A conversation between Jeannette zu Fürstenberg and EON co-founder Charlie Horowitz.
For decades, orbit was a destination. A place to visit and return from, or else a means to an end for observation, navigation, and communication back on Earth. But what happens when space becomes infrastructure? A new generation of founders are treating orbit as an operating environment in and of itself.
Charlie Horowitz is a part of this next generation and his company, EON—cofounded with Tyler Presser—is one of the clearest examples of this shift. After experience at Palantir and Apex Space, Horowitz turned his attention to the vulnerability of inter-continental bandwidth. Today, 95% of the global internet moves through subsea fiber cables. EON is taking it to orbit.
On the eve of announcing EON’s new round of investment, GC’s President Jeannette zu Fürstenberg called Horowitz to talk about lowering the launch cost curve, how orbit is evolving from a location into an economic environment, and why data centers aren’t the most immediate satcom demand.
The Conversation
Jeannette: Hey, Charlie. How are you?
Charlie: I'm good. You know, nobody tells you this, but you start a company and you get a little busy, so just dealing with a lot.
Jeannette: I keep telling all my founders, you need to understand that this is also part of building—to actually take breaks. And we appreciate you taking the time. So why don't we start with the big picture: What is the biggest misconception that people have about space infrastructure right now?
Charlie Horowitz: I think there's actually quite a few. The two big ones: that launch is getting much cheaper, and is therefore “solved”, and that the decrease in launch costs makes space viable. Because of SpaceX, everybody thinks launch is this completely solved problem and you can go tomorrow and launch something. And that's not really true, unless you work for SpaceX. For the rest of the industry, you're at the whim of when a rocket is available.
JZF: For EON, what kind of launch availability do you need?
CH: We're designing a constellation to fill in where subsea fiber can't reach right now, or to supplement existing capacity—orienting our orbits around where we think the important routes are. We're planning on launching in blocks, where a single launch with a heavy rocket, carrying tens of satellites, gives us 24-hour coverage of these routes. We'll continue launching additional satellites to add capacity and deliver new routes.
That means we need a heavy launch vehicle that can deliver a lot of mass to orbit. We are looking into partnering with a range of launch providers who can deliver. One thing we are particularly mindful of is not being overly reliant on Starship, as it is unclear when or even if it will be commercially viable.
JZF: And on the opposite end of that spectrum, what is the biggest overestimation of space infrastructure that people have?
CH: I think the elephant in the room is space data centers. What's nice for us is that's not the market we're serving. We're serving a market that exists today.
As you know, there is this immediate problem: subsea fiber is really hard to build, even without any geopolitics involved. Meanwhile, there's way more data today than there ever has been. So if you want to move it, good luck doing it through subsea fiber. At the same time, SpaceX is doing awesome stuff for commercial use cases—like this phone call we’re having right now, we could do it on Starlink. But for really high throughput, that's not what they do. They do radio predominantly.
When we started hearing about the subsea fiber problem from customers, from governments, and reading about fiber lines getting cut near conflict zones, space became an obvious alternative.
When we started hearing about the subsea fiber problem from customers, from governments, and reading about fiber lines getting cut near conflict zones, space became an obvious alternative.
JZF: Are you anticipating that states would become customers of yours?
CH: We're in an interesting spot in that there's quite a bit of precedent for how subsea is built. A lot of it is sovereign-funded—these are critical infrastructure plays—so a lot of it is either sovereign-funded or sovereign-subsidized. I think that would probably be the same for us, especially for sovereigns that are poorly located. Certain countries require you, anytime you land data there—even if it's going from that country to everywhere else—to pay taxes. So there's a bunch we can do to actually make it more affordable for smaller sovereigns to have a meaningful stake in their own internet, which is pretty cool. And then for larger sovereigns: if you're worried about conflict in the Pacific, we're a really good alternative. If you're worried about conflict in the Middle East, we're a really good alternative. The Baltics, we're a great alternative. Basically anywhere there's a lot of infrastructure and information that needs to be moved and conflict is kind of following, we're a really nice alternative.
JZF: So what in your view is going to happen to launch? Are we entering a period of experimentation?
CH: It depends on where you want to end up in space. If you're going to low Earth orbit (LEO), it's all about getting as much mass to orbit as possible, so a vehicle like Starship makes the most sense. But if you're headed farther out, like medium Earth orbit (MEO), you probably want a third stage. You'll sacrifice payload mass, but you'll gain the performance needed to get there.
So as the space industry matures, I think higher orbits like MEO will become more interesting. We're early on that. I think others will follow suit.
JZF: Now, let's just assume launch is solved and there's an abundance of new stuff popping up—what will that be? What are the most interesting ideas that are currently bottlenecked by launch availability?
CH: In that case, if launch is totally solved, then data centers actually do make sense—if it's affordable, I should say. The economics actually play. If you're able to do whatever you want in space, the next level is just getting communications to be a lot more robust. Any satellite beyond geostationary orbit (GEO) is really hard to talk to. So if you could get networking nodes into orbit for deep space that would open up a host of possibilities: communicating with other satellites, rockets, launch vehicles to the moon and beyond.
JZF: Almost like its own telco network in space. Would that mean space manufacturing could include autonomous entities?
CH: The world (or cosmos) is your oyster. None of the crazy ideas we're talking about today are actually all that new. People have thought of a lot of this for decades. What's new is that a lot of this is becoming affordable, and a lot of the science has caught up to science fiction.
JZF: When you really have infinite intelligence (compute) to solve for these things, and you square AGI with space, what would that unblock?
CH: It would completely change how simulation is done. If you have AGI that's designing you a perfect rocket from the start, and you can actually model out the flow of propellant through an engine, which is really hard to do correctly, that changes everything. Then launch becomes cheaper because you don't have to spend as much on R&D. Same applies for satellites and payloads.
JZF: Tell me about the breakthrough you had around optical technology.
CH: Lasers have been around for decades, since really the '60s. I think Einstein theorized it, then Bell Labs really perfected it, and a few others in Berkeley in the '80s and '90s for fiber. So they exist. The hard part is getting them through the atmosphere.
Stars twinkle because light passing through the atmosphere hits particles that cause scintillation. Lasers run into that same problem. But there are ways to get around it. That and pointing is where we are spending most of our time to deliver an incredible product.
JZF: Are there cultural differences between Old Space and New Space and how are you and Tyler bridging that gap as you recruit the best talent for your team?
CH: We need both, and we need them at the same time. New space moves fast. Old space is proven and reliable. We don't have the luxury of choosing one.
That tension shapes how we hire. We're building a team that merges two worlds: the speed of the startups now shipping impressive hardware at a rate traditional aerospace hasn’t matched, and the discipline of engineers who've spent careers making sure hardware works the first time, every time, in an environment with no room for a second try.
JZF: In your childhood, how much time did you spend thinking about science fiction and space in general?
CH: Not a lot. I grew up in LA, which now has this amazing space scene—it didn’t when I was younger. As with most sector-based scenes, it tends to die and come back. But when I joined Apex, I saw my first launch, and I thought, I'm always doing this, I'm never not gonna do this. I was probably 24, but I felt like I was 10 again. I wish I had gotten into it earlier—there is so much amazing stuff nearby that I didn't know about. But how can you not get romantic about it? It's the coolest stuff in the world. It's beautiful.
Image Credit: Philip Cheung
%201.svg)
