👋 Hey, it’s Jeff Crusey. Your Resident Investor at Per Aspera.

I became a VC in 2008. My peers went off to Silicon Valley to pursue social media apps and I went to Detroit to back energy tech. At the time, Lehman Brothers was collapsing in real-time and the city was in crisis with people bathing in the streets.

My choice was clearly the hard, unsexy path. But I truly felt that being a great industrial investor meant showing up where the backbone of America is, not just where capital is fashionable.

Today, we’re in a special, unprecedented moment: the unsexy is now very sexy! Deep tech is not only back, but it’s being bankrolled by private capital at scales once reserved for sovereign nations.

But, as someone who’s lived through multiple cycles of this business, I’m concerned that bigger isn’t necessarily better.

👇 Find out why.

Forwarded this email? Subscribe for more.

The tinkering, trailblazing pioneers of venture capital were a boisterous sort. They invested not in virtual abstractions, but the real world: semiconductors, computing hardware, early biotechnology, and medical devices. Today, VC has matured into a sanitized asset class whose center of gravity sits on the balance sheets of global financial behemoths.

The last decade was dominated by a highly successful software-first, Moneyball-like playbook. But the era of software maximalism is coming to an end, due to saturation, slowing growth, AI, and the West’s deindustrialization reckoning.

Investors are diversifying back into the complexity, physicality, and technical depth of the real world. Deep tech’s share of VC has doubled from ~10% a decade ago to over 20% today (or 35-40%, if you count AI infrastructure). A generation of founders and investors are coming of age tackling hard technical problems that can’t be commoditized overnight.

The collision between the technology we need and the capital stack we have raises hard questions:

Today’s Antimemo tracks the punishing math of megafunds, why specialists are getting sidelined, and how founders should face and embrace this structural friction.

It’s essential reading for founders, my fellow investors, and LPs — and sure to ruffle a few feathers. Give it a read, and let me know what you think. What resonated? What don’t you agree with?

Per Aspera’s own Dan Goldin (9th NASA Administrator) will be making a big statement on asteroid mining at the Space Economy Summit by Economist Impact, where Editor-in-Chief Ryan Duffy will also be moderating on the space economy.

The Space Economy Summit by Economist Impact will be November 5–6, 2025 in Orlando. Dan G. and Ryan D. will join other industry experts to explore the next leap forward for space and the benefits to businesses here on Earth.

Register: Secure your spot here.

Jet engines are a tough nut to crack. COMAC, China’s state-backed planemaker, has reportedly cut delivery targets for its flagship C919 by two-thirds, expecting just ~25 of the single-aisle jets to reach customers this year (a far cry from Airbus’s ~820 and Boeing’s 430-450 mostly single-aisle deliveries expected in 2025). Chinese ambitions for homegrown aviation are sky-high (ha, ha), but the program remains deeply dependent on imported, Western-made avionics, hydraulics, and most critically, the CFM LEAP-1C engines supplied by GE & Safran.

This dependency makes aviation a bargaining chip in broader U.S.-China trade brinkmanship. (The U.S. froze, then unfroze, GE jet engine exports this year.) Chips are also bargaining…chips…but China seems to believe that Huawei’s new inference line will be good enough to get the job done domestically.

But aviation is a built-different beast. Despite China investing $10B-$20B (est.) into domestic jet engine development so far, planespotters analysts see commercial readiness as years away, with ~2030 being the most optimistic consensus forecast.

Do you hear that? With sovereignty, aerospace, geopolitics, critical technology, and unclear timelines all colliding here, what we’re hearing is that we should turn this one over to you all, the smart cookies reading Per Aspera. Time for a reader poll!

*Editor’s note: Jet engines are arguably the summit of national industrial ambition: high-bypass, highly fuel efficient, exceptional reliability, and rigorous safety/certification standards that are much tougher than military engines. There is significant debate when (or even if) China’s first truly commercial-ready engine arrives. We don’t expect you all to have a firm view — so if more of you pick ¯\_(ツ)_/¯ than any other option, we’ll reserve this as a topic for a deep dive in the near-future.

Project Flo, No Mo. A proposed 468-acre, $1B+ datacenter project slated for the southeast side of Indianapolis was pulled minutes before a city council vote, after residents packed the chambers in vocal opposition. Google and its developer had pitched the project as a transformative regional investment, which fell on deaf ears, as the community pointed to perceived costs without a clear upside.

Per Aspera’s take: We’ve written before about the local veto as a real risk factor in the “AI factory” gold rush. This will get worse before it gets better, as the public weighs rising electricity prices and points its finger at AI datacenters, while also grappling with other perceived costs (land conversion, grid strain, water use, and often opaque incentives).

P.S. remember our recent writeup on Doña Ana, NM, and its bold gambit to land an AI megaproject? Well…OpenAI, Oracle, and SoftBank have selected the county for one of three new GW-scale Stargate sites, newly unleashed by the ChatGPT maker’s megadeal with Nvidia. And already, some local officials and residents are speaking out against the project.

PER ASPERA EDITOR’S PICKS
Surplus stories, developments, signals, and data that stood out while we were pulling together today’s edition.

Humanoids are Meta’s next AR-size bet // Meanwhile, MIT roboticist says humanoids won’t learn dexterity // White House weighs mandating a 1:1 ratio of U.S.-made & imported chips // Sila opens silicon anode battery factory in Moses Lake, WA // NATO investing $728M in space // Germany sinking €35B ($41B) into space capabilities through 2030 // ‘Teraflop’ added to Merriam-Webster, along with dumbphone, beast mode, rizz, and 5,000 other new terms.

FRIENDS IN HIGH, HARD PLACES
What did friends of Per Aspera get done this past week?

Niron Magnetics Breaks Ground. Minnesota-based Niron Magnetics just broke ground on a 190,000-sqft. facility in Sartell, MN that will manufacture 1,500 tons of rare earth-free permanent magnets a year at scale (yes, rare earth free). Their Iron Nitride magnets avoid the costly, polluting processing required for rare earths. Full-scale production is targeted for early 2027, which couldn’t come soon enough.

Skyryse (Founding Sponsor 🇺🇸💥) & Skylar. If you’re new here, Skyryse is the creator of SkyOS, an aircraft-agnostic system that simplifies flight operations for pilots. The startup has just unveiled Skylar, an AI flight assistant that monitors ATC and ATIS transmissions, weather, and air traffic; automates checklists; optimizes flight routes; and even calculates fuel burn on the fly, in real time. As CEO Mark Groden notes: “We use AI to talk to our homes and cars — now we’ll use it to talk to our aircraft.” What a time to be alive!

🎥 Watch the Skylar Preview video.

Got news to share? Breaking ground, shipping hardware, or hitting a milestone? We want to hear about it, and so might thousands of your fellow Renaissance-pilled readers. Reply here or drop us a line at [email protected] ↗