#4: Augustus Doricko - Weather-Modification Drones Are Coming Soon

Hey everybody, I'm Christian and this is First Principles. Today we're talking

to Augustus Dorico about the physics and economics of Rainmaker, a cloud seeding drone startup. Basically, they'll have their drones at a farm site or some other place where they want to literally make it rain, and when they see a cloud overhead that has the right composition and things inside of it, they'll fly the drone up there, drop their payload, and it will literally make it rain. It's an unbelievable idea, it sounds really ambitious and it is, but weather modification is

both too cool and too impactful to pass up. So I think you'll enjoy this episode. This is Augustus Dorico with Rainmaker on First Principles. Hey, Gus, thanks for joining us. Happily, Christian. So I am, I'm building Rainmaker Technology Corporation. We're an advanced next generation cloud

seeding company that's looking to enhance precipitation over agricultural watersheds first, then ecosystems in the American West, eventually terraforming all the deserts in the planet, relieving water scarcity and ending hurricanes. Just all of it at once. Well, no, indignation. A very fast baby step was actually the plan. And then the background that led me to Rainmaker was a blessed and circuitous route. Originally, I was studying physics at

UC Berkeley. The tongue-in-cheek thing that I like to say is like, Neil deGrasse Tyson tricked me into studying physics. I love it. And I loved it at the time, but the philosophical implications of E&M were not those which the cosmos made them out to be, at least not in the undergraduate level. And so I started studying data science just to provide myself some optionality, figure out what was going on. I was in the bay, so naturally the AI thing

was in the water. And then when the pandemic hit, I moved to Texas because I had some family there while still doing school online. I became a personal trainer because I like to work out and I wanted to meet some people there. I ended up training the biggest water well driller in Texas. And so he and I just became fast friends while working out. I highly advise anybody work out. There's lots of alpha at the gym, especially if you're thinking about starting a company. High

quality people. And so he told me about the water well drilling industry, about water policy and groundwater policy in specific, and that he was starting a consultancy in addition to the drilling company. So he had this large drilling company that was punching holes in the ground, and he started a consultancy to manage the regulatory compliance for people, drive a truck out to the farm, read an analog flow meter, fill out a report manually,

and fax it in to regulators. And I said, like, OK, first of all, why haven't you automated this? And then second of all, what are groundwater regulations like in the American West? Why are we regulating how much water you pump out of the ground? And so the answer to the second question was, Well, all of the aquifers in the American West, with a handful of rare exceptions, are either in a state of managed depletion or are totally depleted already.

So if you look at Scottsdale, Arizona, the Rio Verde housing development, they're banning people from moving in there, even after some folks purchased homes, because the city is so depleted of water, and groundwater in particular, that they can't service this new development and are banning new housing outright. Thankfully, last year we had outsized amounts of precipitation, ahistorical amounts of precipitation offsetting

this depletion, but that only replenishes the aquifer so much. And regulators realized that if we do not solve the

problem of water scarcity, particularly groundwater scarcity, then there will be total agricultural collapse in the American West. Cities like Phoenix, Tucson, all of Arizona, Salt Lake City, and Las Vegas are going to be uninhabitable between

like 30 and 80 years from now. And so they've started to take better accounting of how much water is being pumped out of the aquifers, so we know exactly what our budget is, and then have started imposing regulations on how much people can use and penalizing them for using

more than their budget. And so what we ended up building at Teraseco, that last company, which automated regulatory compliance with an IoT flow meter and then a bunch of backend software that I wrote for the reporting, what we ended up taking care of was just the inconvenience of

regulatory compliance for farms, HOAs, corporate campuses, and factories. And so I worked on that company in school for a little bit, we iterated, we grew it to the point where we had sufficient revenue for me to be confident to drop out of college, which I did. And then I spent a few years working at Teraseco, initially just in a technical position, building out all the software, building out like the hardware. And then eventually, osmotically learning about water policy and

water infrastructure. And what I realized was that there is a dialectic about how to solve water scarcity in America. And one camp says, sorry, no more agriculture, we hate pistachios, we hate almonds, two minute showers only, low flow toilets. And being more efficient about the water that you use is great. We should be good stewards of the resources that we're blessed to have. But if that's overindexed on, then you end up wandering in these like degrowth premises and

quality of life reduction measures. And so that's problematic. And I think that we're severely overindexed on that. Then the pro-abundance camp says, you know, The Sea of Cortez, the Pacific Ocean, that's all freshwater now, if we will it to be so. Let's desalinate everything. Let's build a $50 billion pipeline from the Sea of Cortez up into Phoenix. Let's spend even more money desalinating on

the Pacific coast of California and sending that inland to Nevada. And although I'm all for desalination, there's a couple of fundamental problems, one of which is the Californian power grid cannot support another, you know, 100 megawatts of desal, let alone like the gigawatts necessary for large scale desal. So that's a problem as it stands. Let's say that we solve that. Great. The second problem is the California Coastal Commission and other regulatory bodies

keep blocking the construction of desal facilities. So I'm not very bullish on our ability to quickly solve water scarcity with them. Eventually, I think we'll cave just out of necessity. But that's another

problem. So say let's solve the energy problem. And then the regulatory problem will still Water is so heavy, and if you're talking about pumping it 300 miles inland over two mountain ranges, then you're just not taking into account any of the pipe maintenance, any of the energy requirements, any of the pump maintenance necessary to service the operation

of pumping this inland. into the interior. So maybe in California, Los Angeles, even the Central Valley, perhaps, you can justify the unit economics of

sending water inland. You can't do that for states like Nevada, Utah, most of Arizona, New Mexico, and Colorado. And so I was distraught to find out that the two solutions to water scarcity are actually non-viable, and I started looking into all these other esoteric means of water production. And I think that brackish groundwater desalination is very promising. That's one technology that I thought about using in the subsequent company. There's about 310 million acre feet of

brackish groundwater in the Western United States. An acre foot, by the way, is if you cover an acre of land Yes, yes, exactly. It's how a thumb is an inch, things like that. And so there's tons of brackish groundwater that's ripe to be used. If some battery guy, shout out Teddy Feldman, or Ethan Loosbrock figures out how to use lithium, then like, great. But nobody knows how to handle the brine yet, so that's non-viable, or at least I didn't think

it was. There is urban water recycling, which is great. You know, I think a lot of people say like, ah, they're using toilet water and gray water for other applications now, like in SF or LA. I actually think that's a good thing. I think that there's reason to be skeptical of it, but we probably should. think of our water infrastructure as more of a closed loop, right? Like it's preposterous that we just spew

so much out into the ocean every day. And like the Hyperion water plant in El Segundo actually does a great job of recycling water, pumping it back into the grid, and also replenishing the aquifers in the Los Angeles basin with this. So that's good, but The lion's share of water use in the American West is agricultural, and you cannot, no matter how much you recycle, offset the demand of agriculture with

urban recycling, so that was non-viable. Then, I, through just perusing around on Wikipedia and places like that, found out about weather modification. Which naturally, if you're an aspirational technologist, if you want to see a sci-fi future, it's exciting, right? Like, can you control clouds? Can you make it rain more? Can you bust hurricanes? This kind of thing. This was all the promise of weather modification in the, you know, 50s, 60s, 70s. And then it fell

off. And we can get into later why it did. But what I found out was by all my independent reading and then going to something called the Weather Modification Association Conference, that weather mod has largely not been innovated in for 70 years, but one critical innovation occurred. Again, we can talk about it more later, but we figured out how to precisely identify the conditions necessary for effective cloud seeding to make

it rain more. And then also we found out how to measure very precisely how much water was produced and how much was exclusively produced by manmade intervention or anthropogenic intervention. And subsequently, for the first time in history, we can quantify the value and the value prop to farmers that makes this

a commercial, a commercializable endeavor. And so it was after attending the Weather Modification Association Conference, where I learned about the status quo of cloud seeding, where I learned about the areas that were ripe for innovation and commercialization that I started Rainmaker for the sake of producing more water at the entire scale of watersheds.

Did you know that there are companies out there that only make money by selling your data? They're called data brokers, and the data that they broker are things like your name and email and phone number, and even the names of your family members. They literally sell your information to scammers and spammers. They are the bad guys. But luckily, there are

good guys, like the sponsor of today's episode, Delete.Me. Delete.me is a service that helps you keep your personal information private, and they're actually helping me do that right now. All I had to do was go to their website and fill out a little questionnaire, and they've now removed my information from literally hundreds of data broker websites. And I know that this is anecdotal and it's just me, but I have seen a decrease in the amount of spam calls I've

gotten ever since I used their service. You should sign up for Delete.me to help protect your personal data. And as a listener of First Principles, you'll actually get 20% off a U.S. consumer plan. All you have to do is go to joindelete.me.com slash FP20 and use the code FP20 at checkout. That's joindelete.me.com code FP20. Check it out. Awesome, let's dive into what Rainmaker does, but I think maybe, can you set the background first of just the super basics, like what are clouds, what

is rain, what is groundwater? Let's start from the super basics and build Yeah, yeah, sure. So if you ask atmospheric scientists what a cloud is, you'll see a million different answers, and sometimes people will short circuit because differentiating between fog and clouds and ice phase clouds and liquid phase clouds, It's a broad space, but simply, right, it's a saturation of water content in the atmosphere, right? Like, you have so

much water that it begins to condense. It's not purely just vapor, but starts to turn into liquid droplets in the atmosphere. The reason why clouds hang up there is because there's convection pushing them up, right? Like, you look at a cloud, first pass, you think, oh, it's a cloud. Naturally, it's in the sky. The second pass is like, how is there water suspended

Yeah, yeah, exactly. But then the third thing you realize is there's there's convection from the earth, there's heat that's naturally pushing them up and keeping them suspended from pressure beneath it. And so, you know, that's that's simply what a cloud is. They come in all shapes and sizes. varied conditions, but the way that precipitation enhancement works is in

one of two ways. There's hydroscopic seeding, which is where you take, say, salt or calcium chloride, and you try to condense the water droplets into larger ones, such that they become heavy enough to fall and overcome the convection and pressure beneath them. And then there's glaciogenic seeding. Glaciogenic, think glacier, freezing,

right? And so what you do is you introduce a catalyst into certain clouds that freezes the water droplets, makes them into big enough and heavy enough ice crystals such that they fall and generally melt back down into rain by the time they reach the ground. And so the way that it works, practically speaking, is this. In certain clouds, you have a supersaturation of supercooled liquid water. And so, like, what is supercooled liquid water? It's water that's below zero degrees Celsius, below

32 degrees Fahrenheit. When you first find out about this, you think, that's

when water freezes. That's peculiar. Well, it turns out, actually, there's like a exponential decay curve explaining how long it takes for water at temperatures between zero Celsius and negative 38 Celsius to freeze. At negative 38, it's almost instantaneous.

At zero, it almost never freezes. You can see this in an experiment in your fridge where, like, if you put a glass of water, or in your freezer, rather, without any particulate in it, it won't freeze, and then you shake it, and that can induce homogeneous nucleation

of ice. Anyway, a bit of a tangent. If you have lots of water that is between, say, zero degrees Celsius and negative 20, right, and the reason why you have it is because more of it's being convected up, more water vapor is being brought up to that altitude where it's below zero, then can be frozen. So there's like a given rate

at how fast water can freeze in these conditions. You That's like parts of the, there's only so much of the cloud that can be freezing at a certain time or something, given that there's hot air going up and it's super cold up there. And so basically it's like that layer of Yeah, roughly. Like it's a meta stable condition. Like it can't last forever, but because more water is being brought up, then like the ice crystals can attach to and freeze naturally. You

have this super saturation of water that's below zero. And so the question then is, if you're trying to freeze it together so that it becomes heavy enough, so that the crystals or the droplets become heavy enough to fall, how do you go about doing that? And so, historically, people use something called silver iodide or dry ice. And silver iodide works very well as a nucleation agent because its crystal structure is almost

identical to ice. And so per every unit surface area, you have the maximum potential for hydrogen bonds to occur to that crystal. And then moreover, it's in this Goldilocks zone of hydrophobicity. And what that means is, if you have a super hydrophobic substance, water

won't be able to bind to it at all. It'll just wick off, right? If you have a super hydrophilic substance, so much water will accumulate on the surface, close to the surface, that none of the ice will be able to freeze and establish, like, it push up for, like, a greater contact angle. Like, you know how ice is hexagonal? It won't be able to push up because the pressure

from that close water layer is too great. Silver iodide is in that Goldilocks zone where it attracts enough water that it can freeze to it at all, but not so

much that it prevents freezing. And so those two features of silver iodide, the crystalline structure being identical or very close to identical, and the hydrophobicity means that when you deposit it in these super cool clouds, you get aggregation of the water droplets onto them, and then this runaway effect where you have the primary isonucleation, you have the initial isonucleation from the silver iodide, and then those crystals break apart, those

induce nucleation of other stuff, more water droplets freeze onto them, and eventually, because you have a runaway effect of isonucleation of heavy crystals, they begin to fall. They begin to precipitate as rain or snow. And that's how it works at the microphysical And that's, that's the, there's a surprisingly small amount of actual material you need to do this in a lot of these clouds, right? Like it's really not that No, no, it's astonishingly low. And so there's, there's two reasons why

that is. The first of all is, like, again, what is a cloud? It's very diffuse, right? Like, you've been in fog before, it's not like you're walking underwater. It's something to the effect of like 80 liters of cloud can be serviced by even like a single gram of the nucleation. Wow. Yeah. And so what you'll end up seeing is single digit kilograms of the nucleation agent dispersed over hundreds of thousands of acres. It's

astonishingly little how much you need. And in addition to how diffused clouds are, the other reason is it's not just the silver iodide based crystals that induce nucleation and

precipitation. All of the ice crystals that break off of those initial Yeah. Okay. So we're going to, let's talk more about the, like the different agents or whatever. Like I think, cause I think that you're not using silver iodide. Is that right? You're using some other stuff or you're like experimenting with other stuff. Correct. Okay. Well, let's talk about that in a second, but let's start with just sort of like the overview of

all of it. So the technical solution is like, you need to identify where these clouds are and when they're in the right conditions. You need to get that stuff up there. You need to get whatever agent it is up there. You have to disperse the agent. It has to be the right chemistry. And then probably the last step is you have to prove that you did it to your customer. That's probably an important step too. So let's actually start with how you identify those

clouds in the first place. Like, can you just pull that from radar? Is it something you Sure. So I think the best way to frame each of these is like, how has it been done historically? How is Rainmaker doing it differently? Historically, in the worst cases, people would just eyeball a cloud and

say like, hmm, that one looks good. And like, you talk to meteorologists and you got to respect their expertise because they are so lost in the sauce and deeply familiar with these processes that they can almost like intuit in an artistic way what good conditions are. But obviously that's not scalable. That doesn't work the lion's share of the time, of times. And so eyeballing stuff is historically done. That's no good. Very low resolution identification of clouds with radar. is also

no good. Basically just saying like, are there clouds? Is there water in the clouds? Like that would be the threshold condition. That's no good either. Rainmaker is different in that we're using very high resolution radar, primarily just radar, to identify when there is a super saturation of super cooled

liquid in cloud. And that's a pretty difficult problem because the If you know anything about radar, right, like you send a beam out and then you get a reflectivity back in, right, like a signal bounces

back. The reflectivity of supercooled water droplets, so like a itty bitty tiny droplet of water that is negative 10 C, is very similar to that of an itty bitty ice crystal that's also negative 10 C. But because of a lot of research done from the University of Colorado Boulder and the National Center for Atmospheric Research, Um, we now know and rainmakers iterating on this internally, how to differentiate with very high resolution radar between super cool liquid water droplets and like ice.

It's very small ice crystals. So we identify when we have super saturation in cloud. of supercooled water droplets between negative 20 and 10 C. As we're doing R&Ds, as we're validating our radar

results, we also have two different in situ measurements of these conditions to validate the radar. So first of all, we have a laser array attached to our research aircraft. that can differentiate between ice and water. And so that's pretty cool. It can tell you the size, concentration, and shape of your ice crystals or droplets. The primary way you differentiate between an ice crystal and a water droplet is just whether it's spherical or not. If it's spherical, it's generally more

like water. Um, so we have that. And then we also have this, uh, thing called the Rosemont icing probe, which is basically, uh, this wire that vibrates at a given rate. Um, and if it starts to freeze up, that means that there was super cold water that can deposit onto it. It'll slow the rate of vibration, that kind of thing. So in addition to the radar that we're using on site, um, we have two other physical in situ observations to validate that.

And we're going to continue using those for. foreseeable future as we validate all of our observations. So that's the way in which we sense the conditions necessary. The next problem, right, is bringing that nucleation agent up to the cloud. There's two ways in which people have historically done this. The first of which is, you know, Not a bad idea in principle, but it bears out really over time. And so there's

something called a ground-based generators, right? And you can basically think about this as like a smoke signal bonfire, where you have a plume of smoke that contains the nucleation agent in it, and you hope enough of that smoke gets up into the cloud, you know, 3,000, 5,000, 20,000 feet above you. And so we could do all the modeling to show you how that works, but you can probably intuit that it doesn't work exceptionally well. Yeah,

yeah, correct. And again, you don't need a ton of material. But if very, very little of it is actually getting to the portions of the cloud that are supercooled, then you would need to disperse lots

more. And it's also worth saying, Part of what's important about the high-resolution sensing is sometimes you'll have mixed-phase clouds, meaning some of it is super-cold water, some of it's ice, and sometimes you only have as much as, like, 100 cubic meters of super-cold liquid, and so you need to very precisely and granularly identify where in cloud—it's not anywhere in the cloud, but where exactly it should be deposited. So the first problem in historic delivery methods is like these

ground-based generators. They show promise, and really they're kind of just an artifact of when we thought it would be a lot easier to conduct weather

modification, precipitation enhancement. So they're not great. The alternative is better, but still not great, and that's using planes, whether they're crop dusters or slightly more sophisticated aircraft. You fly a plane up into the cloud, You either fly above it and drop the nucleation agent into the cloud, or you fly through. It's better to fly through, but there's a couple problems even still with that. First of all, the latency to delivery is

pretty bad, right? Between when you identify conditions that are viable for seeding, And when the plane gets tower clearance, when the pilot's ready, when the plane transits to the location, you can have hours in between identification and actual delivery. And so that means you can miss the conditions because that ice might have frozen or that water might have frozen. It might have dissipated in

another manner. So that's not great. The other thing is, You know, you only need like a kilogram of the nucleation agent, and taking an entire airplane up to carry a one kilogram payload is a very frivolous way to use resources. And so, because it's so expensive to have a plane on call 24-7-3-6-5, people are priced out. So there's... Farmers primarily are priced out of hiring a plane to conduct this operation for them. So there's a high latency problem. There's

a price problem associated with having planes on call all the time. And then also, you know, the turning radius of planes is only so great. You don't have like an F-35 conducting nucleation. That would be cool. It would be awesome. And in the future that we hope to build, we will have F-35s or Predator drones or something like that. China actually is retrofitting their military aircraft for this. We can talk about that maybe. The

obvious solution is to use UAVs and UAS, right? So what Rainmaker is doing is using very cheap drones that can carry these, you know, one to three kilogram payloads up to cloud that are installed on our customers' farms, on our customers' properties. And so you solve the latency problem because you have very quick response times. You can launch, immediately ascend

to the cloud, fly through. precision in the delivery because you can turn much more quickly you can fly through, you know, given hundred cubic meter chunks, and then they're way more cost effective because it's literally one drone without like a pilot on board. And also as we're saying. You're flying through icing conditions, right? And so they're not really safe operations to have manned aircraft fly through. Having a drone do that, provided that you engineer it

in such a manner that it can withstand said icing conditions, is also safer. So our delivery mechanism is cheaper, more precise, faster, and But the drones are flying themselves. So like you might not even be in the loop to say, OK, now go do it or So initially we have pilots on site. The FAA is gradually becoming more lenient and favorable towards UAV applications. But for now, we have a pilot on

site on the customer's property. Eventually, and what we're building out internally already, is a means to port over the coordinates that our radar identifies the highest concentrations of super liquid into a flight path plan for the drones. So they're totally free from humans in the loop. So another consideration, right, is the means by which you disperse the nucleation agent. So you need to aerosolize it so that it's very

fine, right? Like if you just drop a brick, that'll fall out of the cloud very quickly. Even large particles will fall out of the cloud very quickly, and you want them there as long as possible to induce as much nucleation as possible. Historically, people use flares to aerosolize the chemicals. The problem with flares is twofold. First of all, The ratio of the total mass of the flare to the amount of nucleation agent in it is terrible. Like 95 to 99 plus percent of the mass

of each of these flares is made up of the metal casing and then the pyrotechnic. So for every flare you take up, you get very little of the actual nucleation agent, so you need tons of flares. That is not great. And then the other problem is, in burning these flares, you're introducing all of this exogenous particulate that doesn't facilitate ice nucleation. And also you're introducing heat to the supercooled conditions, which you don't necessarily want because

that can interfere with the precipitation enhancement. So what RainMaker is doing is directly emitting the nucleation agent from an atomizer that we've designed. You can basically just think of it as like a fancy spray paint can. This sprays plume of the nucleation agent with a much higher ratio of INA, ice nucleation agent, to total payload mass. And so that makes our operation radically more efficient, drives down the cost because you don't need

all this additional hardware as flares. So that's the different dispersion mechanism. And then the last thing is alternative nucleation agent. Right. So traditionally we use silver iodide and silver iodide works. And in the quantities that you're dispersing this over huge areas. Right. So like a kilogram a couple times per season over hundreds of thousands of acres is totally inconsequential. Right. But the problems with silver iodide are this. One, you

know, it's literally silver, right? And spraying silver in the air is an expensive way to make rain. It can be justified in certain cost profiles, but it's still very expensive. The second problem is it only works at negative six degrees Celsius and below, right? And so what that means is You have limited seasonal and geographic operability where you can use silver iodide, maybe, you know, five times ish or fewer per season where you can see it over Phoenix and using glaciogenic agents.

That's not great because some of the most water depleted regions of the country are the hottest and so you have the fewest supercooled clouds, or at least fewest clouds that are below negative C, although maybe you have them between negative 6C and negative 2C. The last problem is, in sufficiently great quantities, silver

iodide becomes antibacterial. And if we're going to engage in the level of weather modification that we need to, to save agricultural watersheds from being aridified, save the Colorado River Delta and its ecosystem from drying up entirely, and then maybe eventually engage in exciting projects like terraforming deserts to make more land arable and ecologically flourishing, biodiverse. You're going to need to upscale the amount of nucleation you're dispersing. So Rainmaker is working on some

mineral-based and some biologically-based nucleation agent alternatives. Won't share exactly what those are. Eventually, we'll talk about them. But we're trying to make them cheaper, more eco-friendly, and more seasonally and geographically operable because of their hyofreeze. Sure.

So really, the sensing to the diagnostic of conditions, or the diagnosis of conditions, the delivery of the nucleation agent, the dispersion of the nucleation agent, and then the nucleation agent themselves are things that we're all innovating on relative to the state of the industry. And then also, because of our high-resolution radar, We can measure the density of precipitation that's in the path of our drone and the density and volume and subsequently tell farmers exactly

how much water we produced for them. And that makes it much, much easier to establish the value prop rather than just, you know, do some whiteboard math and say like, well, we project about this much based on the micro physics. And instead we So let's talk about the attribution thing just because we're right on it now. Do you want to talk a little bit about like, I forget there's some like fancy name, but like the zigzag pattern that people fly to attribute. You want to talk about

Yeah, yeah, totally. So historically, right, the biggest problem to commercializing cloud seeding was that nobody knew whether or not you were actually causing it to rain, and also nobody knew how much rain you were producing. Just because, you know, I fly up into a cloud, I sprinkle some magic beans, and then it rains, like, that doesn't mean that I caused it to rain, right? Like, it might have

precipitated naturally. And then even if you accept the microphysics, and we say that we understand that the process by which this induces precipitation how much more we got than we would have received naturally, how much more rain

we got from our intervention was impossible to say. The innovation from some folks at NCAR and CU Boulder and Idaho Power was, one, just using higher-resolution radar in the first place, but, two, having the procedural insight to say, okay, if we fly back and forth, if we strafe back and forth perpendicular to the direction of the wind in the area over which we're seating, What you should see what they hypothesized they would see and what they did see was

a track of increased nucleation and precipitation that was clearly and definitively man-made because there was uniform spacing between each precipitation line and that was exactly in accordance with the you know, trajectory of the drone. And so, that attribution, proving out, okay, this is how much rain was produced that was exclusively man-made, it was exclusively man-made at all, that is a critical input, right?

Relative to just saying, you know, here's our calculation, here's how much water we detected would roughly be in there, here's how much we project came down. Solving the attribution problem is probably the foremost consideration of Rainmaker because we need to do something different than cloud seeding companies have done historically, which is not overpromise and not be vague in the results that we offer, but say definitively, this is the upper

bound of how much we could produce, this is how much we did produce. And so, yeah, attribution is of critical import to us. And that's something, Very cool. Yeah, I think I read that Aspen Skiing and some other folks actually did experiment with this back in the 70s and 80s, but it was the attribution problem. It was they didn't know if it was actually causing more snow for their people, for their customers. That was why they didn't continue doing it.

Yeah, dude. I mean, cloud seeding ripped from the 50s through the 70s because we were so excited about its prospects, right? And we should be because the weather is something that ought to be controlled to the extent that we can with all of the second order effects modeled and mitigated. why there should not be more rain or snow when we have the technology to produce it is a question that I don't think anybody's given a good answer to. And granted, should

we play God? People ask that a lot. Isn't that the domain of the sky gods? Or is that really the domain of man? And the answer to that, I think, is one, we shouldn't play God. We should do everything that we need to do model out the risks and be cautious about this. And thankfully, like, our team of atmospheric scientists and the people in academia that we're collaborating with are very cautious and very concerned about those second-order effects.

But, like, I think that we should acknowledge the fact that the trajectory that we're on without intervention will result in the aridification and depopulation of the American West

and other regions in the world. And so if we can take a risk-adjusted If we can take a cautious risk and say this is an intervention with these potential outcomes that would be for better, some of our R&D, some of our water infrastructure spend should be dedicated to this because we'll either die a whimpering death or have a glorious future if we decide to cautiously move forward with weather modification. So yeah, agreed, man. And there was this thing called Project Storm Theory too.

So unfortunate. It was Uber based. It was trying to break hurricane, the Atlantic. Yeah. is trying to break hurricanes over the Atlantic before they hit the eastern seaboard. And we didn't have high enough resolution radar to attribute the results to our interventions. But it's something that we should be thinking about, right? Because hundreds of millions of dollars of property damage and insurance claims come from every hurricane. And if we can stop them, then we obviously ought to. So

That's an area for future product development. Killing hurricanes and nuking hurricanes with drones. Not nuking them. Making it rain earlier. On the topic of really epic names for these things, did you originally think about calling the company Thor? Or is that still a product name Yeah, yeah, yeah, yeah. We were thinking about calling the product, like the full initial cloud seeding suite, Thor, because it was triggering hail and orographic rain was

the idea. Because by introducing the nucleation agent, you can either do hail suppression or induce precipitation rain in the right conditions. So Yeah, we're trying to baptize all of these pagan rain gods and storm gods into the Christian ethos and generally base There we go. Slowly and surely. I love it. Okay, let's talk a little bit more about the agents. So you mentioned that silver

iodide is obviously what they're using today. And you're doing some cool, like you don't have to talk about specifically what they are, like what the chemistries are, but you're doing a cool set of experiments, I think, to like figure out in like a pretty ingenious way to figure out whether there's actually like the ice that's forming and how much is forming and stuff from these different things you're doing. Can you talk a little bit about that? Also, maybe tell the

Yeah, yeah, sure. So I'll speak to the history first. We have a poster actually of the good old boys that discovered it. The analog that I like to share is like the Powerpuff Girls. So the professor, right, he's mixing sugar, spice, everything nice into this vat to create the Powerpuff Girls, but then he spills Agent X into the vat, and that's what gives them their superpowers. So these three guys, Vincent Schaefer, Irving Langmuir, and Bernard Vonnegut, they were

conducting some totally unrelated cloud physics experiments in their cloud chamber. and spilled some dry ice and silver iodide into the tank. And they saw increased plumes of nucleation of ice in their simulated cloud. And so what they realized was, well, if this works in our little simulation, why wouldn't it work in real life out in the

field? And because it was 1946 and 1947, and you could do whatever you want, and dudes rocked, they just took up their own private plane and started emitting silver iodide over Western New York and induced the first man-made snowstorm. So that was incredible. That's how we realized that weather modification was possible at all and that silver iodide and dry ice should promise as nucleation agents. Now the question is, you know, because of those aforementioned reasons, we're

transitioning away from silver iodide. we need to ask, how do we identify new nucleations, right? How do we identify things that perform well at freezing, and how do we ensure that they're eco-friendly? And so you can do pretty simple tests on eco-friendliness, just traditional measures of toxicity, but how you measure the viability of an INA, we're

doing in three ways. The first of all is really high-throughput, low-resolution data, but pretty clever, and that's from the University of Alberta, or it was inspired from some folks at the University of Alberta. My friend Jesse Lee was responsible for designing this. It's essentially a control-break freezer, right? You can even use a traditional freezer if you want to do this experiment at home.

You use a control brick freezer with an array of vials in it, and those vials are filled with an oil that is denser than ice, but less dense than water. Then you put a droplet of your nucleation agent in that, like a droplet of water with the INA, into said vial, and it sinks to the bottom. What happens then, as it's undergoing cooling in the freezer, is when

it freezes, it floats to the top. So you can put a camera in your freezer and then measure how quickly things freeze in different temperature conditions and cooling rates to assess how functional they are as an INA. So our first-pass approach, sort of like our throwing stuff at the wall and seeing what sticks, is with this freeze-float system, that's what we're calling it. That's the first way in which we assess how well something performs in

immersion freezing. There's all these different ways that freezing can occur, but that's like our first pass. Then there's two other instruments that we're building out. One's called an ice spectrometer. That I won't talk about too much, but then the coolest thing is like a cloud chamber.

And so we essentially have, well, essentially are building a controlled box where you can induce very specific humidity, temperature, and pressure conditions, right, because the pressure is radically different at the ground

than it is 5,000, 20,000 feet up. And you can simulate what would happen if you introduce your INA to, you know, say, the top of the cloud, and then it flows through, like, a slightly warmer layer, see how survivable the ice crystal is that you produce up there, and measure the conditions as though it were in situ. And so the cloud chamber that we're building is inspired from some research at the University of Michigan. So, big thanks to those guys for figuring out how to

make cheap cloud chambers. That's cool. How big is that thing going to be? It's about three meters by three meters That's pretty big, actually. It is, it is. You wouldn't want You would suffocate and freeze and

A bad way to go out. That's super cool. And so for the first experiments you're going to do for the first times, or even for the first products you actually deliver for this farm, and we'll talk about that in a bit, but for those first experiments, are you going to use silver iodide, or are you going to use some other So there's two seminal cloud seeding experiments in America, the

ASCII experiment and the Snowy experiment. What we're interested in doing first is just baselining our results against academia. If we can reproduce that, then we know at the very least, like, organizationally Rainmaker's capable of this and everything

that Snowy produced is valid and can be reviewed by us because sometimes there's doctored results. We trust the folks that did those two projects, they're great people. But we want to baseline first with silver iodide. Again, in extremely low quantities over an extremely large area for an extremely large area, mind you, that nobody lives in, for the sake of baselining. So what we're going to do in our first campaign is use... Well,

actually, we're going to first use water, and then eventually we're going to use silver iodide. And then after that, we're going to start later in the year using Very cool. Okay, well, with that, let's move and talk about the sort of more business side of things, which is, you know, who are you selling

this to, and how are they buying it? I mean, maybe it would help to start with just who is your first customer, and how did you meet Yeah, so our first customer, and all of our first customers are large-scale commercial farms. And they're large-scale commercial farms in one of two regimes, either people that have already experienced the

total depletion of the aquifer and surface water near them. So they're farmers that have no water and cannot operate because they have no water, or they're farmers that are in a sufficiently litigious and penalizing regulatory regime such that because they're pumping more than they're budgeted, they're paying hundreds of thousands or millions of dollars per month in overdraft penalties. And so our first customers are all one of the They're

one of these types. And so I won't say where our first customer is or who they are, but they have tens of thousands of acres and they are paying about three quarters of a million dollars every month in overdraft penalties during the growing season. And so what we're trying to do for them is not, you know, help them purely stay alive, but to some extent, increase their margins, prevent them from going under, from spending too

much on penalties. And then what's interesting, too, is if you can increase the amount of precipitation that occurs over a 10-year span, then you can increase the amount of water that a farm has budgeted, right? Because the allocation is determined in part by how much precipitation naturally occurs. So not only are we offsetting the penalties that our customers are receiving, but we're enabling them to draw more groundwater as well. So that's our first customer. They do nuts

Huh, that's sort of counterintuitive. I guess, I mean like, the simple way I thought of it was, oh that just means the rich are getting richer. Like people that have more rainwater are just able to get more groundwater as well. But like the reason for that is that because that's like the replenishment rate sort of how you can think about it. Like you're not gonna deplete the aquifer as long, okay. Exactly. Wow, interesting. So, I mean, I guess this

is a fine time to talk about this. I mean, you are actually able to get non-zero-sum, right? Like, if you can just make it rain and you can't make new clouds, then it would seem on the surface of it like, okay, you're just taking rain from here

and moving it there. It's not actually increasing the net amount of rain that we have in California, for instance. But I know from what we've talked about before that there actually maybe are ways that it's not zero-sum. Do you want to talk Yeah, so there's three ways in which cloud seeding is not zero-sum. So there's one that is definitively and exclusively positive-sum, not like semantically positive-sum.

The first is if you have like the marine cloud layer or the water vapor in California that blows from the coasts to the interior and then recedes back in the evenings or burns off in the evenings, that water is not being deposited or it's not precipitating over most of the coast. And so what we can do is just seed those clouds that would not otherwise, that would instead just go back over the ocean and be recycled into the ocean to do exclusively positive

sum seeding. So there's hundreds of millions of acre feet of water vapor from the coasts that are not precipitating in California. And that is extremely important both for California and everybody else in the Western United States because California draws

water from the Colorado River, right? If we produce more water in California, and exclusively at positive sum manners, then we reduce the extent to which California is dependent on the Colorado River watershed, meaning that Colorado, Utah, Nevada, Arizona, and New Mexico get more of their rights to the water, to the river back. And so that's one positive sum means for producing water. The next is, a

net positive some sense of producing more precipitation. And so what this would look like is, say you have some cloud that's bound to precipitate over, say, Sacramento. It's an easterly. Sometimes this occurs, not always, but sometimes. It's bound to precipitate over Sacramento or, say, it's bound to precipitate over the lower Central Valley at the bottom of a watershed, right? What you can do is you can instead induce precipitation in those clouds at the top of a watershed, and subsequently get

more turns on capital, so to speak, of that water. Meaning, if you induce the precipitation at the top of the watershed, everybody there immediately benefits from it. But then the runoff and the percolation of that water back into the aquifers and streams is usable from everybody downstream of you, rather than just immediately running off into the ocean. And so that's one reason why increased precipitation at the top of the Colorado River watershed would be huge for everybody beneath them. So

that's one consideration. And there's actually some incidents of states in the lower basin paying for cloud seeding, again, in very inefficient old methods, paying for cloud seeding at the top of the watershed. So first of all, you can see the marine cloud layer to induce positive sun precipitation. Two, if you Induced precipitation at the top of the watershed everybody can benefit from that because it doesn't run off into the ocean so quickly. And then this is the third and the most

exciting long term prospect for atmospheric river generation and radical positive some precipitation in the American West and so. Atmospheric rivers are essentially a huge stream of water vapor flowing from the western United States over to the east. These occur, you know, dozens of times per year, and the flux of water through them, right, again, as water vapor and as clouds, not literal, like, streams of water in the sky, the flux of water through them is multiples greater than the

mouth of the Nile River. Right? So these are enormous amounts of water that are traversing the country. Some of it makes it all the way to the Atlantic. Some of it's deposited in the Great Lakes. Some of it does precipitate naturally over the western United States. And that's a really important mechanism for precipitation over the whole country. But what you can do is the following. And this needs to be modeled

out much more robustly before we test anything to this effect. But if you have sort of a speckled pattern of clouds over the western U.S., right? And then you induce precipitation in them. You can reduce the pressure sufficiently in that line from the coast to establish a pressure gradient where water, so to speak, flows downhill from the really high-pressure area of the coast and then into the interior

of the country. And so what that would result in is increased convection from the coast to the interior as far as you can establish that pressure differential. And that increased convection would result in more evaporation from the coasts and essentially the establishment of nearly like an Atlantean level of technology where you have multiple times of the Nile River, Mississippi River flowing west to

east in the United States. And so that's something that probably in conjunction with hurricane busting, we're going to try to do in the coming years. Love it, cool. Okay, let's go back to the farms. So we were talking about how your first customers are farms, they're folks that basically just want more rain to fall over their farm. And so the basic value prop for them is they get greater yield, they stop having to basically spend all this money buying the illicit overdraft fees

on their groundwater. Is that the basic model? That's Yeah, so there's two reasons, right? One, rainwater is more valuable in certain contexts than groundwater, just because there's less sediments, for example, and it percolates and it is absorbed better by plants. Now, you still need irrigation for fertigation, right? Like, one of the best ways to distribute fertilizer to your plants is through these distributed irrigation

networks that you already have established that you can just mix the fertilizer into. But in some contexts, rain is more valuable because it is more readily absorbed by crops. So there's that, first of all. Then the second thing

is cheaper water. And so in most cases in California and other states in the West, water is still very cheap because of the old grandfather rights and like the negotiation about pricing that occurs between farmers and irrigation districts. you can pay something between like a dollar and like $10 per acre foot. And

again, an acre foot is 300,000 gallons, a little over that. But it's kind of funny to think about like you buy a jug of water from the grocery store for like three bucks and you're getting hundreds of thousands of times cheaper water if you're an agriculturalist. So being cost competitive with that price point of water isn't something that we're aiming to do anytime soon. But what is really important to consider is when

you get to your allocation, right? So when a farmer has pumped as much water out of the ground as they're permitted to over the course of a year, the price of water spikes from, say, $2 or $10 to hundreds or thousands of dollars per acre foot. So some of our customers are paying $7,000 per acre feet.

Some of them are paying $500 per acre foot. And so by offering them water at a price that is between the going rate of irrigation water below allocation and the penalty rate above allocation, we get a healthy Very cool. And then, so it's, it's farms, it's like ski resorts. That's actually like another type of customer that wants this. And their, their basic incentive is just increase the amount of snow that they have, extend the ski season. Uh, not,

is it make it better skiing during the season? I don't know. Like, I don't know how they would quantify that. Dude, do you ski? I historically have skied, like my wife got hurt really bad skiing, so she's out for the foreseeable future. I haven't been able to convince her to get back after That has put a damper on the ski plans, but. Fair enough. Well, I got a couple dozen stitches last skiing season in my shins. Uh-oh. Got right back out, so highly recommend.

Great sport. It's snowboarding, actually. Yeah, I mean, if you've ever skied before, then you know, like the artificial snow that's produced is much finer, icier. It's not like the fresh pow-pow that's fun to shred on. And so, yeah, more snow throughout the season so that you have either year-long opens, like what we were hoping for at Mammoth last year, or at the very least extended seasons and higher quality snow. That's all something important. Those

are all things that are important. It's very easy to, I think, because of the level of sci-finess that our technology is, think of all the varied applications, right? Like ski resorts, hail suppression, wildfire fighting, hydroelectric watershed supplementation, hurricane busting, logistics facilitation, right? Like preventing storms from interfering with planes

or cargo ships. But Uh, I'm a big, I'm a big proponent of the Tealian advice that you like nail an initial market before you go out to all of these varied places. And so we're strictly for the foreseeable future, sticking to farms that either do not have water or that are paying excess

penalties for their water consumption. And, um, once we, once we blow it out of the park, once we solve agricultural water scarcity, and at the very least California, but hopefully the rest of the Southwest and Mountain West, then we'll go out and start talking to people about these varied Interesting. I don't know how far along you are in thinking about this, but are you going to have an installation fee to put the thing out there, and then a monitoring fee per month, and then a mission fee or

something? Have you thought about how you're actually going to price this on an ongoing basis Yeah, yeah, totally. And so the way that we've constructed everything so far is either an annual or a multi-year servicing agreement, meaning it's either paid monthly or annually. And the way that we price the contracts

is the following. We do a climatology of our customer's farm, right? And we say, okay, super cold liquid water occurs this frequently and in this concentration over your farm in the last 40 years. Here's the trend that we expect. Here are the amount of opportunities for seeding that we anticipate. And we anticipate this much water from each of our operations to be produced. There's a probability distribution of, you know, the 90th percentile year where there's, you know, two feet of rain and the

10th percentile year where there's three inches, just because we're getting started. We're going to begin with the 30th percentile years and say, OK, you know, in a 30th percentile year, you would get this much precipitation from our intervention. And then it's a really simple calculus between how much water we produce times the mean of the penalty rate and irrigation district rate of water. And

we sell them annual water at that price. And No, I mean, you're giving them, effectively, 50% off on what they would otherwise have to get from the groundwater overcharge thing. Precisely. Very cool. And then what about government? So, I mean, I totally understand the advice and I agree with the advice of just crush one thing, make it clear to customers who you are, get those referrals from this first guy for the next folks. I totally get that. But it sounds like governments are

actually spending money on this today. There's a lot of governments, maybe not this specifically, but definitely a lot of water scarcity stuff. So how do you think about that? Do they have tax credits or something that these farmers could take advantage of? Because government seems like Yeah, so Nevada has a $1.6 million bounty right now out for cloud seeding contracts. Wyoming spends for cloud seeding in their watershed improvement or restoration program. Idaho Power

and Electric pays for this. The Sacramento Municipal Utility District pays for this. Colorado and Utah have paid for cloud seeding historically. So there's plenty of instances of state-level intervention. There's some NOAA pilot agreements that fund cloud seeding, again, in these And so long-term, it makes sense for the state to become a purchaser of cloud-seeking services because, one, you know, it's in the interest of the entire state and, like, the entire nation

that we produce more water for our watersheds. To some extent, it's, like, a multi-state endeavor, so it's natural that it would be, like, a federal contracting

program. Um, and I would even say beyond just like governments domestically, um, the pace car for our operation is the CCP. Um, so right now China is far and away the biggest, uh, weather modifier of any country or organization in the world. They have the, uh, a multi-hundred million dollar annual budget for weather modification operations. They employ a small town of people in their weather modification bureau. It's 30,000 plus folks. And they're

trying to do clouds. Yo, it's huge. And so they're trying to do cloud seeding for a variety of things, one of which being agricultural watershed supplementation, so like the Yangtze River, hydroelectric watershed supplementation, again on the Yangtze and other places. And then for terraforming, right? They're trying to reclaim a lot of the Gobi Desert and Inner Mongolia. and then reduce flooding in other areas by inducing precipitation

outside of floodplains and at other places in the watershed. So, you know, the CCP has a huge budget for this annually. They're using retrofitted military drones for seeding. They have extremely innovative LIDAR and radar-based approaches to identifying seeding conditions and validating results. And And so that's a big deal because controlling the weather obviously has implications beyond just agriculture, right? So I'll

just say that. And we at Rainmaker are very, very aware of that and have an American flag hung up in the laboratory as motivation. So there's that consideration. And then there's a couple other countries that are notable. Thailand has a Royal Rainmaking Department, super base. So they try to do hurricane or monsoon mitigation, flood mitigation with that. The UAE and Saudi are actively trying to terraform their deserts by

inducing precipitation over them. Interesting thing about the UAE, a lot of their roads were built under the assumption that there would be like a millimeter of precipitation annually. So there's no grade in the center of the road that leaves water to run off. So when they conduct these experiments, sometimes they end up flooding the city just because they haven't built their infrastructure to account for rain. But they have a

decamillion-dollar annual program. Saudi spends even more than them. And Israel used to do lots of this, but for geopolitical reasons, it was pressure to stop. But yeah, governments around the world are paying for this. It's actually, in some sense, astonishing. In some sense, no surprise at all that the United States is not as actively engaged in weather modification. And I think part of that has to do with, like, the

way in which we've constructed our environmental ethos in the country. For some reason, environmentalism thus far has been joined at the hip to degrowth and to like an anti-human perspective, one wherein we have minimal interventions and reduce the human footprint on the world. And I think that this is something that's shared by Zoomers and increasingly just people that

are bullish on technology in general. Um, like Rainmaker wants to be part of the clique of people that are responsible for terrapunk, solarpunk, um... climate engineering that is pro-human, pro-environment, and not, like, on the side of, say, you know, big oil and crazy industrialists that are actively destroying the environment or something like that.

And certainly not on the side of de-growthers, but people saying, like, we should have a symbiotic relationship with nature where both humanity and ecology thrives because of our intervention. And so hopefully that message is something that resonates with people and results in more Totally. Are there other companies that you think are on that I think that Casey Hammer and Terraform Industries is totally there. I think that Isaiah Taylor and

Valor Atomics, totally there. You know, really care about cheap, free, clean energy. So there's there's those guys. Those are some great companies. You know, I think that Make Sunsets is very interested in like seeing human thriving. So those are those are some examples. And

That's right. Oh no, we can't get into that. Cut it. No, I love it, dude. So this is a huge technical and business puzzle, right? There's so many different pieces. There's lots of stuff that you're doing uniquely. There's lots of stuff that has been done before. So which pieces of that puzzle do you feel are most like easy handled done and which of the pieces where you feel like like okay we really have a chance to really move the needle and like significantly advance

Yeah, sure. So the way that I break it down is essentially the systems integration piece of the product, and then the, like, deep tech, deep science of the alternative nucleation agent development, right?

So the Snowy folks, in an academic context, at the very least, figured out what sort of radar you need, what sort of radar processing you need, what sort of patterns you ought to be seeding in, and what the minimum viable albeit crude, means of dispersing the nucleation agent is to induce precipitation. So, pairing our, you know, existing radar truck with, you know, the radar processing that they used, and with our drone, with the dispersion mechanism, that

is a systems integration problem. That's logistics, that's operations, that's engineering, but

it's something that can totally be done. I will say, interpolating lower resolution radar that's publicly available, like from NEXRAD, NOAA's operating, radar installations, that is going to be like a serious machine learning problem and something that we're super excited about and will take some serious development because having radar that's not on site, but that's just covering the entire continental US, if we can get it down to a sufficiently low resolution, that will

be a huge unlock for the capital intensiveness of any of our operations. So it's largely systems integration on that side with a little bit of R&D into how to do better radar processing. Then there's the alternative ice nucleation agent development. And so that's some

serious chemistry that we need to engage in. That's something that would, if we succeed as we plan to, we would be able to induce glaciogenic precipitation, and we'd be able to glaciogenically seed over tens of millions of more acres every year across the Western United States, and we'd be able to produce tens of millions more acre-feet of water. So that would radically change the extent to which we can modify our watersheds, our

Very cool, and then how would, so let's say that there's a brilliant young engineer out there, a college student, maybe he's about to choose his major, what would you advise? How does someone learn to participate in this kind of thing? What do they have to know in Yeah, yeah, yeah. Well, probably don't go to college is what I would say first. Step one, drop out. Step one,

drop out of college. But what I would say is, Familiarize yourself with cloud dynamics, with thermodynamics, with baseline level chemistry, and then maybe a little bit of radar processing if you're interested in EE. It's a pretty varied space. Any one of these skill sets that I mentioned would be extremely useful and is what we're hiring for at Rainmaker. Cloud seeding as an industry is super nascent, so there's very few people that have synthesized all

that together, but if you want to, by all means. And then after you've established your baseline context in, you know, chemistry, signal processing, and physics, I would say just read the Snowy research paper, read the ASCII research paper, read all of the papers that have been published or the amount of papers that you can possibly read on cloud, on weather modification. because there's a ton of work that's been done between the

70s and now that's super relevant, between the 40s and now even. Start with like Irving Langmuir's paper and get your baseline context from there. And once you've done that or once you've decided to be sufficiently autodidactic, email me at