Rainmaker’s Secret AI Plan for Superintelligent Weather

In this episode of Sourcery, Augustus Doricko https://x.com/ADoricko⁠(Rainmaker) & Alex Levy https://x.com/alevy⁠ (Atmo) break down how they’re reengineering the weather to solve one of humanity’s biggest challenges: freshwater scarcity. Rainmaker delivers water to farms, ecosystems, and watersheds through a modern precipitation enhancement system—combining radar validation, AI modeling, weather-resistant drones, and sustainable cloud seeding. It’s the only immediate, scalable solution for creating abundant freshwater. Atmo powers the system with next-gen AI meteorology, offering forecasts 100x more precise and 45,000x faster than legacy models—used by the U. S. Air Force, sovereign governments, and renewable energy operators. Together, they’re tackling drought, powering reindustrialization, and restoring water as a strategic resource in an era of climate stress and geopolitical competition. Highlights • How cloud seeding actually works (and what myths to ignore) • The U. S.–China race for weather control and engineered rainfall • Why Atmo’s AI systems outperform NOAA’s current forecasting tools • Rainmaker’s role in America’s reindustrialization—from cooling chip fabs to powering hydro dams • Public perception challenges after the Texas floods—and how the narrative is changing Connect with us: 1. Augustus Doricko: https://x.com/ADoricko 2. Alex Levy: https://x.com/alevy 3. Molly O’Shea: ⁠https://x.com/MollySOShea⁠ 4. Sourcery: ⁠https://x.com/sourceryvc⁠ Brought to you by: • Brex https://brex.com/sourcery—The modern finance platform, combining the world’s smartest corporate card with integrated expense management, banking, bill pay, and travel. As a Sourcery Listener you get: 75,000 points after spending $3,000 on Brex card(s), white-glove onboarding, $5,000 in AWS credits, $2,500 in OpenAI credits, & access to $180k+ in SaaS discounts. On top of $500 toward Brex travel, $300 in cashback, plus exclusive perks (like billboards..) visit → https://brex.com/sourcery • Turing turing.com/sourcery—Turing delivers top-tier talent, data, and tools to help AI labs improve model performance—and enables enterprises to turn those models into powerful, production-ready systems. Visit: turing.com/sourcery • Kalshi https://kalshi.com/sourcery—The largest prediction market and the only legal platform in the US where people can trade directly on the outcomes of future events: https://kalshi.com/sourcery • Fourthwall https://fourthwall.com/—The #1 way to sell merch online — Fourthwall is the easiest way to launch a fully branded merch store—used by big brands like MKBHD, Acquired, & even the Smithsonian. 100+ products. No upfront cost. Visit Fourthwall to start today: https://fourthwall.com/ Follow Sourcery for the latest updates! https://www.sourcery.vc/ Chapters: (00:00) Why water is as critical as energy in AI and industrial growth (02:00) What Rainmaker actually does — cloud seeding 2.0 (03:00) Atmo’s AI meteorology: 100x more precise, 45,000x faster (04:40) How cloud seeding works — science behind artificial rain (06:00) Why Rainmaker + Atmo is a game-changing partnership (07:45) Attribution: Proving man-made precipitation with radar & models (09:15) Why the U. S. is falling behind in weather innovation (11:00) China’s lead in drones, materials, and the weather tech race (13:00) Rainmaker’s drone: the only NATO UAV that flies in icing (14:45) Facing public backlash: Texas floods, chemtrails, and trust (16:30) Making cloud seeding safe — how Rainmaker differs from China (19:00) Cloud seeding for reindustrialization: water as industrial input (21:00) Use cases: hydro, chips, data centers, and agriculture (27:00) Atmo as a U. S. export: modern meteorology for the world (34:00) How AI models simulate physics and save lives

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  I think a lot of people have talked about energy and I think energy is a non-negotiable critical input for everything that we're doing. You not just need energy, but you also need water. The TSMC plant in Arizona, they're planning to recycle a lot of the water that they use.

• 0:13Speaker 01

  TSMC says it will take 4.7 million gallons of water daily to run the first Arizona fab, but it'll bring that demand down to 1 million gallons a day by recycling some 65% of that.

• 0:24Speaker 00

  I like to take long showers in California. I like all of the citrus and fruits that we produce in the central valley. I like all of our clean hydroelectric baseload. And so we need lots of water to keep all of those things humming. AI, data centers that need water, reindustrialization, same way that we need energy, same way that we need steel, we need water as a critical input.

• 0:42Speaker 02

  Extreme weather, drought, other environmental crises around the world are something that humanity has a responsibility to address. I'm sorry to say that the US on the civilian side has fallen significantly behind the rest of the world.

• 0:54Speaker 00

  I would say the United States has the advantage on radar. We're closing the gap on drone capabilities, but it's no surprise China's ahead there. They seem to be willing to use most any mean to accomplish their end, which is global dominance in every field and sector. And so a lot of what they're producing, a lot of what they're dispersing is very toxic. And then with respect to cloud seeding and weather modification, like I've said, it's a really old technology. Generally, it actually doesn't work so well.

• 1:18Molly O'Shea?

  The biggest conflicts that you keep on facing is public perception. The Texas floods is one of the things and you were dragged through the media.

• 1:24Speaker 00

  So.

• 1:36Molly O'Shea?

  Augustus, Alex, welcome to Sorcery.

• 1:38Speaker 00

  Molly, thank you. Thanks.

• 1:40Molly O'Shea?

  Well, I know you have some lovely water in front of you. Should we talk about this?

• 1:44Speaker 00

  Big water, guys.

• 1:45Molly O'Shea?

  Big water, guys?

• 1:46Speaker 00

  It's very important to us.

• 1:47Molly O'Shea?

  Why?

• 1:48Speaker 00

  Well, I mean. I like to take long showers in California. I like all of the citrus and fruits that we produce in the Central Valley. I like all of our clean hydroelectric baseload. And so we need lots of water to keep all of those things humming. AI data centers that need water, reindustrialization, same way that we need energy, same way that we need steel, we need water as a critical input. And what Alex and I are doing is helping identify where that water is and get it from the atmosphere above us down to the ground to our reservoirs and aquifers.

• 2:21Molly O'Shea?

  Augustus, I'm sure everyone knows you because you are a social media figure. You are now a news like segment every week. But Alex, we need to talk about you. So could you share more about Atmo and what you're building?

• 2:39Speaker 02

  Totally. Atmo is the first and leading company in AI for weather. So we started in 2020, and we invented the first deep learning neural networks that could predict the weather. So the same way that LLMs can predict the next word or phrase or paragraph of text, and that's turned into this entire world of AI companies, we did the same thing for the atmosphere where we can make, you know, these high dimensional models that predict the next minute, hour, day, week, month of what the atmosphere is going to do. And that's let us make the most accurate and precise forecasts in the world. And we run these every day now for the most important organizations. So we build next generation forecasting systems for the US Air Force, the Navy, and entire sovereign countries.

• 3:22Speaker 02

  So most recently, we just completed an upgrade to the entire Philippines National Weather Service that increased their level of accuracy by as much as 50% and made it far more detailed than what they had before, which was quite important because they just had 22 typhoons in one year.

• 3:38Molly O'Shea?

  Wow. And now you're partnering with Rainmaker.

• 3:41Speaker 02

  That's exactly right. So we're combining AI meteorology and the radar verified and drone based cloud seeding that Rainmaker does to do ultra precise enhancement of brain.

• 3:50Molly O'Shea?

  What are the kinds of, I guess, like, performance increases that you get from your systems?

• 3:56Speaker 02

  Well, this will be the first time that we're actually going out to the field. So we're going to find out experimentally how much more we can get out of it. So I'd say have us back on in a few months and we'll tell you what those results are, which we'll be able to measure in combination with Rainmaker's radars and also model verification from Atmos. So that's partly how they work together. But in terms of the base increases to the Atmos side of the equation, we've seen models that are up to 100 times higher resolution. So to give you a sense, we're based in San Francisco. And if you took the National Weather Service model for San Francisco, it would treat that whole city roughly on the order of nine pixels. So just like nine big kind of blurry zones. We subdivide that into over 10,000 zones. And so that's the kind of precision that we're going to be bringing to target precisely when and where to do cloud seeding with Rainmaker.

• 4:40Speaker 00

  And so the thing that, I think the three things that are of most interest in the alliance and the partnership that we have are the following. So Alex has spoken to targeting, right? Like with our radar, with his models, we can now find where there is supercooled liquid water in cloud. And just as a brief overview for anybody that isn't familiar with cloud seeding yet, it relies on finding small liquid drops in cloud that are too tiny to precipitate. And if they're below 0 degrees Celsius, below 32 degrees Fahrenheit, we can disperse a material into the cloud such that the little water droplets freeze onto it, become big, heavy snowflakes, then either fall as snow or melt back into rain.

• 5:21Speaker 00

  Now, finding where that liquid is in a very turbulent system or in an area where you don't currently have radar coverage, right? Most of the American West, most of the world, does not have meteorological radar coverage. Finding where that is via ultra-precise models radically extends Rainmaker's capability. But more even than that, the two following things I think are of most interest to me. The first of which is scenario modeling, right? So with our radar, with our probes on our drones, Rainmaker could find supercooled liquid water in the past, right? But maybe there's half a gram per cubic meter of liquid water here. Maybe there's a particular...

• 6:03Speaker 00

  Area of advection going on such that if we seed in this location, we actually wouldn't get as much precipitation as we would if we seeded in this location, right? And what Atmo is going to enable Rainmaker to do is not just help us find where the supercool liquid is, but also map out where in the cloud we could get the ideal outcome from our seeding, right? How can we get as much water on the ground as possible? Or how can we place it in the watershed in the most favorable location, right? Like, we want all of the snow to fall on the tops of the mountains so that that runoff and snow melt lasts as long as possible, right? So, one, I think the targeting is going to be radically improved via the Alliance. Two, the scenario modeling to optimize the amount of precipitation that we're getting beyond just what our radar enables us to do.

• 6:48Speaker 00

  That's going to be great. And then, thirdly, Rainmaker exists. Cloud seeding is viable now because of attribution, right? Because for the first time in the century since cloud seeding was invented, we can use radar to measure how much of the precipitation is manmade, right? And that's been done with radar up until this point. If I fly in a zigzag and I only see precipitation in that zigzag flight track, then we know that it's manmade. But beyond that, right? In really complicated systems where there's already lots of reflective stuff, like lots of ice, and the radar evidence is strong but can be corroborated further, and people can be more confident about the amount of anthropogenic water we produce via the modeling

• 7:32Speaker 00

  that Atmo provides. That is going to be a great capability as well, just so that our customers are confident that they're getting the amount of water that they're paying for. So those three things: improved targeting, scenario modeling to maximize yield, and then improved attribution so that our customers know what they're paying for. That is all something that Atmo is enabling Rainmaker to do.

• 7:52Speaker 02

  Right on. Yeah, for real. Yeah, I mean, it all goes to ultimately the types of constituents that we're trying to answer to. You know, this problem worldwide of water scarcity is a sort of live or die issue for political leaders, national leaders of all stripes around the world. And for some people in some parts of the world, it's becoming extremely severe, really threatening the existence of cities, entire states in some cases. And so they want to know that there are practical solutions that let them reverse this downward trend. And to do that, they're willing to pay very significant sums, I think, to see that resolved. But to Augustus's point, we have to prove what we're doing and how we're doing it in order to have that attribution and ultimately get paid to provide the service of restoring watersheds.

• 8:37Molly O'Shea?

  In the global scope of things, how does America rank for their advanced systems for weather?

• 8:45Speaker 00

  I mean, you could speak to modeling first, yeah.

• 8:48Speaker 02

  I'm sorry to say that the US on the civilian side has fallen significantly behind the rest of the world. So on one hand, on the military side, you know, our DoD partners have been deploying some of the most advanced forecasting models with Atmo over now a period of years. I actually just got here from some work on the East Coast, working with some of our Air Force partners. And we've been running with them multi-scale models, some that go down to a specific Air Force base, some a whole region or tier of operations, some are global in nature. And they're using these and utilizing these for all sorts of different applications. So they're very advanced. Unfortunately, if you go on to the National Weather Service today, none of those models

• 9:32Speaker 02

  that you see there are AI-powered. They're powered by the previous generation systems, which had a lot of limitations. These previous generation models required multi-billion dollar supercomputers, but critically, they're giant, giant stacks of old Fortran code that don't tune and train themselves. So when they make a mistake, like missing Hurricane Sandy in 2015, where other models got it, the only way that gets fixed is if somebody goes in there and hand-patches that to fix it the next time through. So it's a pre-machine learning way of doing it. The post-machine learning way constantly compares what you forecasted with your observations to tune and calibrate. That has yet to come to the US National Weather Service. So we're very, very excited about the possibility of bringing over what we're doing for the military to the public side, but that's very urgently needed, especially considering

• 10:16Speaker 02

  a lot of the forecasting surprises the country's had to the general public over the last couple of years.

• 10:20Speaker 00

  And I would say as well, like if... If it weren't for Atmo, the cuts that are currently underway at NOAA, at the National Science Foundation, at the National Weather Service are further degrading our nation's capabilities to do weather prediction. And so I'm grateful that Atmo exists to fill that gap. And then with respect to cloud seeding and weather modification itself, like I've said, it's a really old technology. And so in many countries, there's these rudimentary forms of cloud seeding where essentially you take a plane up, you dump a bag of salt out of the side of the aircraft, you hope that it rains more, right? Generally, it actually doesn't work so well. However, Rainmaker's primary competitor, it is not an incumbent,

• 11:05Speaker 00

  right? There isn't like another cloud seeding company out there that's trying to end scarcity and water scarcity. It's the Beijing Weather Modification Office through the Chinese Meteorological Administration. And what I would say is, it seems to me as though across three critical components of the tech stack, radar-based validation, UAV-based seeding itself, and then the materials used, the United States seems to be marginally ahead on radar capabilities. We seem to be better at identifying where there is super cold liquid water in cloud, where there is icing cloud, the types of hydrometeors that exist there. But on UAVs, we're probably at parity or just behind China, where Rainmaker has a pretty cool drone,

• 11:48Speaker 00

  the Elijah. It's the only drone in all of NATO that can fly in severe icing conditions. I'm very pleased with our team and impressed with what we've built there in just about a year's time since preliminary design review. That's all great, but China is retrofitting the equivalent of their MQ-9 Reaper. So like our most advanced military drones, the equivalent of that in China, the Wing Long 2, they're using for cloud seeding missions. These are very autonomous, very long-range, very high endurance, high-altitude vehicles with large payloads. That's serious business. Rainmaker is interested in reaching parity with them. Then with respect to materials, Rainmaker uses silvery iodide now. Silvery iodide, that's

• 12:32Speaker 00

  the material that we disperse. It's crystalline structure, is almost identical to ice, so water freezes onto it, right? Now, in principle, there could be materials that have water favorably freeze onto them even more so than silver iodide, or that attract droplets to them and condense vapor onto them even better than, say, salt. And the Chinese Meteorological Administration also through these sort of proxies that exist in the Middle East have developed much, much more sophisticated engineered nanoparticles than any United States operation has. And so I would say the United States has the advantage on radar. We're closing the gap on drone capabilities, but it's no surprise China's ahead there. Their material science

• 13:14Speaker 00

  for cloud seeding is significantly more advanced, but we're working on now at Rainmaker some alternative nucleation agents that, one, are proving to be more effective, and we'll start to roll those out in 2026, but also are not, you know, like, China is an entirely unrestrained country. Like, they seem to be willing to do, they seem to be willing to use most any mean to accomplish their end, which is global dominance in every field and sector. And so a lot of what they're producing, a lot of what they're dispersing is very toxic, both to the environment and agriculture, to humans as well. One, because

• 13:55Speaker 00

  I care about our nation. Two, just simply like for PR sake as well. The materials that Rainmaker is developing right now are, the Rainmaker is developing right now are significantly more safe than anything that China has dispersed thus far.

• 14:09Molly O'Shea?

  Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the US. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Brex's designs help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account.

• 14:40Molly O'Shea?

  You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, ServiceTitan, Hims, Anthropic, Flexport, Robinhood, and Plaid, trust and use Brex. Start today at brex.com/sorcery. That's B-R-E-X dot com slash sorcery.

• 15:11Molly O'Shea?

  Because of that, it seems like one of the biggest conflicts that you keep on facing is public perception. I mean, the Texas floods is one of the things, and you were dragged through the media. And you actually, I mean, you did a world-class job, and people commended you for that. Are you worried about public perception taking this down and ultimately helping China win?

• 15:37Speaker 00

  What I've been really relieved by and really grateful for is, over the course of the last two weeks, people rightfully were skeptical. They rightfully had questions and concerns about a very consequential technology, weather modification and cloud seeding. But in going out to the public, transparently talking about our operations, transparently talking about the state of the technology and where it's been in the past, the overwhelming majority of people that we've spoken to, either in public or in the government, have been extraordinarily favorable and open-minded towards us. A lot of people actually came to me in DC this past week, just off the street from Texas and otherwise. And we got to pray together, both for those affected in Texas, and also they were kind enough to do so for us,

• 16:21Speaker 00

  for our reputation's sake and for our safety's sake. I think I mentioned in some other locations and to you earlier that some people that were misinformed about what cloud seeding is, that believed it to be chemtrails, made threats against. Me, my company, etc. We're safe now. We're taken care of. So first of all, I think that the overwhelming majority of Americans have been open-minded and in favor of cloud seeding after hearing us out. This is true of government as well. Some members of the United States government seem to be interested in capitalizing off of a tragic natural disaster for political gain. That's unfortunate. I will always continue to engage in good faith

• 17:04Speaker 00

  with them. I think that any of their constituents can benefit from cloud seeding, and I hope that they do in the future. Now, I will say as well with respect to our announcement, adding AI into the mix with the concerns that exist about-

• 17:20Speaker 02

  That only lowers the concerns of conspiracy theorists, right?

• 17:23Speaker 00

  Uh, well, no, actually, opposite.

• 17:27Molly O'Shea?

  Super intelligent weather.

• 17:29Speaker 00

  Yeah, yeah, yeah.

• 17:30Speaker 02

  You know, but actually, in a way, I think it does answer one of the legitimate concerns of the sort of massive middle that actually thinks about this issue, right? Ultimately, I think most people can see that extreme weather, drought, other environmental crises around the world are something that humanity has a responsibility to address. I think the great majority of people believe that. The question is how we go about addressing them, how we become more responsible stewards of the world, and try to reverse or ameliorate these problems that we have played a role in causing. And so I think when we say, listen, we want to take the best of computer science, mathematics, physics, apply it to this problem,

• 18:15Speaker 02

  you know, as Augustus was saying, to do simulation and planning around it, to have the most beneficial possible effect with the highest level of rigorous proof and validation. That's exactly what people want to see. And so the course that we're charting in science and technology to show that, time and again, get the reps in, in practice, is going to build that muscle that allows people to trust that operations do what they say they're going to do, and ultimately, I think, towards a positive vision for what this all looks like, which is that humanity is not helpless in the face of these types of drought and water scarcity events. We have something we can do, and we can prove that we're doing it well. And I think that appeals to the great majority of people that are out there that are concerned about this issue.

• 18:54Speaker 00

  Yeah, the motive cloud seeding that we do is called glaciogenic, right? It's glaciogenic or orographic wintertime precipitation enhancement. That's a bunch of gobbledygook words, but simply put, what that means is- Like think of glacier, right? Think of freezing. We freeze water drops in cloud into snowflakes. That means that we're actually better at making snow than anything else.

• 19:20Speaker 00

  If it's sufficiently warm beneath the cloud, then that snow can melt back into rain.

• 19:26Speaker 00

  Rainmaker currently can't, just because the snowflakes that we make are too small. We can't make hail. We can't suppress hail either. But a bunch of cloud seeding programs in the past have tried to do not just precipitation enhancement, but either severe weather mitigation or hail mitigation. Those processes are really, really, really complicated, but ostensibly, they're just physics problems, right? They're just physics or chemistry problems. And so insofar as they are physics and chemistry problems, we're bound by the laws of physics, but we can, with the correct inputs, with the correct sensors, modify those weather systems as well.

• 20:08Speaker 00

  Beyond Rainmaker's current capability, but entirely possible and has been attempted in the past, right? With Project Storm Fury, the United States Weather Bureau, in conjunction with the Air Force, was flying out into the Atlantic to attempt to mitigate the severity of hurricanes. They did so back in the, what, 70s, 60s, 70s? The problem then was this. You know, the theory was that you could precipitate water out of the cloud to mitigate flooding, and also because freezing is exothermic. So when you, when you freeze these little drops into snowflakes, you release some heat, you expand the area but reduce the velocity of the winds

• 20:49Speaker 00

  of the storm. It would also do less damage, wind damage. That is something that the United States government has attempted in the past. It is something that perhaps with a prayerful mindset and appropriate government oversight and regulation to ensure it's not just anybody with a biplane trying to do something like this, I think that it's reasonable to believe that the US government would do in the future.

• 21:12Speaker 05

  In today's high-speed business world, staying ahead means using the smartest tools possible, including the powerful capabilities of artificial intelligence. Meet Turing Intelligence. Turing builds customizable AI systems designed to solve your mission-critical challenges, no matter your industry. From expert guidance to tailored projects, Turing helps top companies realize AI that's more capable, more adaptable, and more effective. With Turing, discover how AI can accelerate your business growth. To learn more, visit turing.com/sourcery, spelt S-O-U-R-C-E-R-Y. That's turing.com/sourcery.

• 21:49Molly O'Shea?

  We are at reindustrialize, so you're going to get the reindustrialize question. What is, let's start with Rainmaker, what is Rainmaker's role in reindustrialization?

• 21:59Speaker 00

  You know, I mentioned earlier, I think a lot of people have talked about energy, and I think energy is a non-negotiable critical input for everything that we're doing, right? Whether it's advanced manufacturing or just simple steel production, right? More, more base components to the economy, whether it is for data centers or for agricultural output. You not just need energy, but you also need water, right? And so what Rainmaker is interested in doing first and foremost is providing as critical an input as energy or minerals, water, to all sectors of the economy. And so the reason why we sell primarily business to government right now is because it's not just one individual farm that benefits from more precipitation over that farm, right?

• 22:44Speaker 00

  It's not just the hydroelectric utility that has a dam on the river that supplies the town. It is everybody in the economy and in the vicinity of that water that benefits. And when I say vicinity, I mean like more snow in Colorado, more snow in Utah. It doesn't just benefit Utah and Colorado. That benefits everybody downstream in the Colorado River, right? So Los Angeles only exists because of water supply from the Colorado. New Mexico, Arizona, Nevada, like I said, Utah. They all need more precipitation in the Upper Rockies. And then also the states in the vicinity of the states that we're seeding in

• 23:29Speaker 00

  that aren't even in the Colorado Basin benefit when there's more snow in the Colorado Basin. So for example, the Platte River, right? Which flows from Colorado into Nebraska. That is pretty radically depleted because Colorado and its constituents and its people and its industry needs water from the Upper Platte. But if you can produce more water in Colorado, then you can actually offset demand from folks in Colorado and industry in Colorado such that people in Nevada or in Nebraska, excuse me, get more. So I think that there's almost this troublesome angle in Rainmaker's sales cycle, which is like, well, who do we sell to? Everybody is a buyer of water. Every sector needs more water.

• 24:12Speaker 00

  And so this, this critical input, this base layer is what we're interested in providing. That's, that's the thing that we can do to facilitate every other sector of the economy, particularly manufacturing, growing and reindustrialization at large.

• 24:27Molly O'Shea?

  I think to make it a little bit more tangible, could you just share the use cases?

• 24:31Speaker 00

  Yeah, 100%, 100%. So you mean for what, for water essentially? So we can say this, cloud seeding produces more water. That's, that's what our bag is. So then who is using that water? So for one,

• 24:47Speaker 00

  As long as we're using hydroelectric, which until either nuclear or solar come online in significantly greater scale than they already have, like clean base power is something that we need more of. And so hydroelectric, more snow on mountains that melts and runs off into those dams that then power factories, that is like one immediate, very tangible use case. If you are cooling a data center, some of that is done just with air, some of that is done with like helium experimentally, but you need water, right, to cool these data centers. If you're trying to make chips, the TSMC plant in Arizona, they're recycling a lot of the water, they're planning to recycle a lot of the water that they use.

• 25:30Speaker 00

  But at the very highest end, you're going to see like a 90% recycling rate, and these are enormous fabs. And so you need to produce more water, not just so that we can create the new chip fab, but also so that in building that chip fab, like people still have water flowing into their residences and can take showers, right? Because one complicated component of reindustrialization is finite resource allocation, right? Like, we want more factories in America, 100%. But we don't necessarily want to sacrifice other sectors if we don't need to, in service of reindustrialization. So I would prefer, one, that we produce the

• 26:12Speaker 00

  requisite water for advanced manufacturing, for stable baseload power, but we do so while maintaining our agricultural production, right? Because, because one option that we could pursue theoretically, that the nation could pursue, is building a bunch of fabs, building a bunch of steel plants, building a bunch of different kinds of manufacturing facilities, and then depriving all of the new residential growth in the American West, or agricultural production in the American West. And so, very tangibly speaking, it's that power, it's that manufacturing, and it's that offset of new manufacturing demand to facilitate continued growth for urban centers, residential communities, and farms.

• 26:53Molly O'Shea?

  Wow. It's a lot. It's a big task.

• 26:57Speaker 00

  I wouldn't have it any other way.

• 26:59Molly O'Shea?

  Alex, same question to you. How does Atmo fit into reindustrialization?

• 27:05Speaker 02

  So Atmo supports the reindustrialization of America, I think, in three ways. First, we're a fundamental service provider to numerous reindustrializers. I'll give you an example of that in a second. The second, I think, advanced meteorology is a critical US export good that does us huge benefit around the world. I'll give you some examples of that. And third and finally, advanced forecasting protects the kind of infrastructure that's getting built now that we need to protect from extreme weather and adverse events. So let me go through those. So in the first case, Atmo is used by numerous constituents to predict the weather in conjunction with what they're doing on an industrial scale.

• 27:49Speaker 02

  So for example, Atmo provides AI meteorology models to all the space launch sites in the United States. So the pace of launch has increased dramatically in the US. This is a huge victory for reindustrialization. And today, the US is by far the majority of volume of space launches. But we're predicting every few minutes of the day extremely precise forecasts for those sites across America. And that allows the increase in that launch cadence because we have to find and thread those windows to launch those rockets. So that's one example. Another is we're doing these ultra-precise weather forecasts for people running increasingly large drone fleets. Rainmaker is a preeminent example of that. And there is much greater drone utilization across a variety of industries and military applications.

• 28:34Speaker 02

  And so again, the ability to precisely predict the weather is a critical input because these drones have certain flight envelopes in which they can and can't operate in. And the range will vary dramatically based on prevailing winds, topography, all sorts of other conditions. So again, you need it for things like that. The growth of the energy industry, particularly renewable energy, is directly tied to weather. So for example, Atmo now operates the most advanced wind forecasting model in Texas. Now, today, we're doing it primarily for hedge funds. And as we did that work over the last year, we learned, very surprisingly, that there are days where the wind power production in Texas is as much as 45% of daily output across Texas. You don't necessarily think of Texas as being a wind state, but for actually a great number of days of the year, it's their single largest energy source.

• 29:19Speaker 02

  And so, again, the ability to predict what's the variable load here is critical to providing sustainable base power and dealing with the peaks and valleys of the power cycle. So, to with solar power. You know, like, one of the first models we ever did was to predict the position of every cloud over solar panels in Germany, so that we could know how much power each was going to make every day. So, in these cases, AI meteorology is this fundamental input to power, to rocket launches, to drone operations. So, we sell that. I think the second thing relates to being an export good. So, one of the grand themes that I think everybody here is trying to fight is the idea that America is getting replaced with other goods and service providers around the world. So, you know, we talked about exports from China. This is probably the most significant example, where there are goods that used to be, let's say, made here in Detroit, Michigan, where we are,

• 30:04Speaker 02

  that are getting supplanted by overseas manufacturers that are doing the same thing.

• 30:10Speaker 02

  Meteorology is an export good. So, in our case, we've helped a number of countries. Philippines is a recent one. Tuvalu is another one that's coming up very soon that we're doing with the United Nations, where Atmo is, in essence, exporting American-made meteorology to foreign countries in order for them to operate their national weather services. This is a huge victory to be able to have a made-in-America technology doing something so critical for a country like the Philippines. That's 118 million people, 7,600 islands. They get the most typhoons of any country, and they're partnering with America, not any other country in the world, but us, in order to do that. And so, it's no surprise, Atmo's had huge support from the State Department, the U.S. Ambassadorial Corps. When we go... Will often go together with with either the sitting or former ambassadors to these

• 30:55Speaker 02

  various countries because they're so passionate about being able to have a made in America meteorology technology doing this critical thing abroad. And then finally, you know, as we build up the type of infrastructure that Augustus was making reference to data centers, manufacturing centers, logistics and transportation hubs, and all of the other connective tissue. You know, these are all big physical assets that are subject to extreme weather, natural disasters. And so, you know, if we're going to face these types of disasters from time to time, the earlier you know that you're going to get kicked in the ass and precisely how you're going to get kicked in the ass allows you to take steps towards protection that mitigates the worst of those harms. In other words, allows you to take those steps that protects that property.

• 31:40Speaker 02

  And so, again, advanced meteorology, like what Atmo has been providing, is going to be of critical interest to these large scale infrastructure projects. So those are the three ways that we tend to interact with reindustrializers.

• 31:51Molly O'Shea?

  Wow. AI is advancing so fast. It's so fast. We're approaching AGI. All of this superintelligence is coming out. A company I recently came across was Turing. Turing helped build and deploy next gen AI AGI systems across various industries. So I'm curious, from your standpoint, like, how do you see AGI and this next wave of like, superintelligence come into all of this?

• 32:15Speaker 02

  No, actually, I was just with the CEO of Turing just a couple of weeks ago. I met him for the first time.

• 32:20Molly O'Shea?

  No way.

• 32:21Speaker 02

  So we had a similar.

• 32:21Molly O'Shea?

  I met him in Paris last week. That's how I know about it.

• 32:24Speaker 02

  No kidding. I met him at a bar in San Francisco two weeks ago.

• 32:27Molly O'Shea?

  Different scenarios.

• 32:29Speaker 02

  Okay, so what's the what's the interaction between, you know, AGI and science? So, you know, just to share a bit of personal history, my my area of study has always been AI and the physical sciences. So back in 2011, I was in the computer science department that originated some of the first work in deep learning neural networks. One of the professors, actually, in our department, Geoffrey Hinton, just won the Nobel Prize about six months ago for inventing deep learning. And the first work that I was doing there was AI for medicinal chemistry. So small molecules. Medicinal chemistry drug discovery. And again, we had this insight that you could take neural networks that were doing things like classifying pictures of dogs and cats and muffins and so forth.

• 33:10Speaker 02

  And repurpose them to actually classify the potential properties of hypothetical molecules that had never been made. Sounded super crazy in 2011. Even now, it's just still being really fully commercialized. And similarly, you know, in 2018, 19, I got really interested along with my co founding team in AI for the atmosphere. Again, the idea that you could take something that's doing, let's say, chat prediction or autonomous robotics and then use it to predict the weather. It's not inherently obvious that you can do that. But what we're seeing is we can make these neural networks dream and envision about physical things. So in essence, it's kind of crazy when you think about it. We are making a biologically inspired neural network. We are moving it into silicon. We are then teaching it to think and dream about physical subjects in the real world.

• 33:52Speaker 02

  And then we're going out and following those instructions to have much more powerful interventions in the world. That's kind of a wild chain when you think about it. Sort of AI that's envisioning and thinking about the physical world, just as as our best scientists can. And so I think for so many of these AGI companies and people that work there, many of whom are investors in Atmo. It's one of the most inspiring parts about this move to AGI. I think people have a lot of hesitations and concerns about AGI, the dislocations that it will pose for society. And I think those are very real. But the great hope is that AGI can provide us an amount of leverage in the physical world that we all live in to do these increasingly innovative and heroic things. And that might be in medicine, or it might be in atmospheric science. And so I think this is one of the great motivating hopes of AGI is that we use it

• 34:36Speaker 02

  to resolve these incredible problems that we weren't capable of doing just on our own, but could with AI.

• 34:42Speaker 00

  Well, and one really, really grounded example of that, right, is in weather modeling. And I'm curious, actually, the extent to which or how you guys do think about this. Like, I think about freezing all the time, right? Like, we need to freeze small drops into bigger snowflakes. Well, there's so many different modes of freezing and like physicists will spend their entire lives thinking about one particular niche, right? So there's like immersion freezing where there's a particle in a water drop and then the drop freezes around that. There's deposition freezing where vapor will freeze onto particles. There's contact freezing where a drop will collide and then freeze that way. And in order to do like a particle wise model of just a meter by

• 35:27Speaker 00

  meter by meter of those interactions of that kind of freezing, like you can maybe solve for that empirically across like a handful of drops, but nowhere near the size and scale to do any meaningful weather prediction, let alone most any meaningful physics simulation, just at like that scale, like cubic meter scale. And developing the physics to explain all these interactions, we could have an army of PhDs doing it forever and still not be great at it, but with better AI, with AGI, and I'm a little bit touchy about that. If anything, I'm a slow takeoff guy. I don't even know if you can really get AGI, you can get something

• 36:12Speaker 00

  that resembles it maybe more and more, but it's not ensouled. So it won't quite have a human thing, which is important, blah, blah, blah. That's more of a theological conversation. But something approximating AGI or just better AI as it exists enables crazy, specific explanation of the physics that are relevant for the atmosphere.

• 36:34Speaker 02

  Yeah, that's super true. And kind of an interesting detail is how much faster these systems are. This has been a big surprise. So like in, so using computers to simulate physics is a very long and old application. So in fact, if you take like the first digital computer, the ENIAC, again, developed here, the first application of it was ballistic trajectory calculation. So if you're going to fire artillery where it lands, but the second ever was numerical weather prediction, predicting the weather very, very crudely, but that was the second application ever run on the digital computer. But the thing that's happened is these physics simulations have required larger and larger computers because they calculate in this extremely exhaustive way,

• 37:16Speaker 02

  where they walk through every single procedural step like Augustus was describing. And the good news is it's scientifically intelligible. The bad news is it can run so slow that you can't really solve practical problems. So you remember AlphaFold, this amazing work from DeepMind, in which they solved protein folding. The big innovation there wasn't just that it was accurately predicting proteins, but quickly. So you could simulate like one of these proteins once a year on like a giant supercomputer. That's fine. But the fact that AlphaFold could do it thousands of times faster, that was the big unlock. Similarly, AI meteorology like we're utilizing here is now running, in our case, together about 45,000 times faster than numerical models of equal resolution so far, which allows us to do these things like consider it the thousands of alternative scenarios and plans

• 38:01Speaker 02

  that Rainmaker could go out there and do and evaluate this whole swath of them. So a big surprise has been not only are these AI systems better at detangling complex systems, but in the sort of pseudo-intuitive way that they reason through their neural networks, can actually do it much, much faster. And this has sort of shocked everybody that you can have a solution that's both more accurate and yet much more computationally efficient.

• 38:25Molly O'Shea?

  And that saves lives.

• 38:26Speaker 02

  And that saves lives.

• 38:27Molly O'Shea?

  More accurate, faster data, can push out emergency responses.

• 38:32Speaker 02

  Exactly.

• 38:32Molly O'Shea?

  Cautions, everything. Then we can mitigate more national disasters.

• 38:36Speaker 05

  Love premium merch just as much as we do? That's why Sorcery uses Fourth Wall. Everyone from creators like Marcus Brownlee to podcasts like Acquired to Orgs, like the Smithsonian Institute, are using Fourth Wall. They let you create and sell premium products without having to stress the details. They handle everything from production, shipping, customer support, taxes, even giveaways. When it's time to level up and make gear that people are actually proud to wear, that's when it's time to use Fourth Wall. And that's why we've trusted Fourth Wall for all of our brand wear at Sorcery since day one. Use my link in the description to get free credits for your first order. And for any VCs, DM me on X and I can get all your portfolio companies set up with a free samples credit deal.

• 39:20Molly O'Shea?

  Okay, as we wrap up, we have...

• 39:25Molly O'Shea?

  Two sections, Brex, and then we're gonna go into some fun Kalshi markets. Okay. For Brex, you guys are really sporting these hats. Where did you get them from?

• 39:35Speaker 00

  I love Brex.

• 39:36Molly O'Shea?

  You love Brex?

• 39:37Speaker 00

  I love Brex.

• 39:38Speaker 02

  Hang on, let's take out my Brex card.

• 39:39Molly O'Shea?

  You have a Brex card?

• 39:41Speaker 02

  Not only do I have a Brex card, but I use it every day, including today, when I used it to check into my hotel here at the Re-Industrialized Conference.

• 39:48Molly O'Shea?

  No way.

• 39:49Speaker 02

  I did.

• 39:50Molly O'Shea?

  So you use Brex for everything?

• 39:51Speaker 02

  I do, and it helps me to label and understand my expenses.

• 39:56Molly O'Shea?

  With AI?

• 39:57Speaker 02

  Possibly with AI, in combination with my great staff.

• 40:01Molly O'Shea?

  So Brex is all about spending smarter and moving faster. You guys are clearly doing that, and I think that involves getting on flights constantly. What do you, yeah, really, what would you say is like the best investment you've made on the spending side of things as you build out both of these companies?

• 40:18Speaker 00

  It's probably all of the tooling that we have in our facility now. Just being able to, having bought or leased and being able to quickly prototype different components for drones, for radar. We're not doing the spinning of boards internally at Rainmaker yet, but that, those purchases in capital equipment to prototype quickly and then manufacture things in low batch volumes internally so that we could iterate. Like, that's been pretty non-negotiable important for our speed.

• 40:57Speaker 02

  Not gonna lie, I'm just thinking of all the flights I take. So we're like halfway into the year. I've done over 100,000 miles so far this year, and all on that credit card.

• 41:07Speaker 00

  Wow.

• 41:08Speaker 02

  Over 100 isn't how many? Like 110,000. No big?

• 41:13Molly O'Shea?

  Where's the furthest you've traveled?

• 41:15Speaker 00

  142, by the way.

• 41:17Speaker 02

  Is that you? You're 142?

• 41:21Molly O'Shea?

  We're gonna have to pull a ranking of the most mileage.

• 41:24Speaker 02

  I actually am a little bit surprised because I I'm off to like Asia every month. So.

• 41:29Speaker 00

  I'm off to the other direction.

• 41:32Speaker 02

  Yeah, yeah, yeah.

• 41:33Molly O'Shea?

  Wow.

• 41:34Speaker 02

  Yeah, probably the farthest I fly probably that in a single shot is is probably Southeast Asia. I'm out there about every.

• 41:40Molly O'Shea?

  Weather is global.

• 41:41Speaker 02

  About four to six.

• 41:42Speaker 00

  So true.

• 41:43Speaker 02

  Yeah, it's true. Although we probably both passed through the Middle East quite a bit.

• 41:46Speaker 00

  Yeah, yeah, I think so.

• 41:48Molly O'Shea?

  All right, these are fun Kalshi markets. We're gonna do a little bit of a rapid fire on this. I want to know your answer. This is a pretty hard question. How many inches of rain do you think there will be in New York City this month?

• 42:02Speaker 00

  New York City this month?

• 42:03Molly O'Shea?

  Yeah.

• 42:04Speaker 00

  Will our answer influence prediction markets?

• 42:07Molly O'Shea?

  Probably.

• 42:07Speaker 00

  Kalshi's.

• 42:08Molly O'Shea?

  And there's like 1.2 million dollars behind this market right now.

• 42:12Speaker 02

  Wait, why is Emma not betting in this market on an automated basis? I'm leaning money on the table.

• 42:17Molly O'Shea?

  Actually, they so their weather markets are very underrated and they have a lot of volume behind it. There's actually someone on their team, Shannon. She's really cool. I like her a lot. She was actually like a beast at doing these weather trades and she made a bag. Not to like, you know, I should talk like a lady.

• 42:36Speaker 02

  This is on whose team?

• 42:37Molly O'Shea?

  She made a lot of money. She works on Kalshi's team now. So she was she was like a super user and now she's working for them.

• 42:43Speaker 02

  She was making bank betting before and they brought her in to work at Kalshi.

• 42:46Molly O'Shea?

  Yeah.

• 42:47Speaker 02

  Okay, interesting.

• 42:48Molly O'Shea?

  But yeah, you should partner with them.

• 42:51Speaker 02

  Yeah, I mean, you know.

• 42:53Molly O'Shea?

  I mean, that might be in. I don't know if that's I don't think that's insider trading if it's predictions, but there is definitely some.

• 42:59Speaker 02

  Someone someone calm.

• 43:00Molly O'Shea?

  Someone will look at you for sure.

• 43:02Speaker 02

  All right, I'm gonna say I'm gonna say 6.75.

• 43:05Speaker 00

  I think just in terms of folks around the table, are you are you placing a bet or no?

• 43:10Molly O'Shea?

  I'll put money behind this.

• 43:11Speaker 00

  Well, then you go next.

• 43:13Molly O'Shea?

  No, you go next.

• 43:14Speaker 00

  Well, because the game theoretic optimal would be for me to say like 6.76 now, right? So I'll go with that.

• 43:23Molly O'Shea?

  I was gonna go with 5.7.

• 43:25Speaker 00

  I win then, I think, is is the outcome of this, provided that there's the on the month of July, to be clear.

• 43:32Molly O'Shea?

  July.

• 43:33Speaker 00

  I think that we'd see. Well, where in New York, in the city?

• 43:36Molly O'Shea?

  New York City.

• 43:37Speaker 00

  In in minute. Okay, yeah, I'm gonna go 6.76.

• 43:39Molly O'Shea?

  Okay, the current forecast is for 5.3.

• 43:45Molly O'Shea?

  And the most, the most money behind it is below six inches. There's about 2% that's above six inches.

• 43:55Speaker 02

  Okay. I'm not going to point out the obvious way that we could do it.

• 43:59Speaker 00

  Rainmaker Technology Corporation. Required to notify NOAA 10 days in advance of any weather modification operation.

• 44:07Speaker 02

  And it is July 17th, so we only have till the 27th.

• 44:12Speaker 00

  We're rapidly deployable, but adhere to state and federal regulations in all weather modification ops.

• 44:18Molly O'Shea?

  Lovely. Well, I can see the collaboration already happening. You should definitely team up with Calci on this. Okay, so going to the China angle of this, right now, markets predict that there's a 29% chance that China overtakes the US economy measured by GDP by 2030. What's your take on this?

• 44:39Speaker 02

  Well, in terms of total GDP, it's plausible. In terms of GDP per capita, that's the real issue.

• 44:45Speaker 00

  I actually am not so concerned with GDP per capita either. Like, I think that, I think that GDP is this thing that economists and politicians have largely used to like abdicate real responsibility to the American people for improving our quality of life. And so just in terms of like the amount of dollars flowing around gluttonous and lecherous financial institutions on Wall Street, like, I don't really care. We could probably create a bunch more fancy fake derivatives that are representative of no value and then radically exceed their GDP. So, one, I'll just put a pin in that and say, you know, like, I'm not a big proponent of the current configuration of the US economy, but I think that

• 45:29Speaker 00

  there's probably like a 0% chance in that outcome, because you can just decide for that not to be so. Like, if it were up to me alone, then I would ensure that that outcome did not come to pass, because I think that's like what an American's responsibility is, but I'm not alone. Like, I think that Alex's making has made and will continue to make inordinate amounts of value for the American economy, and not just in terms of GDP and money flowing around, but in terms of quality of life improvements for our people. And I think the rest of the folks that reindustrialize will do the same. So I would rate that as a, I mean, that's free money. I think that's free money. I do not think that there is any chance in hell that the CCP will do that.

• 46:12Molly O'Shea?

  Okay, free money.

• 46:15Molly O'Shea?

  Last question. Will the UFC host a fight at the White House by the US 250th anniversary?

• 46:24Speaker 02

  By or on?

• 46:25Molly O'Shea?

  By.

• 46:27Speaker 00

  Inclusive of the 4th of July?

• 46:31Molly O'Shea?

  If a UFC fight is held on the White House grounds before July 5th, 2026, the market will resolve to yes.

• 46:39Speaker 00

  I think that's also free money. I think that we live in clown world and there will be a UFC fight there. I agree. Yeah. Yeah, for sure. Didn't Trump post a picture of like an AI image of a UFC octagon on?

• 46:54Molly O'Shea?

  Did he really post it?

• 46:56Speaker 00

  Well, sure.

• 46:58Speaker 02

  Could we add one as well to the betting market? Which is the date on which American Coca-Cola will be converted to real cane sugar from high fructose corn syrup?

• 47:09Molly O'Shea?

  Ooh, I thought you were gonna say add cocaine, but that's, okay, that's probably a better route.

• 47:14Speaker 02

  That'll be for a future administration.

• 47:19Molly O'Shea?

  Okay, guys, this was so much fun. Alex, Augustus, thank you so much for taking the time.

• 47:24Speaker 02

  Molly, thank you.

• 47:25Molly O'Shea?

  And thank you for helping re-industrialize America.

• 47:28Speaker 02

  Right on.

• 47:29Speaker 00

  It's always a pleasure.

• 47:30Molly O'Shea?

  Hey, it's Molly. If you enjoy our interviews, check out our newsletter, sorcery.vc, where we deliver a once a week top deals and tech headlines email and also go deeper on our podcast interviews. Subscribe to Sorcery today and don't forget to subscribe to the podcast on YouTube, Spotify, Apple, or wherever you listen. Link in description to sign up.