The company building underwater drones for $50,000
John CollisonChris SaccaWill O'Brien
Ulysses’ base Mako can sell for as little as $50,000 to navies, versus roughly $5 million for the common U.S. Navy autonomous underwater vehicle; the work it targets is also often supported by ships costing a few hundred thousand dollars a day. Will O’Brien’s thesis is that lower-cost, persistent robotics can move ocean operations from occasional expeditions to infrastructure, unlocking new conservation, energy, and defense markets. “We need to be building millions of underwater drones.”
The system pairs a 2,000-nautical-mile autonomous surface craft, Leviathan, with modular Mako underwater vehicles. Leviathan transports, recovers, wirelessly charges, and connects the drones to Starlink; Mako’s modular battery packs, propulsion, stereo cameras, sonar, and robotic-manipulation capabilities adapt one platform to multiple jobs. O’Brien calls it dropping “the kids off to school.”
Underwater communications are the central technical constraint, forcing intelligence onto the vehicle. Acoustic links deliver only about 10 kilobytes per second over five kilometers, so Mako carries roughly 100 times more compute than other autonomous underwater vehicles, including a full NVIDIA 5090 and Intel CPU. Missions currently use relatively rudimentary local autonomy, such as generating a mapping route and avoiding rocks, but the architecture anticipates more complex behavior.
Manufacturing capacity, not customer demand, is now Ulysses’ limiting factor. The company is declining contracts, plans a factory next year that could rank among the world’s largest underwater-drone facilities by unit count, and hopes to produce several hundred vehicles—maybe 1,000—while one supply-chain supplier makes only 100–200 of a component annually. O’Brien is progressively bringing sonar and inertial navigation in-house because existing maritime supply chains cannot support his intended scale, lead time, and cost.
Seagrass restoration supplied the initial wedge and the reusable technical stack. Seagrass covers 2% of the seafloor yet holds 20% of ocean carbon, and about 20% of commercially important fish stocks rely on it. O’Brien links declining fish stocks to overfishing and habitat loss, while restoration occurs 100–1,000 times more slowly than ecosystems are lost. Restoration entails mapping habitat, modeling suitable sites, collecting seeds, and depositing them; Ulysses can then reuse the sensing, navigation, autonomy, and surface-vehicle platform elsewhere.
Early commercial traction spans mine countermeasures and offshore-wind maintenance. Ulysses won its first U.S. Navy and offshore-wind contracts in its second year; for floating wind, its robots can inspect power cables daily, assess structural integrity, detect discarded fishing lines, and eventually cut entangling lines away. The intended autonomous alternative would reduce dependence on a crewed, weather-constrained ship operation and support more continuous monitoring.
John Collison challenges Ulysses’ dependence on slow government and energy buyers, but O’Brien and Chris Sacca see procurement friction becoming a moat after award. Clearances, master service agreements, and contracting vehicles delay revenue, yet replacement is equally slow, making successful deployments “very sticky” and, in Sacca’s words, “very reliable and very bankable.” The larger strategic opportunity is persistent monitoring of fragile undersea cables and critical waterways rather than today’s episodic patrols.
1. Guinness gave O’Brien a model for industrial innovation
O’Brien’s favorite example is Guinness’ in-house R&D operation, which began as early as the 1890s, hired smart people, and effectively told them, “Come in here and find a problem to work on.” John Collison’s translation: Guinness had its own version of Xerox PARC.
William Sealy Gosset developed small-sample statistics because Guinness needed to estimate hop quality without testing hundreds of samples and losing a few percentage points on every batch. He published under the pseudonym “Student,” O’Brien says, because Guinness did not want Anheuser-Busch or Heineken learning about its research advantage.
Project ACORN—originally Project Dynamite until shipping parcels bearing that label between London and Dublin proved awkward—created the canned-Guinness widget as drinkers shifted toward consuming at home. O’Brien’s point is that a familiar pint conceals decades of technical, manufacturing, and marketing invention.
O’Brien also admires Guinness’ bar-quality system, the 7 Cs: crafted presentation, consistent freshness, counter visibility, correct gas, correct temperature, clean lines, and clean glassware.
That appreciation came from an entrepreneurial Irish family where “business was the talk of the table.” Guinness showed him that consequential technology need not look like software or arrive from Silicon Valley; a homegrown, high-volume business could treat research, process control, and brand as one operating system.
2. Seagrass was the wedge into a much larger ocean platform
Ulysses began with the observation that autonomous systems would perform an increasing share of critical work. After comparing air, land, and sea—and then surface with undersea—the founders saw underwater jobs that were either not being performed, not performed at scale, or too expensive to attempt.
The first use case emerged when a cofounder met a marine biologist manually taking photographs, collecting water samples, and planting seeds. Having worked on aerial delivery drones at Manna, he asked why an underwater equivalent could not map ecosystems, collect seagrass seeds, and restore damaged habitat.
O’Brien calls seagrass “your rapper’s favorite rapper,” overshadowed by its “very good-looking friend,” coral. As stated in the episode, it covers 2% of the seafloor but holds 20% of ocean carbon, and roughly 20% of commercially important fish stocks depend on it. O’Brien attributes declining fish stocks to overfishing and habitat loss, while ecosystems are being restored 100–1,000 times more slowly than they are lost.
Restoration starts by mapping existing beds, then applying a habitat-suitability model using depth, water clarity, and other conditions. Ulysses identifies flowering plants, collects their seeds, and deposits them in suitable areas—a technical workflow it can now repeat without substantial new R&D, leaving field operations as the main scaling constraint.
3. Lowering access costs could turn the ocean into an economic frontier
O’Brien describes himself as a “pro-growth environmentalist” and “eco-pragmatist”: a flourishing future requires both a consistently growing economy and a biosphere capable of supporting it. His childhood beside the sea supplied the emotional pull; the lack of knowledge about most of the ocean supplied the frontier. He says more of Mars’ surface has been mapped than the depths of the ocean, that more than 75% of Earth is effectively unseen, and that fewer than 100 people think about this work daily.
His framework begins with expensive expeditions justified by one high-value use case: pelts in the American West or defense satellites in space. Railroads and reusable launch then lower access costs, increase persistence, and enable uses that were impossible to justify during the expedition phase.
Collison makes the SpaceX analogy explicit: cheaper access to orbit enabled businesses such as Starlink. O’Brien’s ocean equivalent is persistent robotic presence—systems that remain available, gather information, and perform work routinely instead of requiring a fresh ship campaign for every task.
Conservation still encounters “green tape.” O’Brien cites a potentially world-leading Australian seagrass-restoration project in an endangered location that Ulysses has pursued for a long time, only to find permitting the “single biggest barrier.”
4. Leviathan and Mako separate efficient transit from underwater work
Full XYZ access requires two machines because surface transit has less drag. The 2,000-nautical-mile Leviathan carries Mako to the worksite, launches it, calls it home after the mission, and recovers it—the surface craft “drops the kids off to school.”
Existing underwater vehicles were designed primarily to gather data, not physically manipulate their environment. Mako therefore uses a general-purpose, Lego-like architecture: add battery packs for another 150 nautical miles, change the rear-thrust module for speed, or fit stereo cameras, sonar, and task-specific implements.
Those early architectural decisions let Ulysses respond when the U.S. Navy or offshore-energy companies arrived with differently shaped problems. The common platform is the vehicle, sensing stack, navigation, autonomy, and surface support; customer-specific capability comes through modules rather than a new robot.
5. Ocean physics makes local autonomy and surface support mandatory
Corrosion can be mitigated through coatings and anodizing, while pressure defines the operating envelope. Ulysses currently operates to roughly 1,500 meters and has been crush-tested to 2,200; O’Brien calls 6,000 meters “full ocean depth,” although the deepest ocean reaches about 10,000 meters.
Communications are the harder constraint. Acoustic transmission resembles “a dial-up modem,” delivering about 10 kilobytes per second across five kilometers, so operators cannot continuously stream perception and decisions between the vehicle and shore.
Mako consequently runs the loop locally, carrying around 100 times more onboard compute than other autonomous underwater vehicles, including a full NVIDIA 5090 and Intel CPU. Present autonomy remains “relatively rudimentary”: an operator draws a polygon, the system generates a lawnmower-style route, and the vehicle navigates over or around obstacles.
Leviathan provides the missing infrastructure: wireless charging, Starlink connectivity, and live diagnostics after recovery. O’Brien’s longer-term design adds replacement modules and “vending machine style” repair capability, reducing dependence on a distant technician or port.
6. Scaling demands consumer-style supply chains and vertical integration
About 20 million aerial drones are built annually versus perhaps 1,000–2,000 underwater drones, O’Brien says. To reach millions, Ulysses must build on electric-vehicle, drone, and consumer-electronics supply chains that already operate at volume—not traditional maritime suppliers.
The first vehicle was roughly 80% off-the-shelf; Ulysses is now on generation five and expects to bring much more in-house over the coming year. Software was built almost entirely from scratch because the modular vehicle differs from conventional underwater robots.
Sonar and inertial navigation are next. Ulysses currently buys both, but is hiring an acoustics team and has internal navigation prototypes intended to replace existing suppliers. O’Brien says the initial navigation system is intended to cost a fraction of the expensive, long-lead-time supplier’s system while using a supply chain capable of scaling.
Collison presses that building an inertial-navigation system is unusually difficult. O’Brien declines to reveal the approach—mentioning fiber-optic gyros and MEMS only as possible methods—but says scale and lead time justify the work. One supply-chain supplier currently makes only 100–200 of a component per year.
Ulysses’ immediate problem is “building things,” not selling them: it is turning contracts away and plans a facility next year that could be among the world’s largest underwater-drone factories by unit count. The near-term ambition is several hundred vehicles, perhaps 1,000.
7. Navy and offshore wind extend the shared-use platform
In its second year, Ulysses secured its first U.S. Navy contract for mine countermeasures and its first offshore-wind contract. O’Brien jokes that it was “finding mines and hunting mines before it was cool,” but the signal is that one architecture crossed conservation, defense, and energy.
Floating offshore wind creates exposed power cables that can snag abandoned fishing nets and lines, potentially entangling dolphins and whales. Ulysses is building a surface-underwater system to map those cables daily, inspect structural integrity, detect fishing lines and entanglements, and cut fishing line away.
The incumbent workflow begins with a ship costing a few hundred thousand dollars per day, a tethered remotely operated vehicle, and rotating crews needed for 24-hour operation. Headcount expands into operators, managers, captain, mate, cook, and sometimes a doctor; weather and seasickness further constrain working conditions.
Mako’s base price can be as low as $50,000 when sold to navies, against roughly $5 million for the most common autonomous underwater vehicle the U.S. Navy buys today. Advanced sensing and navigation narrow the gap, but O’Brien still expects about an order-of-magnitude advantage; Leviathan costs less than $1 million and more than perhaps $500,000.
Ulysses generally sells vehicles to navies but provides offshore customers a service. The intended autonomous alternative would remove much of the labor burden, permit operation in worse conditions, and let customers substitute persistent capability for sporadic ship charters.
8. Slow procurement can become a moat around persistent infrastructure
Collison’s pushback is that “bad buyers kill companies”: governments and large energy firms impose clearances, safety reviews, master service agreements, and multiyear procurement cycles, potentially exhausting a startup before revenue arrives.
O’Brien accepts the delays but reverses the conclusion. Venture-backed companies can fund the R&D and waiting period; once a contract is won, they become difficult to displace because customers do not want to repeat a two-year qualification process. Sacca says that once companies reach the contracting phase, the work can become “very reliable and very bankable.”
The cultural motivation is broader than procurement. O’Brien contrasts “make something people want” with a newer hardware-founder impulse to “make something the world needs,” linking physical-world companies, faith, and post-COVID dissatisfaction only tentatively: “perhaps” founders forced deeper into digital life now want tangible impact.
Ulysses’ sharpest strategic case is undersea infrastructure. O’Brien says there are about 200 undersea cables today and identifies Ireland as a potential choke point because cables running from the U.S. to Europe largely cross there. He says cables are typically damaged by accidental anchor draggings, though some incidents are not accidental. Sacca separately warns that cables can also be snooped or tapped.
O’Brien’s prescription is persistent, 24/7 robotic monitoring around cables and critical waterways because manual patrols are prohibitively expensive. Unlike defense companies focused on making vehicles “go boom,” he wants to solve the harder general-purpose robotics problem. His stated endgame is to replace episodic ocean campaigns with permanent robotic presence—and help expand the ocean from a roughly $3 trillion economy into “a $10 trillion-plus place for commerce.”
Full transcript
We build robots to solve some of the most critical tasks in our oceans. We need to be building millions of underwater drones.
Saltwater and the ocean are very hostile environments, so that's harder than people realize when you're building a robot that lives in the ocean.
Will O'Brien is the cofounder and president of Ulysses. The company started doing seagrass restoration and is now expanding into everything from mapping the seabed to keeping ports secure. O'Brien's larger ambition is to make ocean operations so cheap and routine that they support entirely new industries.
Before we get to Ulysses—we have lots to talk about—but, just getting to Ireland's most innovative company, explain the technical innovations to Chris.
1. Guinness Built an Innovation Engine
The Guinness Company is a great example of business. Its growth largely derived from technical innovations. You don't think of it that way; you think of just a pint of Guinness. But the truth is, they were starting an in-house research and development team as early as the 1890s. It was a research and development team where you would just hire extremely smart people and say, “Come in here and find a problem to work on.” Much like—
Wait, they had Xerox PARC in Guinness?
They did, as early as the 1890s.
And it was undirected research?
Yeah, just do what you want. Many technical innovations came from it. For example, anyone who's—I studied math in college, and one of the tests from that is the Student's t-test. This is how you derive either the standard deviation if you know the mean, or the mean if you know the standard deviation, from a small sample of less than 30. This was a problem at Guinness because one of the biggest determinants of the quality of a pint of Guinness is the amount of lipids, or fats, in hops. They were having to take hundreds of hops to see whether this was a high-quality batch, and they were losing a few percent per batch. That means a lot when you're in the high-volume, low-margin food-and-beverage industry.
William Sealy Gosset had to publish under a pseudonym, the Student. He couldn't publish his research that he was doing with Fisher, another famous statistician at the time in the early 1920s. He published in the first Journal of Mathematica, which is now a very famous mathematics journal, this innovation, this breakthrough he'd found of how to do small-sample statistics.
But he had to publish it, and that's why it's still called the Student's t-test to this day: They didn't want Anheuser-Busch, Heineken, or other people to know that they were hiring these researchers to solve problems for them.
Oh, so they didn't want people to associate it with Guinness?
Exactly. They didn't want their competitors to know. So they were getting better margins. There was a project called Project ACORN, standing for Advanced Cans of Rich Nectar, which was basically: How do we make a Guinness-at-home version? It was the '80s; people were starting to drink less and drink more at home. The price of a pint was going from £1 to £1.50. People were up in arms, so they were drinking at home. Guinness asked, “How do we do these things at home?” Initially, the project was called Project Dynamite, but, as you'd imagine, in the 1980s, shipping parcels with the word “Dynamite” on them between London and Dublin would cause some concern.
They'd get confused with all the other parcels.
Yeah. So they stopped labeling them Project Dynamite, renamed it Project ACORN, and developed the—
The widget.
The widget. The widget was voted above the internet in the 1990s' top 100 inventions of the 20th century. Marketing is another place where Guinness was really standout. The Guinness Institute—
The Nigeria thing.
Nigeria, the movie, the film industry in Nigeria.
They produced the James Bond of Nigeria, and that's why people know Guinness is very popular in Nigeria. It's one of the top markets, but it's because they made a movie where—
It's the top market, actually.
Okay, yeah.
It's the number-one market.
Was it part of the movie or something?
Yeah. But the marketing, I think, is really one of the areas where Guinness was most standout. I think “Guinness Is Good for You” is this iconic campaign. And the toucan—there's one up there—the iconic Guinness toucan bird. They've won more Cannes Lion advertising awards than nearly any other company. I have the Guinness quality standards on my phone—
Why do you know all this stuff? I'm just curious.
I'm just very passionate about the—
Did you start with it because you're an entrepreneur and a founder, and you're looking for examples of innovation, and there happened to be one in your backyard?
Or is it that you studied as much in the pubs as I did?
I think it was both. I just thought it was a very underrated company. These are the 7 Cs of Guinness quality standards. This is a PDF that I have printed.
Is it a PDF that they make for bars or something?
Yeah, exactly. They have the 7 Cs. The quality standards are: crafted presentation, consistent freshness, counter visibility, correct gas, correct temperature, clean lines, and clean glassware. Those are the things you need.
This is a good system.
Yeah, that's overall life advice, honestly. Keep clean.
Yeah, exactly.
For me, I come from a family of entrepreneurs. My dad's an entrepreneur, my mum had her own business, both my granddads—you know, one ran a pub, the other one was a farmer and had a few other businesses. It's a very classic Irish story. But I always grew up in a world of commerce and business, and my mum and dad have siblings. None of them—none of the men, anyway, at least, the Murphys or O'Briens—work for anybody else. They're all entrepreneurial and a bit disagreeable.
So I grew up in a household where business was the talk of the table. It was revered. I've always been an appreciator of good businesses. I've always revered technology and companies like Stripe and others. But to see one that's homegrown like this, and that I don't think anybody really talks about, was really cool and exciting.
Speaking of entrepreneurship, we do have to get to Ulysses at some point. By the way, do you need another Guinness? I will note that it's 20 minutes in. I'm just belying Irish stereotypes. You can pour these ones and see how you do.
Absolutely. All right.
Thanks, Chris. Wow, value-add. What a value-add investor.
You guys get started talking about this company. I'm only here for one reason.
What does Ulysses actually do?
We build robots to solve some of the most critical tasks in our oceans. That's the simple one-liner. I think we started the company 3 years ago under this general observation that it was very clear that autonomous systems and robotics were going to do an increasing share of the world's GDP and an increasing share of the critical work.
My cofounders and I were looking at potential domains in which to deploy robotic systems. After a cursory look at air, land, and sea, the sea became very interesting. Looking at the surface and undersea, undersea became very, very interesting. We could quickly identify a lot of very interesting jobs that, frankly, were not being done underwater, were not being done at scale, or were not affordable enough to do. The idea for building underwater robots came from that.
You started with seagrass, right? Maybe describe why you started with seagrass and how you ended up pivoting to the robots.
2. Ulysses Starts With Seagrass
Yes. The first one, when we started looking at tasks that robots could do, was this whole spectrum of ocean-conservation-related tasks. One of my cofounders was on a surf trip. He met a marine biologist, and she described her job. Her job was taking photos underwater, planting seeds, gathering water samples—all these very manual tasks. He'd been building aerial drones for food delivery with Manna and thought, “Surely you could build an underwater equivalent of one of these aerial drones for me to do these sorts of tasks.” That's what led us on.
And what were the tasks?
The task initially was the full spectrum of ecosystem restoration: Map the seabed, understand what's down there, collect the seeds wherever there are seeds—we were looking at flowering plants at the time, seagrasses—and then take them and plant them in areas that need to be restored.
And what caused it?
Really, really quick: I have to say, I used your one-pull method.
I hate to concede that there's no appreciable difference between the pints, but, yeah, nothing sacred anymore.
No, no, it's the ceremony.
This is the one shot.
Yeah.
The one shot.
Okay. With seagrass, why does it need to be restored, period? What's going on in the oceans?
So, a few reasons.
If you take current projections for how fish stocks are going to do in and around the 2040s, many of the critical fish stocks in the ocean could be totally wiped out. Basically, the population curves are not looking good for many fish species. Billions of people rely on fish as their primary source of protein or income every day.
Just overfishing in general?
Overfishing, yes. A big driver as well is habitat loss. You have a lot of factors driving it, and basically, the rate at which we restore these ecosystems is 100 to 1,000 times lower than the rate at which we lose them.
Okay, so it's important to restore the seagrass.
Yeah.
And we've lost a lot of it. And you started with that as part of the need to restore the seagrass, you ended up making underwater robots, which is how you got into the underwater robot space?
Yes, exactly. We started solving the seagrass problem, but very quickly into solving that problem, it was very clear: Wow, there's a lot of other similarly shaped problems in the ocean.
3. Ulysses Builds a Modular Robot
From first principles, if you want to build the ocean robot company, you need a few things. You need full XYZ access in the ocean as a domain. To do that, you need a surface craft because there's less drag on the surface, so that's where you should be doing most of your transit. Then you drop an underwater vehicle to go down.
So we developed this autonomous surface vehicle called Leviathan, with a 2,000-nautical-mile range. That goes and drops the kids off at school, so to speak, to go do what they want. The underwater vehicle is called Mako. It's launched into the water, and it goes and finds the location where it needs to do the job. It scans something or robotically manipulates something, and then Leviathan will say, “Come home.” It goes in, and that's the end of the mission.
The state of underwater vehicles before we started building Mako was primarily focused on gathering data, not doing robotics. The vehicle looks a lot different if you want to build something that can physically manipulate things underwater.
There were a few design decisions we had to make there. We wanted to build one general-purpose form factor that could solve a lot of different problems. Our underwater vehicle, Mako, is made of these Lego-like building blocks. If you want to add, say, an extra 150 nautical miles of endurance and range, you just add extra battery packs because we knew some people would want that.
Other people would want the vehicle to go faster, so you have a different rear-thrust module. Some people want to take images underwater, so we have stereo cameras that can give you 3D imaging. Some people want sonar and bathymetric maps, so we have a sonar front module.
There were a few very deliberate early design decisions that enabled us to adopt new applications when the U.S. Navy comes knocking and says, “Hey, we've got some problems you'd like to work on,” or when offshore energy companies come knocking. Because we made these early design decisions in the architecture, it was very easy for us to adapt.
Can we dig into that for a second? First of all, just for all the geeks, I'd love to hear why seagrass and how it compares as a carbon sequester to everything else out there.
Yeah, seagrass is your rapper's favorite rapper or something like that. It's the most slept-on plant out there. It gets zero attention. Coral is the very good-looking friend that gets all the attention, but seagrass is really who you should be talking to.
Coral gets a lot of the funding, and seagrass is really this unsung hero of the ocean. It covers 2% of the seafloor but holds 20% of the ocean's carbon. The ocean is the world's largest natural carbon sink anyway, so it's very important in that regard.
About 20% of commercially important fish stocks rely on seagrasses as a habitat, so it's a very important species. It doesn't really get that much attention. It grows up and down the coastlines of Ireland, the U.K., here in California, the East Coast, and Australia. It's actually the same species in Ireland and California.
We're working across several locations. We've got projects in Virginia, Southern Florida, California, now in Sausalito, and a few other locations. We're also in the Great Barrier Reef.
While you started the company to do seagrass restoration, you're now mostly focused on the robot. You still haven't left seagrass because you're now working on seagrass projects.
Fundamentally, the tech platform we're building—again, because we made these early design decisions—means that you build the technology once to do the seagrass restoration, and then scaling that business is more of an operational problem. There's no net-new R&D that we have to do to serve new contracts. We've already built the underlying IP.
It's just: Do you have enough operators who can go in the field and operate it? The core platform—the underwater-vehicle sensing stack, navigation stack, autonomy, and surface vehicle—is something that can then go off and work with folks in the energy sector or defense sector.
What do you have to do to restore seagrass?
First of all, you have to understand where to restore. You have to map and understand where there is seagrass and where there isn't. On top of that, you'd run a pretty rudimentary model called a habitat-suitability model to say, based on the depth, the water clarity, and these other features, that you think certain areas would be good for restoration.
Then you identify where the plants are flowering and producing seeds, and you go and collect the seeds. You can see videos of this on the Ulysses Twitter account. We just posted a seed-collection video yesterday. Then you bring the seeds to the locations that need them, which you've identified in the habitat-suitability model, and you deposit them there.
I want to dive into that a little bit, because Lower Carbon Capital, our firm, has been involved with you since before the pre-seed—
The first investor, yeah.
And so we should talk about that later and all the hustle that made that happen. But I can say with confidence that, from the beginning, you've had the vision of how to build a universal thing and make it extensible to all of these applications. Both the interventions and the data you're gathering were literally in the original presentation.
How do you choose seagrass initially? I'm curious: How much of that was mission-driven? How much of that is caring about the climate? How much of it is who you can recruit initially? And how much of it is that that's where the money was for the first contracts?
I think it's probably some combination of all of them, but it was very much a passion thing to begin with. I grew up by the seaside. My mom would put me in a wetsuit and send me down to the beach, and that was my babysitter in the summers.
I'm a very pro-growth environmentalist and an eco-pragmatist, maybe—that's how I describe myself. A beautiful, flourishing future for humanity has an economy that's consistently growing and a biosphere that supports that consistently growing economy. Carbon removal is one of those core technologies that I think enables that, along with clean energy and other things.
4. The Ocean Becomes a Frontier
The ocean as a frontier is a true frontier. We've mapped more of the surface of Mars than we have the depths of our ocean, and when I heard that and meditated on it a little more, I was like, “God, this is crazy.”
There's a part of Earth—over 75% of it—that we have no eyes on and have no real clue what's down there. Fewer than 100 people are thinking about this and working on it on a daily basis. How long was all of Western civilization focused on exploration? Now we're doing space, but we have this whole other frontier here, and you're telling me there's a fraction of people actually working on this day to day?
I was like, “This is crazy. This is insane.” Put the money aside and just think: Humans need to explore, discover, and learn.
Part of your thesis is, if I understand this, just like SpaceX, if you lower the cost to orbit, then there are other things you can do. You can do Starlink if you can get things up into orbit cheaply. Similarly, you believe that you can unlock new technologies if you lower the cost of doing things in the ocean?
Yes. I think there's this very generalizable framework for how frontiers get developed. The American West is a good example of this. You have this exploratory period where people go there and go back, and they typically go there and go back for some kind of very high-value reason, where the cost of going there is justified because you're recovering fur or getting pelts or something like that.
There's a single high-value initial use case.
And then in space, it's for defense reasons. You're putting satellites up there to watch the Russians. So there's this initial expedition period, and then you move from an expedition period to, “Okay, well, what happens if we spend more time here and get more precise?”
What happens when we not only gather information from here or do singular things, but do more things and become more persistent in our settlement there?
So then, in the American West, you have the development of the railroad, and access becomes much easier. You get all this development of places like California. In space, you get reusable launch. In maritime, it's like, okay, if you can actually go and have persistent robotic systems out there doing the important work, then there are a lot of new things that you can do.
If you want to go plant a bunch of seagrass, do you get approval from anyone for that, or is it just—seagrass away?
Unfortunately, I think this is one of the saddest things about the environmental movement: you have this concept of green tape, right? Like red tape and green tape. That is the single biggest barrier. For example, there is a certain very endangered location of seagrasses in Australia where we would love to do a project. We've been trying to do a project there for a long time, and it would probably be the biggest seagrass restoration project in the world, in a very important location, but we're kind of banging our heads against the wall.
How did you end up in the Chinese prison?
This is the topic.
I was in—
I want to get back to the robots.
There was an interesting time in the 2010s when China was on this ascending path or whatever, and my mum was like, “Will, you've got to learn Mandarin.” So I go to Shanghai. I'm about 15 or 16 years of age, and I go there. We end up in some night market, and we were there a week before. We go back a week later, and they falsely accuse us of shoplifting.
I think it was this kind of scam they had. They would bring these white kids or something in, and then—
The scam—who had it?
The people in the—because what would happen is, when they accuse you of the crime—
You have to pay a fine on the spot.
You have to pay—well, you get brought to the police station. And if you remember, this is before instant money transfers, right? Sorry, Marcus. Terrible. Crazy time. So we do Western Union, and that takes—China is 9 hours ahead of Ireland, and it's 10 o'clock at night, so my mum's not awake. You ring her at 3 in the morning, or whatever—or it's 9 in the morning Ireland time, so she's awake. So I'm like, “Mommy, I'm in this prison cell in Shanghai. Can you Western Union me the money and get me out?” She's like, “Okay, all right.”
So they—the value of the phones that we shoplifted—
Of course, nowadays they just write it off as a deepfake AI scam.
I was going to say, now that's an ElevenLabs call. That's an ElevenLabs call for sure. I've had to build a safe word with my family.
Send it to me in iTunes vouchers.
Yeah, yeah, yeah, exactly.
One of my favorite parts of Stripe is Works in Progress, our magazine dedicated to new ideas driving economic and technological progress. Take the latest issue landing on doorsteps right now. It covers how China's different cities compete to grow their local economies, why regulations are holding up ways to eradicate mosquito-borne diseases, how hedge funds make millions predicting the outcomes of drug trials, and about 130 pages of other fascinating stuff you're unlikely to read elsewhere. And we want Works in Progress to be as beautiful to look at and hold as it is interesting to read. And call me biased, but I think it's the most beautiful magazine in print today. To read Works in Progress online or to get the print edition delivered to your door, head to worksinprogress.co.
5. Ulysses Brings Manufacturing In House
As I look back—and I did, I looked back at the original pitch you came to us with—but as I look back at that and think about how, from day 1, you were thinking about all the ways this company could— all the markets it could address and what it could do. At the core of it was making the vehicle.
Mm-hmm.
If you had come to a venture capitalist 10 or 12 years ago, or maybe 6 years ago, and said, “Hey, we are going to verticalize manufacturing. We are actually going to build things in the U.S. We're not going to outsource them to Foxconn or Shenzhen or whatever. And for those components where the supply chain might be dicey, we will literally make them from scratch so that we control everything—the entire vertical stack of our device”—that would get you shown to the parking lot.
So how verticalized are you guys, and where did you start? Where are you going to?
It's this iterative process. The first vehicle has maybe 80% off-the-shelf pieces, and we're on Generation 5 now. Over the course of the next year, we're going to bring a lot of it in-house.
To make an underwater drone, you can break it up into the mech, the elec, and the software. The software is pretty much all from scratch, in-house. We're not really building on top of any other primitives. We've built it up from the ground up because we're building a very different type of vehicle. The modularity requires a lot of interesting things.
Presumably, there are hardware actuation frameworks that you use?
Yeah, there are things, but frankly, we've eschewed a lot of them. There's OpenRAWs, and there are a few different ones. We're eschewing a lot of them and building a lot from the ground up because our vehicle looks a lot different from a lot of other underwater robots.
Then there's the elec. In that, you've got sensors and the internals of the vehicle. If we talk about the sensors and that sort of thing, how do you perceive? You're typically using a sonar system of some description. Again, right now, we're buying our sonar systems off the shelf. But we're hiring and building up our acoustics team now to start building them in-house next year.
Then there's the navigation stack as part of the vehicle. There are inertial navigation systems. Again, right now, we buy them off the shelf, but we've got internal prototypes that we're going to start building in-house ourselves to replace these existing suppliers.
It seems very hard to build an inertial navigation system in-house. Explain how an inertial navigation system works, and then—
There are different ways of building them. You've got fiber-optic gyros, which use— that's one way. You've got MEMS sensors, which are in your iPhone or whatever, that are kind of like accelerometers that you can use as well. I won't go into how we're building ours, but yes: initial prototypes. Our initial one is to put it at a fraction of the price of the very expensive, long-lead-time supplier that we're using right now, and to build on a supply chain that we could scale.
But I'm curious: why build your own inertial navigation system? Is it cost? Is it lead time? Is it that, ultimately, you need to get the device—
Scale. For us, it's scale and lead time. There are about 20 million aerial drones built every year. There are maybe 1,000 to 2,000 underwater drones. For us to really fully manifest our vision, we need to be building millions of underwater drones.
Therefore, we look at supply chains and ask, “Can we build on that supply chain?” Insofar as we can, we try to build as much as possible on top of electric vehicle supply chains, drone supply chains—where you have to buy from somewhere else—or consumer electronics supply chains, where things are made in the order of millions, and not maritime supply chains.
So you're building at small scale with an eye toward a very large scale, and everything needs to have a production line in order to do that?
Yeah. We will open a facility next year that could be one of the largest underwater drone factories in the world by number of units. And I think there's a decent—right now, where our company is, our main problem is building things. We are saying no to contracts. We have more demand than we can supply.
The things that we think about on a daily basis right now are not, “How do we sell more?” It's, “How do we build a factory? And how do we vertically integrate some of these parts of the stack?” For example, there is one part of our supply chain where we got the supplier to tell us how many of these components they make per year right now, and they make 100 to 200 per annum. There's no way that we're hoping to make a few hundred vehicles next year, maybe 1,000. So we can't build on top of that.
We find this all the time, both with our companies and with our own investment thesis. As you start diving in, somebody brings you a cool idea or a cool prototype, and it's cobbled together in the same way that Elon was getting components from Mercedes and shit like that to put the first one together. But then, as you start scaling, you start to notice bottlenecks, whether they're intrinsic to the way you're building something or to rolling it out, et cetera.
With each of those bottlenecks, you now know there's an opportunity to resolve that. It's not unlike what you guys have done with Stripe. We've done that in our own investing thesis, where we start to see grid resilience is a thing. Then you start to see, “Oh shit, you know what there's a shortage of? Transformers.” And so we've got these guys, Giga, who taught themselves how to build transformers in their backyard using YouTube videos. Somehow nobody died. And now they... And then they went to China to see how China manufactures them and brought that back.
I love fucking—and now they make transformers in a way that's not all hand-spun.
We definitely haven't done that either.
And so, I love this path, and it's something that I know your audience will love: We have now started to see companies like this really get to be supply-chain experts and see what's standing in the way of their 100x from today. It really is a religion that came out of SpaceX and Tesla.
My cofounder, literally all he works on is manufacturing. We have 4 cofounders. Colm, all he does is pretty much manufacturing. Anduril is the same. One cofounder is pretty much all on manufacturing. I think it's just the default state.
6. Underwater Robots Fight Physics
Saltwater and the ocean are very hostile environments. What's harder than people realize when you're building a robot that lives in the ocean?
Corrosion and pressure are huge external forces. But corrosion, you can get around it by coating, by anodizing the surface of the metal, which gives it some resistance. Pressure, again, you can mitigate that, but going very deep sea, which would be 6,000 meters, is what we call full ocean depth. The ocean goes to about 10,000 meters in its deepest parts, but full ocean depth, as a technical term, is 6,000 meters. That's very difficult.
We currently operate at up to 1,500 meters. We've been crush-tested to 2,200 meters, but we're typically operating in that 1,500-meter range, which is where most of the important economic activity today happens. In the future, I think much more will happen in the deep sea, particularly in critical minerals. But today, a lot of the economic activity happens there. So pressure is one problem. I think the biggest problem, though, is comms underwater.
I was going to ask that.
Yeah, yeah.
So how do you communicate with the vehicle?
A lot of it has to run locally. You have to do a lot of sensing because you can acoustically—
Because communications are very low-bandwidth?
Yes. You can do like 10 kilobytes per second. Over like five kilometers.
Over HF or what?
Acoustically. Ping—like a dial-up modem. And then just being far away from anything is very difficult. Being far from someone who can repair you or recharge you is very difficult.
That's why the surface vehicle is such an important part of our stack. You pick it up, you put it on deck, and it wirelessly charges. Leviathan wirelessly charges underwater vehicles and then connects us to the internet so that we can get a live diagnostic on the vehicle: what's working, what's not working. If someone needs to come in and do an intervention, in the future the surface craft will actually be able to repair the vehicle and have replacement modules, kind of vending-machine style—
The surface vehicle is autonomous too.
Yeah, and it's connected to Starlink so we can see. Starlink is a huge enabling technology for us. That's one part of it. That's why we go for the surface vehicle.
Another thing is that we have around 100 times more onboard compute than any other autonomous underwater vehicle. We run a full NVIDIA 5090 onboard with an Intel CPU. That's very important to our autonomy stack. Again, if you want to do anything intelligent robotically underwater, you need a lot of compute; you need to be able to sense-make to a large degree.
Okay. Does the Mako, if it's doing some mapping or whatever, get sent down with a fully built plan and program? Is it basically running the full loop of sensing and deciding locally on the GPUs on the device?
Yeah, right now the autonomy is relatively rudimentary. We haven't got into very complex autonomy behaviors just yet, but not many of our tasks require it yet. But yes.
But to the extent that it is autonomous, it is fully locally autonomous.
Yeah, exactly. It will be: draw a polygon, let's map this area. It auto-generates a lawnmower track, and then it'll have some rudimentary behavior: if I see a rock, let me just go above that, or if I see something else, go around it.
And then what applications are you having the most success with the robot right now? What are you selling the robot for?
Right now, there are 2 domains in which we've had a lot of success in the last year. Nature was what we focused on in the first year. Last year, this being the second year of the business, we started to look at other use cases. Last year, we got our first contract with the US Navy, and we got our first contract with an offshore wind developer as well.
The US Navy was—actually, we were doing mine countermeasures, finding mines and hunting mines, before it was cool, before the Strait of Hormuz popped off. So we were working on some work related to that. And then the offshore wind—
I haven't heard “the Strait of Hormuz popped off.” That's pretty sweet. Way to appeal to his younger listeners.
And then the offshore wind contract was where—actually, there's this problem in California where they're trying to build floating offshore wind. One of the objections to floating offshore wind is that you put these power cables in the water; they attract abandoned fishing lines. The fishing lines then catch dolphins and whales and these sorts of things, and that's one of the biggest obstructions to offshore wind in California.
And so, could you build—We're building this sensing system, a surface and underwater vehicle that will map these cables on a daily basis, look at their structural integrity, see if they're in good condition, and also see if lines are getting caught on them. If there is a line caught on them, we're building an implement that will cut the fishing lines off the floating offshore-wind lines.
And so why are the offshore wind turbines getting caught up in lines? What's the problem?
Basically, when you build a floating offshore wind terminal and then you have the cable running up, that cable's in the water.
I see.
And you just have—
It's the power cable that's feeding the power back to it.
Exactly. And then you might just have a fishing net floating through the water that has fallen off some Japanese fishing line. It's debris, and it'll get caught in that.
Can you bring that over?
Excellent. Actually, look at you guys.
So it's so weird. I didn't mean to set you up, but when we talk about craft and beauty—
Are you doing a product placement right here?
This is, man. This is a tequila that we make, actually.
Wow.
And it's the only 100-point tequila in history.
You're drinking with two Irishmen and you didn't bring whiskey. You're in an Irish bar.
It's because—
This is, uh... And you bring your own.
I am an American and we are imperialists.
You bring your own.
This is your tequila?
Yes. And so this is Komos Tequila. Komos with a K.
Do you not think post the TV show Silicon Valley, it's a bit on the nose to make your own tequila?
Yes. Yeah. Tres Comas. No. This is actually literally called Komos. Uh, K-O-M-O-S. But this is—it's the only 100% tequila in history. It's very unorthodox, but it's partly aged in wine barrels. And then it also spends some time aerating, so it's like the smoothest-drinking tequila ever. But it makes no sense to do it this way. Uh, nobody had done it before, but we care about that stuff. And so I want to say cheers, but I... There is a weird tie-in to what you just said about design and beauty. Cheers. That is the smoothest tequila in the world. One of the great cultural things about you and you is you guys both, I think, would identify as heterodox. But what is it about you guys who—
Don't ask where I'm from.
Seem to be—No, no, but I think you're emblematic of the movement right now. What is it about you guys? What do you have in common? What's changed?
I think the marked difference is that the founders look very different, right? I think you alluded to this earlier: they're kind of rah-rah, smoking, go to church on Sunday. They're very—maybe a bit louder—and typically work on hardware rather than software.
No disrespect to Y Combinator, but there's the “make something people want” moniker, or something like that, and I think this group is much more, “make something the world needs.” It's a different motivation. There is seemingly more of a spiritual reason why people start their businesses, I think. That's definitely true for myself, or at least a motivation in it.
I often think of there being head- and heart-type founders, where Jeff Bezos is the canonical head-based founder, or head-oriented founder. The internet was growing at 30% per annum, and I saw that: What would be the best thing to sell on the internet? Books are good because you can move them easily and they don't go off.
Dashboards.
Yeah. Then he starts Amazon. But Steve Jobs is the canonical heart-based founder: the bicycle for the mind, the hippie on acid, all this kind of—
This kind of other way. I think this era of high-tech founders is way more oriented toward the heart-based side, or something like that. And, yeah, there's this interesting spiritual vibe. I think there are potentially interesting reasons for that.
A lot of us are very impacted by something like COVID. A lot of us are typically in our late 20s now, but we were in college or something like that when it happened, and maybe feel a little bit cheated. Or we were forced to be in the digital world, so we'd like to cause impact in the physical world. Perhaps. There are maybe some potentially interesting reasons there.
Or maybe we're just more of a product of the times that we're in, where more people are going back to organized religion. Or we grew up through maybe crazy woke-era stuff or something like that, and we're very colored by that.
Yeah.
I know you want to ask more about the robots, but do you want to talk about faith for a second? That was something you would never talk about 20 years ago in Silicon Valley. But I'm curious about that. That has been a real sea change. How does that—do you think it matters? Where does that come from?
I think it speaks to an underlying change in the world. It's very important to me. I've met a lot of my closest friends here in San Francisco through church. I live with 3 other guys; all of them are Christian, and all of them run tech companies.
So, yeah, it has been how I've found my community here. It's definitely noticeable. All of my friends who are Christian are also building companies in the real world or in some capacity. So, using my small sample, I might have to use Student's t-test to see the reliability of this, but there is definitely a correlation here between guys who believe in God and believe in heaven and believe, "I need to live a good life or else I don't go to heaven," and then build these companies that make it rain, that do autonomous cargo, that do ocean robots. There's definitely a connection.
Making it rain in particular is, you know—
That is literally biblical.
I still have more questions on the robots.
Sorry, let's go.
I keep bringing us back to the robots. If I had a bunch of offshore wind turbines—I don't, but if I did—and I wanted them to be flotsam-and-jetsam-free, what are my options right now? Do I pay someone with a boat to go inspect them? Is this just: you sell cheaper inspections, or—
The broad shape—
How are you selling to the... What's the pitch meeting look like?
The broad shape, again, the beauty of our business is that the way business is done—or maritime operations are done—for the Navy, ocean conservation, and energy folks is pretty much the same. Big, expensive surface ship. And as you know, we're not building ships fast enough these days, so there's a backlog even to get the ship. Once you get the ship, you're chartering it at an order of maybe a few hundred thousand dollars a day. On that ship—
What do they cost to make, by the way? We haven't really talked about the core economics of what you're making versus what they're making.
So, on the underwater vehicle, we can sell them. We primarily only sell them when we're working with navies. With the offshore industry folks, they want to buy services from us, so we're running this as a services business. But if you're working with a navy, we can sell them for as little as $50,000.
By contrast, the most common autonomous underwater vehicle that the US Navy is buying today is somewhere on the order of maybe $5 million. So, yeah—
So you're targeting a two-orders-of-magnitude cost improvement?
Now, that's for the base vehicle. Once you add the leather seats and the other nice things—
Premium subscription.
Exactly. But frankly, when you add advanced sensing and advanced navigation... But even still, then you get to one order.
You have a low-cost platform that you can then add sensors and everything to.
Yeah. So that's the underwater vehicle. And the surface vehicle is less than $1 million, more than maybe half a million, but it depends.
So Chris was flagging earlier: bad buyers kill companies. And I think this is true. If you're selling to governments or dealing with multiyear procurement cycles, fairly frequently, the company can run out of money before the buyer has a chance to finish the procurement process.
It strikes me that you kind of have bad buyers. You have some private companies buying from you and everything like this, but it's long procurement cycles and some selling to governments. So how do you deal with this? Or you can push back on the premise, but the fact is that, ultimately, your time to revenue is longer than for other companies.
So right now, the 2 main groups we're selling to are the Navy and offshore energy. And, yes, both of them have long processes. On the energy side, you have to do all these health and safety things. You have to get onto a master services agreement. You have to ladder your way up to that. Government is kind of the same. You have to get on the contract vehicle, get the clearances, all that stuff.
I think this is actually the shape of business that venture-scale companies are very well suited to go after. We can afford to invest the money in the R&D and wait, because once you win those contracts, they're very sticky, actually. They're very sticky, and you're kind of hard to pull out. They don't want to wait 2 years to put someone else in against you.
I've always been attracted to this. My first startup was kind of the first scooter-sharing startup in the UK and Ireland. That's where I learned to sell to government, and I loved it. I loved it for 2 reasons.
Once you won a contract, it was very difficult to lose them because they're just very sticky customers. But also, you're seeing your technology go into the hands of people who really need it. My underlying thesis around my own entrepreneurial path has been that my first startup was going into transport and trying to change that.
This is now going into the public domain, a public good, working with defense or going into the ocean in general, which is a public good, and trying to build and deploy technology into it. But 60 years ago, the people working or thinking about these problems—most commonly transport and things—were not guys who were here in San Francisco. They were in some government job or something like that.
My underlying premise around technology has always been that public services or public-domain technologies are maybe the most interesting because, if you think the smartest person in 1960 was probably going to work in government, they're now going to work in San Francisco.
Ergo, an increasing share of the things that were previously the reserve of government are now going to be the reserve of private companies. So there is a future where Ulysses is the largest navy in the world. We're their largest steward of the ocean—
The US Navy is pretty big.
Yes.
I bet he has more craft soon. But I'll tell you, as an investor, once these companies actually get to the contracting phase, they're very reliable and very bankable. So it may take some time, but that's what our business is: to bridge that to where it comes through.
Strangely, whether it's as a hard-tech investor or even in some of the decarbonization work we do, one of the most sober and consistent customers for the work we do is the Department of Defense.
They're actually a good buyer?
When you get to the people who actually get shit done at these organizations, they're incredibly well-informed, and the decisions are apolitical and very sober. And so when somebody comes inbound to him... That's the other thing, like when you see the Department of Defense or the Navy coming inbound to a startup, that's a new muscle that they didn't used to have, so they're paying attention.
And when they come in and say, "Look, you solve a problem for us at 1/100th of the cost we pay today," once we get to that contracting process, it may take a while to fill out your master service agreement, your SLAs and shit, but the reality is that's very bankable.
And they're using you for mine-clearing stuff, then, maybe?
Yeah. So I might actually just go back to one of your original questions: how is work done in the ocean today, and how are we different? Because it makes the rest of the things clear and, again, draws a consistent through line.
Most important work today involves paying a few hundred thousand dollars for a ship per day to go out and do something on that ship. Because you're going out there, you typically have a bunch of guys on board. If you want to do something underwater, you put this tethered vehicle in and remotely operate it; you look at the screens, take shifts, and go do the thing underwater.
You find the mine, repair the pipeline, whatever it may be. Again, they all have this shape: plant seagrass. But the problem is, due to union laws and labor laws, people can only work maybe 8 hours a day, and you want to work 24 hours because you're going out there anyway.
So you need 3 people then.
And then each of those people has a manager, and then you have the ship captain, and then you have the second mate, and then you have the guy who needs to cook the spaghetti, and then you have the doctor because you've gone over the 8 people and you left from the U.S. port. So you end up with this massive crew, hundreds of thousands of dollars a day. And that's a broad shape of all the tasks underwater.
Then the idea is, if you can do that autonomously, you don't have any of the labor cost, you have control over Starlink, and you have all these nice—
Can you work in worse sea conditions than—
Yeah, exactly. That's another thing. And you go fast as well. You can go fast through the water. When people are on board, you can't. They get seasick.
Do you compete with the neo-primes? Are you guys doing things for some of these, like Saronic or those kinds of folks?
Saronic doesn't have an underwater vehicle. They're building ships, and they're just going to become a new shipyard.
Got it.
And I think where we're different, maybe, from Anduril or others is that the thing we're thinking about is, again, robots will do an increasing share of the work in the future. Who builds the underwater ones? A lot of the folks in the defense sector are thinking, “How do we make things go boom?” Frankly, that's not an interesting technical problem for us. Any numbnuts can put C4 on a vehicle and make it go boom.
The very difficult, interesting technical question is, how do you build a general-purpose underwater robotic platform? That's the thing that we think unlocks the ocean at scale. That's the thing that turns it from a $3 trillion economy into a $10 trillion-plus place for commerce.
The undersea cables—how worried should we be about—
Very.
Absolutely. Totally asymmetric. Jesus—totally.
Do you guys have any undersea cable angle?
7. Undersea Cables Need Constant Patrols
Yeah, we have one. We have some work in that domain. I can't talk about it too much.
Top secret.
I can't talk about it too much. But yes, there is basically a need to create a—again, what we're trying to advocate is that the maritime domain move away from this episodic, singular-campaign thing, more like we're creating permanent settlements there, like we're trying to do in space.
A part of that would be a persistent defense network around all critical waterways. A very critical waterway is off the west coast of Ireland. All of the cables that run from the U.S. into Europe pretty much cross Ireland, so that's 1 area that you might highlight as a choke point. There are a number of these. There are a number of them in Asia and in the Middle East.
All of these areas should just be persistently monitored 24/7, but it's much too expensive to do that manually. So yes, that's a massive metric.
100 years ago, I used to co-lead the team at Google that built all the data centers, and we also did the transoceanic fiber inside. Data centers are fragile, and fiber is even more fragile. The way you take a data center offline is you bring a pair of fucking garden shears and cut the fiber that goes through.
And there are many fewer cables than people think.
Yeah, there are. And by the way, I found one in Oregon. No one used to publish their fiber maps, and so the way I would find them to figure out where I wanted to site a data center is I would drive around the coast until I saw “Call before you dig” signs.
Once in Oregon, I encountered some “Call before you dig” signs—
They’re like, “Oh, start digging”—
The street was called Five Nines. What kind of—only a fucking geek would name a street Five Nines. I was like, “Yep, I got it now.” Then we'd follow the “Call before you dig” signs.
If you really want to take a data center line offline, until we have Starlink and other diverse opportunities—diverse providers in that space—you literally snip the fiber and it's offline, right?
Yeah.
I say this partly because we should be aware of this—partly because we need to protect it, and partly also for other humans. In the event we do get to a FOOM moment in AI, we might all have to do our part. We might have to all do our part to kill the data centers.
That's funny.
I'm sure I just put myself on the naughty list as a result of saying that.
That's a bad fact to have in the pretraining—
But the reality is, right now, it is the Wild West. When you bring up the Wild West metaphor for what you do, it really is that. And we do need protection of those resources out there.
Yeah, so we have about—I think it's about 200 undersea cables today, and they typically only get taken out by anchor draggings, accidental for the most part, though some are not. But if you did want to do a targeted demolition of many of them, it's not crazy hard.
But I also don't know what—They get snooped, and so I get read in, and they get snooped, and—
As in tapped?
Yeah, like, it's not hard for China to go listen to those things and read everything. It's actually the least fortified. It's hard for them to get into a hardened facility in downtown San Francisco that's running all your traffic.
But on Five Nine Street in Oregon.
Yes. It's very easy for them to hang out and listen. And so we need patrols there for sure.
We're going to have to leave it there.
Good.
Chris, thank you guys.
Yeah, right on, man.
Cheers, guys.
Cheers.