🫡 It’s your PA team here. Your eyes do not deceive you — this is a real, live issue of Per Aspera, our first since mid-August. When Joy signed us off for a short summer break and programming overhaul, we promised a new-and-improved PA for the fall. That starts today, with some very exciting new additions and editions coming to inboxes in the not-too-distant future.
(To the several hundred of you who joined during the quiet stretch: welcome, this is not the usual cadence, and you picked a good time to show up.)
Leading off today: Ep. 2 of Hard Reflections with a close friend of PA — someone who saw early that there are many kinds of American capital, and has spent a decade standing up institutions for each of them, all pointed at the technologists who can tip the U.S.-China balance in our favor.
Good to be back. Let’s get into it. 👇️
IN THIS WEEK’S EDITION:
🎙️ 'Violence of Action' w/ Jordan Blashek
🏆 Global Space Awards
📊 The datacenter dilemma
📰 YC industrials + space power
🌊 Seaglider Era
Forwarded this? Join Per Aspera today.
EP 002 / This week we sit down with the man, the myth, the legend, Jordan Blashek. Dan first crossed paths with Jordan in Austin, where the two had been booked to interview each other. Dan, eminently impressed with Jordan’s wisdom, came away calling him “a man of 35 going on 85,” which is about as close to a standing ovation as you can get from Dan.
The story of how Jordan got here starts with a Princeton degree and walking away from medical school for the Marines, whose recruiter looked him over and told him to come back when he could do 20 pull-ups. Nine months later, he could. In the Corps, Jordan’s first deployment was counter-piracy duty off the Horn of Africa, where the battle group's daily brief included the position of the Chinese submarine shadowing it… this, on a mission that had nothing to do with China. On his second, training the Afghan army, the local soldiers kept repeating a refrain: you're just here to steal our minerals. He knew that was wrong right up until he looked into it, and found major lithium and rare-earth deposits, with Chinese firms already working them.
Jordan with his interpreter and several Afghan National Army Soldiers (2013)
Jordan left the service convinced America was distracted by the wrong wars. He had a prescient sense that the nation’s preeminent strategic focus would shift back from counterterrorism to great power competition, and that a showdown with China was looming — one that would not be fought kinetically but with technology and markets. This is where he knew he had to be.
Civilian life became a decade-long pursuit to answer the question: who pays for the hard stuff, and is there a better way? Jordan found creative ways to answer this: first, building Eric Schmidt’s family office (with a patient capital ethos before it was cool), then cofounding the first venture fund licensed to take Pentagon money (back when most investors still dismissed deep tech startups as “science projects”), then talking New Mexico’s ~$80B sovereign fund into seeding a $50M venture studio. Now, at Overmatch, with Evan Loomis and Morgan Hitzig, he backs companies he intends to see through to the public markets.
THE INTERVIEW. We’ve wanted to get Jordan “on the record” for over a year, and our wide-ranging interview went twice as long as we initially allotted. We covered it all:
The 9/11 morning that set his course ("my mom woke me up and said, 'we're at war'")
Four years of road trips across a divided America with a close friend from the other political party, and the book that came out of them
Is this whole defense-and-deep-tech venture wave just the new SaaS, destined to overheat and wash out?
The capital gap shifts right: we’re flush from seed through Series B, with a wall at Series C, D, and E, where factories, tooling, and inventory are financed
Hard tech's only two real commercial demand engines: the space economy and the AI buildout
The boring minerals deal everyone else passed on, that went from zero to nine figures of revenue in two years
The moral case for going public early (we agree!) … and more broad-based, compounding public ownership
The four things he looks for in a founder, the Marine’s answer to move-fast-and-break-things… “slow is smooth, smooth is fast,” and the magic of when it all comes together (“When you do have the pieces in place, that violence of action is so important.")
Nothing in this business is clean or cut-and-dried — which is exactly why Jordan makes a perfect second guest for Hard Reflections. Big thanks to you, Jordan, for letting the PA community dig into your worldview, motivations, life lessons, and that 85-year-old’s wisdom. 👇️ Read the full interview online.
Get someone their flowers. The Global Space Awards, judged by a distinguished panel that includes our very own Jeff Crusey, are the annual ceremony honoring the path-breaking players, projects, and programs across the space economy — from executives and engineers to mad scientists and startup founders doing their life's work.
Last year's edition drew 500+ nominations from more than 30 countries; this year is pacing for a similar draw with a ~week left on the clock.
Honorees get their night at London's Roundhouse on December 11, alongside a carefully curated group of a few hundred leaders from across the international space business.
The Global Space Awards are built around the conviction that “a new era demands a new benchmark,” which we so happen to agree with. Let's honor the space economy just like we do mature industries. We don't need any more dog-and-pony shows that lift up the loudest logos or most familiar faces. Rather, we need true recognition of the pioneers shipping world-changing work.
So, if someone on your team, your cap table, or your factory floor did the best work of their career this year, put their name in before the window closes.
Find out more and nominate someone below. The process only takes a few minutes, and the deadline to nominate is next Monday, September 21.
A HOT TAKE: DATACENTERS ARE GOOD FOR AMERICA. We know… this isn’t (or shouldn’t actually be) a hot take, but based solely off of public opinion in these United States of America, it 100% is a hot take today. But before we make our case, let’s be clear on where we stand (as we’ve already done in #016, #019, #022, #025, #034, #039, #044, and #052): Datacenters are now among the most disliked pieces of infrastructure in the country. They poll worse than coal plants did a year ago, and fracking, at its most politically radioactive, was never this unpopular. They poll worse than the IRS, which at 38% favorability is the least popular agency in the alphabet soup, worse than socialism, and even worse than pineapple on pizza. No wonder, then, that opposition to new builds has become the rare cause uniting both political parties at every level below the White House.
The facts, meanwhile, argue in favor of a buildout: Datacenters are currently a huge contributing factor to U.S. GDP growth. They are the entire growth engine of U.S. private construction (with second-order benefits for trades, towns, and industries we care a whole lot about, as documented in #037, #051, and #055). In fact, ex-datacenters, construction on everything else combined sits $100B+ below late-2023 levels.
Looking at one specific qualm, H₂O, American datacenters directly consumed ~17.4B gallons of water in 2023. California’s almonds consume roughly 90 times that (as we pointed out in The End of Thirst Traps). Looking at another qualm, land use — our good friend and collaborator Alan Simon pointed out on X… datacenters’ projected 2028 footprint (~900,000 acres) is less than half what American golf courses occupy:
BUT… Numbers don’t defend themselves, and so far, nobody in tech, industrials, or public office has figured out how to effectively make the public case for datacenters. On the timeline, we often see people in our circles dunking on the “stupidity” of datacenter opposition — which, surprise surprise, has yet to change a single person’s mind, and never will. The hyperscalers are spending, we’d guess, billions on PR for this… and whatever they’re doing, polls suggest it isn’t going very well. The frontier labs are even worse, with their main contribution to date having been musing quixotically and aloud about how many jobs their products will erase. (Though they’ve stopped doing this recently, thank the heavens.)
OPEN THREAD… It’s a tough needle to thread. Being able to sell this — to make the case for the buildout to the country — is an existentially valuable national skill right now, and nobody seems to have figured it out quite yet.
ATOMS V. AGENTS @ DEMO DAY. Every few months, Y Combinator runs startupland's closest thing to a rite of passage, in which a couple hundred startup founders, given three minutes apiece, take the stage to deliver a polished, nervous, and (hopefully) well-practiced pitch on the dent they intend to make in the universe. Every vintage has its meta: the 2010s batches minted consumer marketplaces (Airbnb, DoorDash, Instacart) and fintech juggernauts (Stripe, Coinbase, Brex); by 2020 B2B SaaS was especially en vogue; and the last few years have belonged to AI/agents. Hard tech was always an outlier, though as plenty of YC alumni … including Boom, Relativity, Astranis, Cruise, and (disclosure with a grin) Array Labs … would tell you, the “YC of hard tech is YC.” Still, going purely off statistics, YC has largely been a software factory. Which is why we were very excited to meet the latest batch last week. There were 29 robotics companies, and more defense startups (8) than in all four batches last year combined. True to form, we saw some big swings, from floating nuclear-powered datacenters (Atomarine) to computer chips built from living neurons (Frontier). It would behoove us to mention that one such Overton Window-busting “big swing” from the Fall ‘24 cohort, Starcloud, became one of the fastest YC companies to ever hit unicorn status.
THE RECEIPTS. Longtime readers of Per Aspera will have seen this vibe shift coming. Last summer (#005) we clocked hard tech at ~11% of Demo Day while AI agents inhaled "every molecule of startup oxygen mindshare.” Shortly thereafter (#007), YC answered the moment with its "Techno-Industrialist" request for startups. They’ve lived up to it: in this latest batch, a quarter of the ~200 companies could fairly be categorized as “hard tech” / industrials, now the second-largest category in the batch behind only B2B, and up 40x from the trough a few years back, by YC’s own telling. Viva la renaissance in hard pursuits!
SPACE POWER. The White House has issued a 1,000-launch-a-year target via NSPM-17, the first rewrite of national space transportation policy since 2013. Specifically the rule states that U.S. ranges should support 1,000+ launches and reentries a year by 2030. (For context we’ve had 176 licensed launches and reentries to date this year.) The memo directs the alphabet soup to expand launch and reentry capacity, open more federal range access to commercial operators, modernize pad infrastructure, tackle airspace and spectrum constraints, strengthen the space transportation industrial base, and buy more commercial transport for federal missions. It also starts two clocks: agencies have 180 days to produce federal range-scheduling criteria (with schedules published regularly thereafter), and DOT is tasked with identifying sites for new spaceports.
Breaking the fourth wall for a moment… this is squarely what we’ve got our sights set on for next week — launch cadence and capacity, and why it will be a big problem for the U.S. in the coming years. Sell-side analysts, VCs, and price-insensitive assured-access buyers (AKA the government) love to say: launch is solved. Sure, from a purely technological standpoint, that’s true. But the market reality for commercial buyers is a completely different story.
So stay tuned for our full thinking next week. In the meantime, we want your take.
How many launches will America hit in 2030?
And if you buy launch for a living — write in with what you're seeing on price and availability. Anonymous on request; your answers will contribute to next week’s writeup.
SEAGLIDER ERA, DAY ONE. On September 2, REGENT's 15,000-pound, all-electric Viceroy prototype lifted off Narragansett Bay with two pilots aboard — the world's first human-crewed seaglider flight. The float-foil-fly craft moves in three stages: it leaves the dock like a boat, rises up onto hydrofoils as it gathers speed, then lifts off and flies a few dozen feet above the water, riding the cushion of air between wing and sea that lets it carry more weight on less power. The maiden hop was about 2,000 feet at 33 feet up.
The production Viceroy carries 12 passengers at 180 mph on coastal hops, and is regulated as a boat rather than a plane, which provides a far faster road sea lane to market. Deliveries are slated to begin in early 2027 against a $9B+ order book of 600+ vessels.
Congrats to Billy Thalheimer and the REGENT crew — years of hard work culminating in 30 very good seconds. 👏
|