to about$27 billion. And so it's still a significant amount of money in mass. And that I think the criticisms have been around wasteful spending and promotion of dangerous ideologies. And I think as I was kind of unpacking it, it seems like there was a grant to an organization called EcoHealth Alliance, which collaborated with the Wuhan Institute on virology, which was one.

29:13And then there's some research related to topics like gender identity and climate change, which the administration has deemed politically motivated. I think that's the steel man of it. I don't know. I mean, you guys help me out here. Look, I think this is a classic. I mean, this is just like politicians on both sides. It's like identify the problem and propose the absolute wrong solution. I don't think the argument that like the NIH is wasteful is wrong. Like there's plenty of things you could point to in how they think about grant reviews and who does those grant reviews. The overhead question, which is we can talk about of like how the universities tend to skim like a really large percentage of the grants.

29:56Funding some really stupid things that they shouldn't be funding here and there. But like, absolutely. Like reform is a great idea. I think it's one of those kind of industries or not industries, like agencies where no one's really taken like a critical eye. And I think Jay, the new guy running it, seems like a great reformer. But reform it within the existing budget. Cutting the budget is just cutting science, right? Like what we'd be better off doing is saying, hey, we could be getting more for our 40 billion or whatnot. And instead of just saying like, actually we're going to chop 40 % of it, like reform rather than cut.

30:33And no, obviously we, we, we decide like, oh, we're just going to cut it because we're mad at some like diversity study. And like, you could show up and cut the diversity research if you want, but then keep, keep the money and use it for biology research and cancer and Alzheimer's and all these things. Like, yes, maybe the mission creep has gotten in there. So I think it's a reasonable steel man, but the answer should be reform, not, not, not budget cuts. I think it's important for people to understand the level of care that's being used to do these NIH cuts. So there was a story that the journal Science reported on where the NIH announced that they were going to cut 80 % of the largest longitudinal study for women's health that has ever existed.

31:19I believe it's called the Women's Health Initiative. And the first people were just like, what are you doing? This is the exact sort of study that you said you wanted more of. It's about understanding the root causes of chronic disease and figuring out how we can make menopause more comfortable for hundreds of billions of women going into the future in this country and every other. Why would you cut this? So science and the New York Times calls the NIH, contacts the NIH, and says, you do realize you cut 80 % of the largest longitudinal women's study in American history. 48 hours later, they reverse the cuts.

32:03And then RFK goes, he either went on Twitter to say this, or he said it to a reporter, said, we haven't cut this program anymore. And the New York Times and Science Journal were lying about the idea that we were cutting it in the first place. The quotes from the NIH spokesperson are right there. Like, we're going to look into these cuts. Now we reverse the cuts. And then RFK comes in and says the cuts weren't even happening in the first place. This is the level of honesty and care that is being used to cut$18 billion worth of science. And by the way, it's of a piece with the level, I think, of honesty and care that's being used to cut just about everything across government with Doge.

32:43I mean, this is an administration that in order to turn the Department of Energy into something that was more focused on nuclear security, accidentally fired half the people who work in the only administration in DOE that has the words nuclear security in it. they're just slashing and burning and then asking questions later and acting affronted when anyone points out that they're going way too fast and not operating with any kind of care for A plus excellence in their efforts to reform some things that do truly need reform. So there is this like parallel world where I would be so optimistic that folks like Jay, whom I want to be clear, I've had nothing but polite and frankly, like quite interesting conversations with in my history as a economic and science reporter.

33:35I want to believe in a world where these folks are going to cut 18 billion dollars worth of science funding and find the exact right 18 billion dollars worth of science to cut. But let's be honest. Think about this from the perspective of like a VC. What percent of VC investments actually like pay out 10, 15 percent if you're doing a great job, 20? The response from the LPs can't be, hey, Andreessen Horowitz, we realize that 80 % of your investments fail. So how about this? We're going to give you 80 % less money. Just quintuple the efficiency with which you invest. It's a bullshit way of thinking about any kind of high risk, high reward strategy.

34:13And when you're dealing with the process of discovery, whether it's in science or technology, you simply are taking on the not just risk of, but inevitability of enormous amounts of failure because the few things that you put your money on are going to have returns that make the whole thing worth it. And there is, in fact, research showing that because of the enormous returns of understanding basic facts about biology, basic research science is one of the highest, most efficient forms of government spending that exists. So this story is like really, really important to science. It's important to me, but I really do urge people to think about the degree to which it's illustrative of a broader philosophy in government right now, which is, I think Zach said it perfectly, the correct observation that there's something imperfect in government followed by a profoundly imperfect and even destructive solution.

35:09There's actually this really interesting paper, I wish I had it, that talks about in biology in particular, just like NIH funding, this like crowd in effect. Yeah. So one of the downsides of government spending is like it takes the place of efficient private investment and you get worse outcomes because the government is like materially worse at allocating money versus the market. And that's true in a lot of settings. Part of why maybe in certain places, Derek and I probably don't agree about like what the government should be doing. In science, there's this notion of crowding in actually, which is like where you see more government investment in basic research, you actually tend to see the private investment follow.

35:47because what happens, and it's not rocket science, right? It's just like the government funding chases things that private wouldn't. It finds interesting nuggets of science. Biology is usually the best example. And then the private investment follows behind it to say, oh, shoot, now that we learned this thing about disease or about biology, now we're ready to fund what comes downstream of that, which is like the medicine, right? Like the drug discovery parts of it. There's actually a really effective taxpayer dollar. I would bet it pays back. I mean, you'd have to trace the dollar. But I would bet you it pays back in spades on just like investment in the United States.

36:23None of this logic is being considered, obviously, by the folks in charge. And, you know, this is this is what we get for electing angry people with with no real ideals underneath them. They're just angry, you know, just pissed and make bad decisions. Ironically, you know, it was one of the greatest crowd in sort of policy architectures the last few years is Operation Warp Speed. Operation Warp Speed had a pull funding where they said, if you make a vaccine that passes these criteria, this level of efficacy in phase three clinical trials, et cetera, we will give you billions of dollars to buy a ton of your vaccines.

37:03And the reason that was so important, rather than just what's called push funding, which is just, oh, you're a Medera now, I'll cut you a check for$500 million, good luck with the money, is that let's say that, you know, you, Logan, Derek, Zach, Jay Bhattacharya, all these people, we're all running pharma companies in 2020 trying to build a COVID vaccine. If I'm a little bit worried that Logan and Zach are ahead of me right now in the from a pipeline, I might pull back my investment by thinking, wait, if Logan and Zach beat me to market and I'm like the third or fourth COVID vaccine, then I'm not going to make any money on this vaccine.

37:36No one's going to take my vaccine. So I'm going to pull back on my investment. But if the federal government is guaranteeing me the same$5 billion it's guaranteeing to Logan and Zach, I'm going to have peddled a medal on this vaccine development process. and I'm going to maybe find other investors who can pull in money as well. So it has like, I think the way that sort of pull funding mechanisms can work in addition to push funding mechanisms like simple NIH grants is a really beautiful way of government reconceiving its relationship to invention. And this is the kind of thinking that I was like hopeful we might get, like I'm not a Republican, but like I was like, there are smart Republican tech people who are flying to Washington.

38:15And I know that they know a lot about how to use markets to create products that don't exist yet. And wouldn't it be cool if you took all this NIH spending, which is push funding, here's scientists, here's an R01 grant, you know, a couple hundred thousand millions of dollars, you're off for four or five years, do your thing. With pull funding, if you build this miracle drug for Alzheimer's, let's say, we will give you$10 billion, even if you were the 10th company to build it. Now everyone get in and take in a bunch of money from private banks and build some miracle vaccine or some miracle drug in Alzheimer's.

38:50I was hoping we were going to get more of that. And instead, you know, we didn't get, it seems like we're not getting anything on the building side. We're getting everything on the destruction side. It's this concept of like having a guaranteed customer on the other side. Exactly. It's demand uncertainty. Yeah. Yeah. Yeah. We actually have this issue. And what's crazy to me is people have known about this issue for 15 years, which is in antibiotic resistant bacteria, essentially, where you've got a mark, there is no market for new antibiotics, which I don't think most people realize, but like, there's almost no way to get paid to develop a new antibiotic.

39:30And it has to do with the way we reimburse for use of antibiotics in the inpatient setting, meaning once you're hospitalized. We don't have to go into all the nuance of it, but essentially like the hospital gets paid kind of like a fixed rate for certain hospitalizations. And if you add more drugs into the cost structure, like the hospital eats some of the costs. So they don't really have an incentive to use these new antibiotics. Therefore, there's no real customer for them. They also won't buy ahead, right? So they won't buy and stockpile because it's too expensive. And so basically like there's no guaranteed market for antibiotic resistant bacteria, like for new antibiotics, the best way of saying it.

40:10And by the way, this is a major, major killer of Americans, right? At a certain point, like you get some sort of, you know, of bacterial infection, the existing antibiotics we have don't work, or you've seen them before, and your body has developed resistance, and you die in the hospital of a really nasty infection, essentially. And like this kills tens of thousands of Americans. And it's super well known. And the reason nobody funds this kind of research is there's no customer on the other side that you can bank on. Perfect spot for like another Operation Workspeed style model, which is to say, okay, here's what we're going to do.

40:47United States government is going to be a guaranteed buyer of your next antibiotic. We'll stockpile and then we'll distribute out to the hospitals as needed. But you need that guaranteed buyer or nobody is going to fund the underlying drug discovery. the amount of money that goes into new antibiotic discovery in the United States is like basically zero. It's really sad. And like, it doesn't matter. Let me tell you, you get a bacterial infection. It doesn't matter if you're Jeff Bezos or the guy on the street, like you're going to die from this thing. If we don't have proper antibiotics, there is no like safety option that you can throw money at, at that point in time.

41:25And it's just like another example of like, We know these gaps exist in scientific funding and for whatever reason cannot get the government to act in an intelligent way because it gets lost in whatever other healthcare reform. There's multiple examples of this where you need the government as a customer and your version of Paul. By the way, another great way of incentivizing people to build something that works rather than just spend the money on essentially giving government handouts. If you pay for the actual product on the other side of it, there's a real incentive and then let the markets work behind it.

42:01But yeah, we're not getting any of this. We're getting this and Trump coin. So it's great. I mean, if we were to try to figure out why this area has been cut, and we've talked about the implications, but there's the time horizon point of this, right? Which is hard. It's totally kicking the can. Again, investing in this is kicking the can down the road of the problems that might come up from it or might save, stave off. And it's kind of a rounding error of the overall budget in the grand scheme of things, right? $18 billion on a$35 trillion budget. It's just no one wants to go after the entitlement spending that actually causes the problems for the budget and trying to find different reforms around it.

42:46And so this, you know, we have these weird things that no one really advocates for in a meaningful way and they get conflated with a bunch of ideologies and stuff. And so it just ends up getting cut. It's a shame because we will never know the counterfactual of what would have happened otherwise. Right. We can sit here and complain about it and be like, oh, this is such a shame. And you know what? In 10 years, no one's going to know what the path not taken would have been. and when the next pandemic or whatever it is comes up and we don't have the drug, who's to say we would have had it back then if this had been kept up, which makes all the incentives really difficult to invest behind this and keep it going.

43:22Well, this is also, this is the drug pricing argument, right? It's the same underlying. Maybe unpack that. Yeah. Give the overview, I guess, or a little primer on that. Yeah. I mean, look, I've been on all sides of this, right? both as a startup. I worked at pharma when they bought my last company. Now we fund early science. So I've kind of like seen all pieces of this. And it's the same idea of if the reward on the other side for taking risk is not big enough, you simply don't take the risk in the first place, right? And like that applies. That's not just like a drug discovery thing, right? Like that applies in almost every context.

44:03You need a big - Venture, anything. Absolutely. And one of the challenges in drug discovery is simply, it mostly doesn't work. It is absurdly complicated to go from, I have an idea for a potential future drug, and actually getting through all the layers of complexity of testing, and just optimization of the actual drug itself to get something to market that actually works. And so if the reward on the other side is not big enough, it doesn't get funded in the first place because as an investor, you do these expected value calculations. Like we do that at CurieBio. Absolutely. We look at the number of patients.

44:45We look at the chances of success. We look at the potential pricing on the other side of it. We do all the math and we make sure that based on what we think is the chance of success of this happening, whether to write the check in the first place. And so we have this giant debate about drug pricing, right? Because drug prices are very expensive in the United States. That is absolutely true. But part of the reason why we have premium drug pricing in the US is to incentivize innovation in the first place. And if you want cheap drugs now, you will get them and then you will get materially fewer drugs in the future.

45:21I kind of think of it as like, are you comfortable knowing that basically all the medicine that's ever going to exist, exist right now, or is in the pipeline. Because if you, if you materially drop pricing, people are simply not going to invest in the first place. And this is the debate because it's somewhat of a, of an ethics question of like, how do you think about investing in the future versus, versus investing now? And so drug pricing can get really, really, you know, uh, contentious because people are on the hook for, for, for decent numbers, but it is what drives innovation, High drug prices are what drive people to invest in this in the first place.

45:57And that is the trade that we are making. I would also mention drugs as a category, think pills and injections and all the small molecules, antibodies, they go generic. So when you pay very high prices today, you are only paying them for a subset of the overall time, think like eight to 12 years roughly. And then you get them and your kids and the future kids get it essentially for free. right? So there is a giant downstream reward in drugs. Do you know what doesn't go generic? Hospitals, doctors, right? Surgeons don't go generic. We don't have a generic surgeon doing knee replacements and hip replacements.

46:40And we argue about drug pricing because it's very acute and you see it, but actually as a percentage of the premium dollar, meaning for every dollar we invest in healthcare, basically less than a dime of every dollar actually goes to medicine. The rest is going to the doctors. And so like, we talk about the drugs is like this big problem, but actually like, it's a beautiful incentive to make new things that eventually get really, really cheap. And all the stuff that where the healthcare costs actually really go, which is to the physicians, that just kind of continues to go up and up and up. And all of this nuance, I think just gets lost in what is a very emotional debate of like, why is this, you know, therapy I'm trying to go on costing me a few thousand bucks to get on it.

47:19And this, this is why it's like, you know, that's what got it there in the first place. Zach, what do you say to people about subsidizing the rest of the world with our innovation? We do. Yeah, we absolutely do. It's just the cost of doing business. This is my same, like Trump gets the problem right and then he gets the solution wrong. He's always like the Europeans are freeloading on American innovation in biotech. Like, yeah, yes, they absolutely are. You know, we pay a premium as the United States. that is what drives the investment in the incentive in the first place. Like on a typical drug, the United States is likely to be 50, sometimes 60 % of like all global revenue for that drug.

48:00So it's really driven by the United States. And then kind of see like a little bit in Japan and Europe kind of tags, tags behind. And what the Europeans do is they basically say, we're going to negotiate like collectively for the whole country. And we're only willing to pay a certain price. But because as a pharmaceutical company, you've already spent all this money on R &D. And so now you're just trying to get every dollar you possibly can. Eventually, you kind of cave to the Europeans. You're like, fine, because it's all incremental margin at that point in time. And really, the thing that funded it was the United States.

48:32So it's very true, like the Europeans freeload on what we do. Now, if the idea is like, hey, let's bring US pricing down to what the Europeans pay, congratulations, what you've now gotten are no new drugs. Like, that's it. Drugs will not be invented because, like, nobody is going to take this insane risk for a very low price. I mean, if I, at some point we should do it longer. Do you support Trump the same way that he's, like, you know, sending J.D. Vance to Germany to be like, you guys need to spend more on military because it's absurd that America spends a trillion dollars to create sort of like a military, a shell over Europe rather than Europe investing in its own defense.

49:12Is this something that you think the Trump administration should do for pharma? Yes, I do. I do because it will incentivize even more innovation. And it's unfair to the American taxpayer that the Europeans can basically free ride on the fact that the United States is an incredibly rich country, which sometimes I think we forget. And as a result, we're the ones driving all this innovation. They don't pay enough. Now, the tactics for doing that, I'm not sure. Hopefully, a very smart person could figure out how we kind of hold them hostage and threaten them a little bit. Maybe we ban exports if you don't pay enough or something along those lines.

49:52Like, hey, if you want to sell your pharmaceutical product in the United States, you cannot sell it in Europe unless the Europeans pay, whatever, 20 % off of what we pay. There's little things like that I'm sure you could do. uh but yeah it's absolutely it's absolutely true and they know it too right also by the way access to medicine in europe is materially worse than it is in the u.s because of this because they take their time right and they slow down the conversations and so we have stuff as a patient in the u.s we have drugs available here that europeans don't get for sometimes years behind us like it's not the access is really poor again right problem like he does get it right in the sense that like they are freeloading i just the solution is to pull them up not to pull us down if you pull them up things are good if you pull us down you know that disease that you're worried about getting in 10 to 20 years is not going to get solved before we hopped on uh derek you were asking about manufacturing a little bit and the speed uh do you want to dive into that briefly uh manufacturing speed of of of what everything or what you're saying oh we were talking about specifically?

51:00Yeah. It wasn't manufacturing. Yeah, no. This was a paper that Patrick Hollison tweeted out that I read. I thought it was really interesting. It's not the manufacturing of drugs. It's about the fact that the number of new drugs approved per billion dollars spent on R &D has declined by 50 % roughly every decade since the 1950s. And that means that the efficiency efficiency of medical R &D has declined roughly 80-fold in inflation in just terms in the last 70 years. I mean, that's just extraordinary. This is not just productivity flatlining in medical R &D. It's going down. And to drop the point home of just how strange and important that is, like if productivity stopped growing for overall manufacturing or productivity stopped growing for like agriculture, I'm trying to think of industries where we just had absolutely extraordinary liftoff curves in terms of productivity growth, it would be a catastrophe for the human race, for the US economy.

52:05And so the fact that we're seeing this slowdown in productivity in medical science, which I consider obviously, like Zach, incredibly important, is a really interesting phenomenon. And the paper looked at a couple of different factors that they considered primary causes of this slowdown in R &D productivity. One is what they call the better than the Beatles problem, which they summarized as like the idea that like yesterday's blockbusters, Zach was just saying this, is going to become today's generic, which means that every time we succeed in developing a drug, say statins or something, you know, PSK9 inhibitors, well, that means that in order to develop something that's better than the thing on the market, it could be very, very, very difficult, certainly maybe harder than developing the original statin.

52:50And not every industry is like this. If you're Crest and you build Crest toothpaste, you don't need to make a better toothpaste every single year. You can just print out the same Crest toothpaste and you're going to be totally fine. But in science, every breakthrough is a breakthrough that you can't have again. And unlike in manufacturing, you're not doing the same thing over and over, which makes you like really, really good at, say, making solar panels. That was number one, this better than the Beatles problem. By the way, to use your crest analogy for one second, the other view, the like twist on that analogy is imagine after 10 years of your toothpaste being on market, the government federally mandated that toothpaste was essentially free and you could no longer sell it.

53:35Right. Then you would have to create a new crest, essentially, which might be good for overall toothpaste. We might in that world, depending on how much people were willing to pay for, you know, new state of the art toothpaste, get a world in which like you brush once and your teeth are clean for the year. But in any case, this is this is the idea. And that not every industry is like medical science, where you are constantly having to come up with entirely new products because the product that you did come up with last year is going to be generic in a handful of years. That's number one. And the other one they mentioned is a regulatory problem, which is a familiar theme among folks that I read.

54:14I don't think everything is a problem of regulation, but I think that there's a lot of bad rules out there. And the paper quoted the Novo Nordisk, the Danish company that makes Ozembek. Their CEO once pointed out that, quote, quote, if printed and stacked, the millions of pages of regulatory documentation required for FDA approval with a total of nine million electronic links would exceed the height of the Empire State Building. So, you know, this is a this is not a study that came up with this statistic. It's the CEO of a pharma company. So he's fairly self-interested in terms of dramatizing the problem of regulatory burden.

54:51But those were two factors. And I'd be interested to know what Zach thinks. But I think this is like a really, really important area. Like if we found a way to come to make breakthrough science more efficient, it would just be like this is like this is the holy grail that people who are AI boosters talk about. like when you ask them, what really is AI going to do for the human race? They're like, it's going to make it much, much faster to come up with the sort of breakthrough discoveries in biology that help us live healthier and longer lives. So if we could do this with or without AI, it would obviously be a BFD.

55:29Look, the better than the Beatles problem is real. uh and if i had to if i had to pick like one reason for decline in productivity as it relates to medicine specifically there's kind of other angles here but like new new drugs that is probably the biggest and and far away and and i don't it's funny because i read the paper and they talk about like oh we're going to propose a bunch of solutions here and that's actually the one part of the paper they don't propose a solution uh which i i don't have a great idea either, but to try and what's the solution? The Beatles exist, you know, the drugs. I mean, yeah.

56:06I mean, my solution is if you just get America as rich as humanly possible to your abundance part, like then we just, we have more money to go and invest in these things. And frankly, the reality of, of, of science may be that the incremental drug is harder to find, right? The low hanging fruit has been found. The thing that you have to remember as just kind of like as a consumer of this stuff is the minute a new treatment launches, when you run a clinical trial for your next drug, you're not comparing against placebo in the vast majority of cases. What you compare against is what's called standard of care, which is basically like, what is the best available treatment for the population that you're focused on today?

56:54And what's happened over the last 40 years, because we've actually gotten pretty good at some of these things is that standard of care just keeps getting better. And in a way, our success causes the problem, right? Like the better the standard of care gets, the higher the bar is for you to beat. And the more that bar goes up, the more chances of failure. And so hence what I think is like a natural productivity challenge. You know, look, the hope obviously is new technologies allow you to make better drugs, the challenge is the amount of money that you have to spend before you know whether it remotely has a chance of being better is in the like 100 million plus.

57:35And that's the challenge right there, which is like, until you stick a new idea, a new drug in a human being, the amount of information that you have to predict whether this thing is going to work is still pretty limited. You've got some proxies worked on a cell, you know, in a Petri dish, it worked in a mouse, maybe it worked in a dog, maybe it worked in a monkey. And like, those are great incremental data points. But, you know, the number of things that like work in a mouse that don't work in a human is pretty extensive. There's your problem right there. Zach, how do you feel about the work that's being done on like digital cells?

58:09This idea that we could eventually create a computer model. I'm not explaining this to you, I'm explaining to the audience, I know you know what it is, but we can essentially create a computer model that captures the full complexity of the human. It's immune system, it's cellular system, metabolic, et cetera. And so you could essentially test, run a digital test of your molecular hypothesis for a drug in this computer before you dose it in a person and get something like 80 to 90 % fidelity or certainty on whether the molecule would do the thing in the human that it did in the computerized human?

58:55Is this a direction of science that you're interested in and optimistic in? No. No chance. I mean, close your eyes and look forward 100 years. maybe. Why? Let's just like one basic thing. To build a computer model that is accurate, you need accurate training data, right? Because the computer learns from observations, right? It has to have some input to make a prediction. And that training data needs to be accurate, and it needs to be at scale, right? You know, in the same, like, why are these LLMs so good at, like, I can make a Jalen Brunson as the Pope image this morning in like seven minutes, you know, seven seconds.

59:37Like it's cause I don't know, there's a lot of fucking pictures of the Pope and Jalen Brunson on the internet. And so like the computer can kind of learn pretty quickly. Also, by the way, if it gets it wrong, I just ask it to make another one. Not so true. You know, imagine what happens when these predictions are, uh, the, I don't know, my, my, my co-founder, as I asked him the same question to my co-founder, uh, who's a scientist, not me. And he gave me this really simple explanation, which I really loved, which is basically like when we observe what is happening in a cell in research. We are essentially observing that cell while it exists on a piece of plastic.

1:00:13That is not how this cell actually works in the body, right? Because we can't observe the cells in your body because you're alive. You can't come in and build a camera that looks like all your living cells. And so all of our understanding of how cells work at a molecular level essentially is like in some way fundamentally inaccurate. We can get directionally right and we obviously do learn things and you learn a lot from trial and error and all that kind of stuff. But like the input data for the model is like a Petri dash and literally a piece of plastic. And so we are so far from even understanding what happens in a cell that like from observation, let alone being able to feed that data into a model.

1:00:59And so it's not that I think the models won't be incrementally valuable. Like, yes, it'd be great to have a predictive model that kind of like gives me directionally where I go. But the idea that you could rely on that model from an accuracy standpoint is, I just think like fundamentally incorrect. I mean, think about it this way. Like how many times inside Gmail does Gmail, who has literally all the text of your emails ever, plus all the texts of humankind in the history of humankind to train from, it has like the largest data set ever created. How many times is the email response that it predicts for you, correct?

1:01:35You know, like - 5 %? I don't know, 65%. I don't know, I'd say 65. 65, maybe you give a lot of simple, like, thumbs up answers. I might have, like, very simple email exchanges of people. So I'm like, can I interview you? Yes. You know, and that's got, like, it's got, like, a billion X the amount of trading data. And so it's a trading data problem. And I think these tools are great. And I'm thankful that people are working on them. but like not it's not a panacea because it's a petri dish and then by the way even if it does make a prediction that you rely on you still have to go make the thing you have to physically manufacture it and then you have to confirm if the prediction was right and you still got to go like stick it in a cell line stick it in a mouse stick it in a dog stick it in a monkey and you got to do those in like sequentially you know you don't get to verify it right away like you do with the digital.

1:02:25I mean, look, if you, if you want to create a system where it's like any scientist can make any drug and put it in a human being, Rand Paul would like that system, I guess the libertarian system, right? Like, you know, like, ah, fuck it. Like, let's see what happens to people. Like if they want to do it, that's up to them. But like, we did that for a while. This is before we had the FDA. And then we got things like thalidomide, which is why we have the FDA in the first place. Right. Which was like a sedative that we were, we were giving to people and women in particular. And it turns out it causes major birthday effects.

1:02:58But we didn't catch it because we didn't test it. So like there's your trade, right? Which is like, yeah, we could move science a lot faster if we're willing to kill people along the way. It doesn't seem like we're willing to do that as a society for the right reasons. And so you still have this like, I think it was like the make test iterate cycle and the make test iterate cycle in computers is immediate. And the make test iterate cycle in atoms in physical stuff is not, plus you got the input. So no, I don't, I don't, I don't, I don't know. I don't think so. I actually, I mean, this is going to sound so ridiculous.

1:03:32I know you have to jump. I actually think like the only real way to incrementally produce better drugs over an extended period of time is to just throw money at it. Like you just need the reward to get bigger and bigger and bigger because the risk actually kind of goes up, not down as we get better. And there will be incremental things, absolutely, that make this faster and cheaper. And my entire venture fund is based on this idea of we can make things incrementally better, faster, cheaper. But there are certain constraints in a dog is not a person, a monkey is not a person, a rat is not a person, and they are only so predictive until you get into human trials and it's expensive.

1:04:17It's just real. On that optimistic note, thanks Derek. Thanks Zach for doing this. This is great. Absolutely. This is fun.