I sat down with Kevin Slavin to ask what it would take to make the next pandemic optional. We talk about wastewater, airplanes, genomic surveillance, and the political and commercial machinery needed to use tools we already have. This is part 3 of 4.

Transcript

Kevin: Whatever. And then when 2020 happened, it was just I was oh my God, I think this is our moment. I think that all these people that I worked with and had contact with, I think this ability to detect what microbes are in the environment was suddenly super fucking relevant. It was and so it was it's January of 2020.

And of course there's no human diagnostic tests, but I was but know how I would be able to find.

Pablos: It. And I know.

Kevin: How Jessica would be able to find it. I know how Chris would be able to find it. And Marianna Mattis, it's, of course it's going to be in the sewers. Right. basically from that moment on awakened, the possibility that this interest could become really consequential. Yeah. and so I spent basically five years, trying commercialize it and make it as useful as possible in the early pandemic for, just the presence SARS CoV 2 of COVID in an environment.

Which, I figured out ways to do it with surface testing and airborne testing and things that. but then it led to, 2023. about three years ago, I started working in proper, biosecurity in really, which was a real extension of the stuff that I had been doing 13 years prior, which was basically, we were building the infrastructure to detect the emergence of something weird and destructive anywhere on the earth, which was it's not a moonshot type thing, but it's a, it's a project Corona.

When they were we'll put satellites in space and we'll be able to see where all the Soviet missiles are. that was cool. Crazy. that was insane. They didn't have digital cameras, right? it was just okay, we're going to, we're going to do that at the microbial level. Okay. got going, and it was doing what's called genomic surveillance where, which was in a way, it was just a different way of saying what I had been doing and what other, these other folks had been doing for years, which is basically you just gather microbial material and you sequence it and you can tell what's there.

Yep. but we were doing it in between the CDC on the one hand and the intelligence community on the other. Okay. we were gathering genomic material from, what they would call austere circumstances, right?

Pablos: Unusual places, wet markets and, just unusual places.

Kevin: Right. just unusual places. Okay. and there's a lot of different ways that you can gather it. and there's a lot of different ways that you can sequence it. And,.

Pablos: As an example, I don't know, sticking a fishing pole in a sewer under a manhole.

Kevin: Public example is a good one. the, one aspect of the work that we were doing, which was public because it was with the CDC was airplanes. Cool. There were, there were, I don't think this program survived, but there were, and I think it has survived. a number of airports, in the United States. And some other countries where when long haul flights would come in, 250 people would get off the plane and then the little truck would come to empty out the wastewater.

And instead of just dumping it all in the landfill, it would take a sample of that. And that would go to our lab. And we'd be able to see SFO was one of them. We'd be able to see okay, 250 people just got off at SFO. and here's what arrived with them. And you don't know who it came from.

There's no HIPAA. it's not, it's not a non volitional test. but it's important to know what has arrived with them. And that was interesting data, but it's much more interesting data. If you think about it, it's not arrived at SFO. It's just departed from Stockholm or Shanghai or.

Pablos: Lagos or wherever.

Kevin: And so now you can actually get a, that's one signal that tells you what's happening in Stockholm or Shanghai or Lagos. Right. and so one is public and then you can, your mind can imagine a thousand other ways that you could gather.

Pablos: That type of thing. But even for that one, it seems sequencing isn't all that expensive at this point. Yeah. how hard could it be to just do every flight, every international flight at the least? that must have been speculated at the time. how do we do that?

Kevin: Yeah. No, two, there there's, there's, there, variables logistics, time and money. from a logistics point of view, it's actually really complicated. no one ever really thought about who owns that wastewater actually. Right. that was a complicated question. Right.

Pablos: Who owns it? Does it belong to Delta or does it belong to S4? Or the.

Kevin: People who search, who does.

Pablos: It belong to? I own my shit once I put it in the can. Right. I.

Kevin: Mean, just, you're just asking questions that no one's ever asked before. But, there were, and again, I don't know where it is.

Pablos: Sounds a very European approach to going after this problem. maybe. it was definitely. Europe asked who owns their shit with a waste. It was.

Kevin: Definitely different in.

Pablos: Europe than it.

Kevin: Was in the States. even in the States, there were.

Pablos: Regulatory questions. Yeah, right.

Kevin: There, if you're looking for something that what you're looking for, we were looking for SARS CoV 2 whatever, that you can do very quickly and very cheaply. Yes. But if you are looking for something and you don't know what you're looking for, then you have to use what's called metagenomics. And metagenomics, I'll just, for the sake of the podcast, I'll just, say, basically, genomics is, I take a swab of your spit, I look it up against the human genome, which is, fully characterized and I know that you're human.

And so I just, I'm looking at, I have a reference point for the DNA that I just got. But if I give you a bowl of soup or, a test tube full of poop or whatever it is, where do you.

Pablos: Start? with.

Kevin: A bowl of soup, do you start with the tomato genome or the cow genome or whatever? And so you have to do what's called metagenomics, where you basically break the, you break the existing DNA roughly a bajillion fragments. And then it's a complex computational task to reassemble those fragments into whatever they're supposed to be probabilistically, given all the givens against the entire tree of life.

Right. Yeah. When I was first doing this in 2013, that was a computationally unbelievable. Yeah. it was.

Pablos: Heavy. But now I could do it on Apple watch. Broadly. yeah.

Kevin: But it's still, it's still a task. It's still, there's still compute.

Pablos: Yeah, compute is the thing we're good at now. Yeah. if we project out a hundred years when humans are way cooler than they are now, and we don't have all this regulatory bullshit because somehow AI solved it for us, then what we would do is we'd have something sample, every airplane, every subway, every hour forever that would go into a giant machine learning model would figure out if something new, popped up.

Exactly. Then we'd take the new thing and we'd immediately start simulating it and figure out, is it dangerous? Is it good or bad? Is it, where does it go? Figure out, do we need a, some vaccine for it? Exactly. Design that. How, and we would basically have rapid response when a new thing popped up. Exactly. the real question now is not we have the toolkit to do that.

Yeah. The question isn't really should we do that? We know we should do it. The question is really can humans get our shit together and actually do it? and so that's the, that's basically every technology lands in some story this. Yeah. what we know is that in the, in a hundred years or it takes us 200, we get to a point where something a COVID would get spotted on the first plane, hopefully before it took off, but certainly by the time it landed, maybe we could put the sequencer on the plane.

By the time it lands, we know there's something novel on it. Yeah. We would, stop, drop and roll right there. Yeah. Figure out how to contain it, which would be step one for sure. Keep it from going anywhere else. Yeah. And then while we contain it, all the computations happening to plan a vaccine and intervention, whatever needs to be created, manufacture it, distribute it, make sure it's in your, pop tarts tomorrow morning.

Kevin: And there are folks, who are currently building every aspect.

Pablos: Of what you.

Kevin: Just described. it's underway. I should just give a shout out, Nikki Romanek and Ashish Jha, who are former CDC, they have a company now called BioRadar where they are setting out to do that first part,.

Pablos: The detection. Great. Yeah. Somebody throw money at them.

Kevin: Right. As best I understand,.

Pablos: That's underway. Okay.

Kevin: Cool. and there are folks who are just working on being able to rapidly create mRNA vaccines against any target, once you have the sequence and.

Pablos: Then being able.

Kevin: To scale, and build Bioforges that can just create a lot. Yes. There are folks that work on that. Yes. And pieces of this, are being built globally. Right. And, I think the main thing it really, at a certain point, it just requires political.

Pablos: Will. Yeah, that's right.

Kevin: And that's a, that's.

Pablos: A trick. It's a.

Kevin: Trick. but on the science and technology side, things are moving right along in that. A lot of people are very worried about biological threats in part because we just lived through one and in part because maybe there are things that make it easier now. I'm not so sure about that, they definitely will become more.

Pablos: Frequent. But this is exciting thing because I think people are panicked because they don't, they're oh my God, the next COVID is going to be a hundred times worse and it's coming next year. And, but, probably is true in the grand screen. We're not done with them, but we have volition in this. We have something we could do.

And honestly, I don't know of any technology missing to dramatically improve our ability to respond. Totally agree. We have the toolkit. We may not have the political will and the business models and the infrastructure machinery in place for humans to actually do the right thing, but we actually have everything we would need to do the right thing.

I agree. And I think that even knowing that is very important for people just to know. They think we're all going to hell in a handbasket. The truth is we have a self driving car that goes to hell in a handbasket. All you got to do is tap on the screen, pick a new destination.

Kevin: But it was also for me, working in that for a couple years, first of all, it was really interesting because it was working with, I was really working with folks who came from a deep public health background and folks who really came from a deep intelligence community background. And those were those are just fascinating conversations, right? because one is, essentially about, national security and the other is about global health.

And those don't have to be at odds. But what happens when they're at odds?

Pablos: There's all kinds of.

Kevin: Interesting questions. And for me, the thing there were two things that emerged in that time that were really surprising and interesting lessons. one of them was I'm simplifying this in part because parts of it that just can't be talked about. basically sometimes we would get data signals from places, other countries, places that were worrisome. Yeah. And but you don't know, could just be a weird artifact on the plate could be that the person who was handling the sample had something.

You don't always it's a first signal. It's not a definitive. Yeah. And so we would find something worrisome from some country. Let's just call it Sweden to just because no one is worried.

Pablos: About Sweden. Let's Call it Narnia. Right. something from Narnia showed.

Kevin: Up. Some worrisome sample out of Narnia. And, effectively you would get Narnia on the on Zoom and just be hey, we found this worrisome situation. Can we compare notes? And what was fascinating to me, and this was in 2023. This was when there was still a lot of international cooperation. What was fascinating to me is that the answer was always no.

We don't share that data. And I was what do you mean you don't share that? we're in a public health emergency. This is about global health. We found something that we don't think about.

Pablos: Can we.

Kevin: Just help each other here? And the answer was basically always no.

Pablos: Those Narnians are motherfuckers. Nobody knows that. Yeah, people think that they're all magical, whatever. But in.

Kevin: Fact, they're very possessive of.

Pablos: Their data. I have almost no data on Narnia after all these years.

Kevin: And it's true of basically every nation on Earth is that they don't share non human data. And that was shocking to me. Now it feels oh, of course they don't, because I understand why not. Which is a whole bunch of things happened about 20 years ago that led to, it started with Indonesia, where there was a whole bunch of things that, broadly speaking, pharma companies were taking advantage of the fact that folks were sharing data freely.

And they were commercializing it to extraordinary ends. And the folks who had put in the work, and also from whence that data emerged, felt well, we gave you access to it, what's in it for.

Pablos: Us? And the.

Kevin: Answer, at least in 2006, was is in.

Pablos: It for you.

Kevin: And so that led to a crisis. And basically, everybody stopped sharing anything at all. And that is the world that we've been living in since then. And that is, it's not the reason, but it's one of the contributing reasons.

Pablos: Let's break that down. Is this the glia monster example is one of them? Are there others? Or what was the origin? What was the problem in Indonesia?

Kevin: It was our friend H5N1, which is the flu that is 20 years later in all these American cows. It just made its way to Australia a couple weeks ago for the first time. And it was spreading and it was generating variants that were concerning. When it hit Indonesia, I don't quote me, but I think the number was that 50% of the people who got it died.

It was either 30% or 50%. Very bad number. plague numbers. And Indonesians were good scientists. they isolated that variant of the H5N1 flu at that time and they shared it. And an Australian pharma company was cool, we're going to develop a whole bunch of vaccines for it. And then they tried to sell that back to the Indonesians for some amount that the Indonesians didn't feel was fair.

And so then six months later, when there was another variant that was far more consequential, it was when it first spread into mammals, also Indonesia, the WHO was the World Health Organization was well, can you give us the data? And the quote from the Minister of Health of Indonesia was, we're not stupid. We know that this data has value.

And that was the moment where a whole bunch of people just started wringing their hands, how is this going to work? And it led to this weird policy and field called ABS, which is access and benefit sharing. And where basically, and this was it comes out of the Convention for Biological Diversity, which is a multinational NGO that is trying to solve these problems.

And they basically said, look, you have to break it up into, there are two constituents in any of these. There are the folks who are providing the data, which could be Indonesia, and there are the folks that are using the data, which could be Pharma Company X. And the idea is that if Pharma Company X wants access to that data, whoever provided that data needs to share in the commercial benefits of that access.

Okay. Which doesn't feel an extremely contentious.

Pablos: Idea, right? Yeah.

Kevin: It feels oh, okay, we can get it. But it was very contentious. And it led, so everybody convenes in Nagoya. This is 2010, I think. Maybe it's a little earlier. And 193 nations. And they're okay, here's the idea. Access over here. Benefit sharing over here. And how are we going to make that work? And basically 130 nations voted for some version of ABS.

And 63 nations voted against it. Because that would really mess up capitalism.

Pablos: The U S and Australia vote against it.

Kevin: The U S led the vote.

Pablos: Against it. And.

Kevin: A lot of other folks voted.

Pablos: Against it. Basically, all the countries that have pharma vote against.

Kevin: It. Broadly speaking. And so it was 130 people who were we want to play the game this. And then 63 people were we want to flip the table. And so the table got flipped. And we've been living in the wake of that ever since. And what's.

Pablos: And the bottom line means if I'm any one of those majority countries that wants to get a piece of the action, if something economically viable comes out of a discovery made in my sovereign nation, if I'm one of those guys, it means if I find an H1N2 or whatever's next, I want to make sure that when Australia tries to sell me a vaccine or sell anybody a vaccine, I'm getting a piece of that.

And so that was what the structure of this benefit sharing thing was. Exactly. And basically it all fell to pieces because those countries that are discovering data can sign on with each other, but it doesn't matter because all the money is coming from the ones who wouldn't sign up. Okay. Pretty typical. Yeah.

Kevin: There's a really interesting lawyer named Bart Van Voren who did an analysis of over 10 years, how much has actually been transacted as a result of Nagoya? Okay. And the answer was around $55 million in total over 10 years, which is that's not a lot of money in the scheme of all of this, across all these countries.

I think the intentions of it are correct. I think the general structure of it is right. But what are the incentives to comply? There's none. that's been messy. And then what makes it especially complex is that up until now, we've basically just been talking about pathogens, about the things that we need to get that data because they're scary.

But there's also, depending on how you count somewhere between half a billion and a billion species on Earth, we have full genomic sequences on somewhere between 20 000 and 50 000 which is obviously.

Pablos: Absurd.

Kevin: It's obviously not the way.

Pablos: Tens of thousands out of billions.

Kevin: It's absurd. And what's the bottleneck? It's not a technological bottleneck. It's not purely an economic bottleneck, because sequencing is not a very expensive thing to.

Pablos: Do these days.

Kevin: It has to do with the sovereign rights around that data. And this is in the wake of Nagoya, which is that all the life sciences have been broadly sophisticated. Not all of them, but most of them have, in their modern instantiation, been built on the back of discoveries that we've made from digital sequence information that came from nature.

Whether that's lizards or flowers or deep sea protozoa or cone snails or the bacteria that grow on certain trees. Those are what have allowed us to have all the magic that we make. With everything from painkillers to cancer therapeutics to cold water detergents. Those come from specific protozoa.

Pablos: Give me a couple examples.

Kevin: Prealt is a non opioid painkiller that's often given as if nothing else works and you have back pain or certain types of cancer. Prealt is given and that is derived directly from the DNA of the cone.

Pablos: Snail which I.

Kevin: Think is Brazil if I remember right. there's that. There's a periwinkle flower that only grows in Madagascar that the digital sequence information produced vincristine which is a common cancer therapeutic. Oh, and there's a really good one.

Pablos: I think I saw that in the movie.

Kevin: Was it called the Madagascar Periwinkle?

Pablos: It's one of the characters in the movie with the lemurs and whatever.

Kevin: And then there's something in the soil just on Rapa Nui on Easter Island, which is technically Chile. There's something that is just in the soil microbes there that from whence derives rapamycin. And rapamycin is essential for transplants. It has an immunological function that is essential when you're doing transplants. Wow. But it comes.

Pablos: From… Chile is getting rich off this.

Kevin: Yeah, Chile, that's why they're a top GDP nation. It's because they have this incredible natural resource that was basically.

Pablos: Stolen from it. Yeah, okay.

Kevin: And there's lots and lots of these. Sometimes they end up in litigation. Sometimes that litigation is successful, sometimes it's not. More commonly, they end up with patents being revoked from major markets because Brazil is well, if you took it without a permission, you can't sell it here. And we don't recognize your patent or export bans. And sometimes it really is litigation.

Yeah. And now increasingly, even just as an academic, if you publish a paper with digital sequence information that's not properly acknowledging its source, they can demand that the paper be retracted. it's not even just on a commercial level, just on a knowledge distribution level. And so I really woke up to this through biosecurity, but it's a bigger story.

And so that was half of it. And the other half of it was when we were working in biosecurity, I can't say too much about it, but there were folks from companies that would sometimes approach the folks who were generating this very unique data set through genomic surveillance. And they would say, we'd be really interested in licensing your data.

It's well, that data, it's not really, we can't really license it because in a weird way, maybe we're not supposed to have it. I'm not going to say we're not supposed to have it. That's not exactly true, but it was procured through surveillance. And by the way, for me, if you're doing that because you're looking for things that could kill people, great.

It's fine. But just learning oh, so the folks who use data really can't get the data that they need. And so for me, that was the other half of it was just realizing oh, I see, you have a whole bunch of people who don't want to share any data. And you have a whole bunch of people who really need the data that isn't being shared.

And policy alone is not going to solve that. I just, I don't want to put down because it's there's really amazing people who are doing amazing work in policy. I don't want to say that it's wrong or that it failed. It didn't. But I just think it's not enough. it's not enough on its own. And that's what led to the work that we're doing now, which is basically trying to build a marketplace.

We are building a marketplace between the folks who have that sovereign data on one side and the folks who are trying to get access to that sovereign data on the other side. And, one of those rare circumstances where you're I think the thing that's missing that could help solve this problem is capitalism. Right. OK. I don't feel that way very.

Pablos: Often. It's not that predisposed to that. Yeah. It's not it's not.

Kevin: The tool that I reach for, to solve a problem.

Pablos: It's the tool I reach for. I know. But.

Kevin: It's very real it's not a technology problem. not purely an economic problem. It's an incentives. Exactly. Right. it's and it's what marketplaces produce is they align.

Pablos: Incentives. Yeah. OK.

Kevin: In ways that policy can't. Because is basically trying to tell you what you should do as opposed to what's in your.

Pablos: Interest to do.

Kevin: And marketplaces have to operate in the interests of both parties that are transacting. OK. that's what we're doing. that's this very it's a very weird walk geography to this. it is still that same question, which is what is it that makes Indonesia different? What is it that makes Brazil different? What is it that makes Easter Island different?

And the answer really is, in this in these invisible layers of life.

Pablos: There's this view that I have often expressed as computational maximalism, which is how I think of, my belief that more compute is always better. that's but that isn't a comprehensive worldview. It's dependent on other fundamental belief, which is that I'm a data maximalist. more data is better. There isn't something I don't want data on or where I want less data on.

I want it I do want to measure everything about my body. I would to know what's in my microbiome. I would to know what all these neurons are doing. I would I can't measure all that right now. Yeah. But the quest to measure it, I think, is important. Yeah. And so to me, there's just no question that everything that can be measured should be measured.

Right. And, we can decide later what we want to do with that data and we can argue about who gets access to it and what all that stuff is. it absolutely is the right thing. Yeah. To try and learn as much as we can about us, about our environment, about these things. And that just comes from what you can measure.

Yeah. And now that we have the tools to measure. It's you said, got tens of thousands of living things that we've sequenced the genome of out of billions and billions. Yeah. And we've got to get them all. We just got to get them all. We have to. Yeah. I agree. Some plan.

Kevin: Yeah. Right. I think my belief is no one should be allowed to leave the planet and go to another one until we know everything about.

Pablos: This. There we go. I that. I think we should maybe send a few people. I got a couple of test.

Kevin: Runs. But I really think that should be the gating condition. Right. I don't.

Pablos: Want to get to fuck up other worlds until we actually understand what this.

Kevin: Is. Because this 20 000 out of a billion is bullshit. Yeah. Right.

Pablos: Listeners who don't know a billion is a thousand millions.

Kevin: Yes, It is. Yeah, it is hard to get your head around the magnitude what we are missing. Right. every time I try to present it visually, it's no, one pixel doesn't do it. Right.

Pablos: There you go. Right. one pixel on your screen. Yeah. Right. it's how many we've.

Kevin: Sequenced. Yeah. Right. Yeah. to realize that they're OK, systems that are necessary to unlock that knowledge, some of them are scientific. Some of them are technical. But it turns out some of them are geopolitical. Yeah. Right. And it's just it's that was the wait. What? Oh, that's how it works. that is how the world works. we we've drawn.

Pablos: That's how the world is.

Kevin: That's a very it's a very good distinction. Yeah. It is how the world is. And all the lines that we drew on maps that, languages end or, transit ends or whatever, that. There are also all these other forms of life that are trapped behind these arbitrary lines. Right. no one ever thought about it because we didn't know that they were there.

We didn't know how important they were. they're there. we just it's there's a very it's a very it's a very it's challenging and satisfying to have to figure out, how do we navigate between scientific, technological and geopolitical systems to unlock it?

Links

Recorded on August 13, 2026