Newsletter 041

Howdy folks, and happy Friday. It’s been quite a week since we last ‘spoke.’

Re: Operation Epic Fury, if you're anything like us, you’re stealing a few minutes from an already overdrawn day to find more updates on the latest. Unfortunately, but predictably, a lot of what you’ll find online has obsessively fixated on one narrow slice of a much bigger story: the cost of shooting down cheap drones with expensive interceptors. The most viral version goes something like this: we are shooting down $10,000 drones with $4M Patriot missiles! We will run out of all air defense systems by Sunday!!!

There’s some directional truthiness embedded in this meme, given the cost-exchange issues posed by cheap drones, but it doesn’t tell the full story — far from it — because it tends to rely on wrong/dated info, conflates various pieces of a layered defense, and ignores the offensive campaign suppressing launch rates. Most importantly, we think, it’s too quick to surrender to sky-is-falling doomerism over the harder but necessary work of finding and fielding solutions at scale.

A more productive conversation might start with the drone that opened Pandora’s Box, a most peculiar story that reads like it’s straight out of a le Carré novel…

IN THIS WEEK’S EDITION:
🌊 Ships, Shaheds & Scale
🧪 Chemicals are underrated
🍁 Greetings from Canada
🏰 Payloads, Privatization, & Physical AI
📬 Feedback from Wisconsin

Forwarded this email? Subscribe to Per Aspera here.

The world’s most consequential drone started life on a West German drafting table during the Cold War, scattered through the post-1989 arms bazaar, and resurfaced in the hands of the mullahs. To understand the conflicts of today and wars of tomorrow, we must follow the Shahed.

The Cold War Concept That Wouldn’t Die: In the 1980s, the West Germans (Dornier GmbH) and Americans (Texas Instruments) partnered to develop a disposable anti-radar drone called the DAR, a cheap, single-use system designed to loiter over and dive into Soviet radar sites. Standard Cold War Stuff. Only the war ended, and the Germans shelved the system.

What came next is hard to definitively trace, but goes something like this: South Africa, building its own version of the concept, sold their technical designs to Israel (as did the Germans, reportedly). From this, Israel created Harpy, a one-way, radar-seeking drone that became quite the consequential arms export — leading to Washington temporarily freezing Israel out of the F-35 program.

The Story of the Shahed. Our central antagonist, frozen out of the global arms market, took an interest in this concept. Iran acquired the South African documentation, reverse-engineered a German piston motor, and pulled in COTS avionics. What came out the other end was a new family of drones, starting with the smaller Shahed-131. The larger Shahed-136 was publicly unveiled Dec. 2021 with a delta wing and 50kg warhead. The key point: this could be built in enormous quantities by people without access to sophisticated anything.

A Drone Spreads Its Delta Wings: That design has since been franchised to Russia, which created the Geran drones used extensively in Ukraine, and “exported” to Yemen, where the Houthis stood up their own production lines and used the drone to shut down shipping in the Red Sea. And then, this past weekend, the drone completed the most improbable leg of its journey: CENTCOM confirmed the first combat use of LUCAS, an American reverse-engineered Shahed, over Iran. LUCAS is an amazing success story, featuring a ~15-person, bootstrapped shop in Arizona, SOCOM and CENTCOM embracing an unconventional strategy, and a remarkably fast time-to-field.

Today’s field note, available in full online, traces the arc of the Shahed through the lens of its industrial history. We unpack why it became one of the most proliferative arms in modern history, grapple with where we go from here, and draw seven (early) lessons. Here are three:

The other four are in our 1,500-word piece. If you’re looking to go deeper than the expensive interceptor bad, cheap drone good meme waging saturation attacks on our timelines, media, and minds … look no further.

One of our more under-discussed, industrial-base bottlenecks seems to be: chemical processing. At $673B annual revenue, the U.S. has one of the largest chemical industries on earth (hey, that ain’t nothing!) — but it's a volume industry of polyethylene, ammonia, chlorine (~98% pure) for plastics, fertilizer, and water treatment.

What fabs and gigafactories need is something different: purification chemistry. Hydrofluoric acid distilled in quartz-lined vessels to 99.9999% purity. Lithium recrystallized to electrochemical spec. Solvents cleaned to parts per trillion. Closer to pharmaceutical manufacturing than petrochemicals — and while the US has some domestic capability here, it's not at the scale the buildout assumes. For the key feedstocks that enter these processes, we’re highly import-reliant.

All megaprojects, gigafactories, fabs, and rearmament production lines share one assumption: the chemical inputs will be there. While some sector-specific funding is out there — hundreds of millions of dollars flowing through DPA Title III into munitions‑related chemicals, and up to $325 million in CHIPS funding for Hemlock’s polysilicon capacity — we don’t have a holistic program focused on purification as a shared, cross-cutting layer underneath it all. For chips alone, McKinsey estimates ~$9B in chemical supply chain investment needed, with ~$4B still unfunded. Japan, Germany, and South Korea already have a lot of this at scale. The Renaissance, for now, runs on allied chemistry — which is much better than, say, the rare-earths situation.

If you are an American chemist living this daily, we want to hear what you are seeing: 💌 [email protected]

FROM THE DESK OF RYAN DUFFY: I spent the last ~100 hours in Toronto to attend PDAC, the Super Bowl of Mining. It was buzzing this year. Critical minerals were center-staged, sovereigns came to play, and industry/investors showed up in equal force.

But why was I here? At Array Labs, where I lead BD, we’re in the business of providing reliable, very-high-resolution elevation data/analytics to customers who need to understand and measure what’s happening in the real world. In this context, think of it as upstream telemetry for the people and machines that mine, move, and refine rocks for a living. What rocks have moved where? How much? When? Across my portfolio? What about across the world? In the tropics, where clouds make optical collections tricky? At remote sites with no planes, drones, or boots on the ground? Are we on track or trailing on our targets?

On the floor, hundreds of junior explorers (early-stage, sometimes publicly listed operators with claims but no operational mines) packed into the “Investors Exchange,” alongside some 90 pavilions. The event is a leading indicator for mining sentiment and financing conditions over the next 6-12 months. From this macro perch, you could feel immense excitement around exploration, due to rising commodity prices (copper, gold, etc.) and broader secular tailwinds (see below). But some investors, with battle scars from past commodity cycles and the boom-and-bust nature of this business, fret that this excitement (or, in their words, exuberance) is an indicator of late-cycle froth.

This time, it’s different?! These are famous last words but the case is at least compelling. You have sovereign-backed supply-side guarantees; the rapid formation of free-world, full-stack alliances; a near-daily rate of new, deepening public/private tie-ups being announced; and the broader catalysts of electrification, rearmament, the AI Supercycle, and datacenter builds.

Playing host at PDAC, Canada used the event to roll out the $1.5B First and Last Mile Fund, a $2B Critical Minerals Sovereign Fund, and 30 new partnerships under its ”Alliance” set to unlock $12.1B in project capital with partner nations. You can also feel the acceleration at a regional level: Ontario, arguably the province doing the most to move the needle, has One Project, One Process, cut mine permitting timelines in half, signed economic partnership agreements with First Nations in the Ring of Fire (a 5,000 km² belt of nickel, chromite, and copper worth an estimated ~$22B), and announced this week that road construction starts this June, five years ahead of schedule. Broadly speaking, there’s a sentiment among the Canadians that they indeed have “cards to play” with their resource reserves and central role across production, processing, and refining. And rightfully so.

💭 Is this useful? The above is an experimental format: fire-from-the-hip, off-the-cuff, conversational dispatches that could be from any one of us (me, Dan, Joy, Jeff), sent from wherever we might be, concerning what we’re working on/thinking about. Let us know if you want more of this — if not, we’ll take the hint!

001 / FROM THE DEPT. OF FBC… The Space Force's top acquisition general and now Senior Advisor, Maj. Gen. Stephen Purdy, recently told a finance crowd in Dallas that the “path to mass-produced launch” is more or less solved, and that the bottleneck has moved to payloads. USSF is looking to get cost-effective, mass-produced payloads out the door fast: "Two to three years is too slow. We've got to get down to one week." Of the STRATFIs awarded to startups since 2020, only one went to a launch company, with a clear funding trend toward spacecraft and sensors. Affordable at scale. Much faster: 1x/wk, not 3x/yr. Sound familiar? Read up on Faster Better Cheaper — Part I and Part II — in case you haven't yet.

002 / THE ARMY WANTS A CO-FOUNDER… In the late '90s, the Army had a $20B housing backlog it couldn't fix through appropriations. So it turned to privatization: 50-year leases, private capital, 86,000 homes across 44 installations, and $32B raised. It worked. Now the Army hopes to run it back across a $150B infrastructure backlog spanning energy, manufacturing, supply chains, and critical minerals that Congress can't close with appropriations alone. The Strategic Capital Initiative is an open call for JVs, long-term leases, and dual-use manufacturing on Army land, where private partners get stable demand and commercial upside. There’s a lot of you who might be a good fit for this. The RFI is live on SAM.gov, responses due April 2.

003 / BACK TO THE FLOOR… OpenAI's former Chief Research Officer, Bob McGrew, has founded Arda, a startup that deploys video-based AI models onto factory floors to learn how physical tasks get done, then automates them ($70M raise, $700M valuation, Founders Fund and Accel leading). McGrew was the second engineer at Palantir, where he helped build the forward deployed engineering playbook we wrote about in our popular FDE antimemo (which earned his stamp of approval). Arda, it seems, is looking to channel a similar thesis: get as close to the problem as possible, learn in the field, feed it back into the platform. McGrew is the latest addition to a long list of tech titans who are taking their talents into the arena of hardware, factories, and the physical economy (eg, Jeff Bezos & Project Prometheus). You just can't sit this era out!

Mike from Wisconsin wrote in after our piece on America's copper paradox — where we laid out how the U.S. mines copper concentrate, ships it to China for refining, then buys it back as finished product. Our argument: the Great Lakes corridor has the water, the workforce, and the industrial infrastructure to bring that processing home. Michigan alone sits on 20% of the world's surface freshwater. The smelter-grade water is right there.

👇 Mike disagrees.

Getting pushback is part of what makes this community (and this country) so great. Keep it coming!