CHAPTERS
- 0:00 – 2:14
AD50 and the wake-up call to rebuild America’s industrial base
Elena Burger opens by asking Connor Love and Erin Price-Wright about the AD50 list and why the American industrial base became the dominant theme. Erin argues that defense readiness depends on having real industrial capacity, and recent conflicts have exposed the consequences of decades of hollowing-out.
- •AD50 frames the moment as a “wake-up call” to start building again
- •Defense capability is tied directly to a broader industrial base, not just defense-specific suppliers
- •Recent world conflicts made industrial shortfalls visible in battlefield outcomes for soldiers and allies
- •Connor describes multiple forces converging: geopolitics, capital markets, and entrepreneurs choosing to build
- 2:14 – 3:30
Seeing the industrial stack end-to-end: inputs → factories → systems
Connor explains how AD50 is organized around the full chain from raw materials to factories to end systems. He argues many people focus only on end products, but resilience and competitiveness require viewing the ecosystem as an integrated whole.
- •AD50 is structured around three lenses: inputs, factories, and systems
- •The industrial base should be evaluated as a totality from ground extraction to end product
- •Many companies are intentionally designing with supply-chain realities in mind
- •The approach emphasizes that each pillar is necessary for the final system to succeed
- 3:30 – 5:09
Defense demand as a catalyst—and the limits of “dual-use” thinking
Connor describes how government demand has historically seeded technologies that later spread broadly, but says private capital is now often the initial driver. He notes that while many startups sell into government first, the factories and capabilities they build inevitably spill over into civilian markets.
- •Historically, government-led projects created long-tail commercial benefits (space, internet, nuclear energy)
- •A shift is underway: private industry and private capital increasingly lead development
- •Factories, materials, and energy capacity built for defense can serve many civilian applications
- •Connor cautions that chasing too many use cases at once can be a startup trap even if dual-use is real
- 5:09 – 6:51
DJI’s arc: from consumer drones to defense scale to robot actuators
Erin uses DJI as an example of knowledge transfer between consumer, defense, and broader industrial robotics. She argues that manufacturing capability and components—like motors and actuators—translate across domains, and that consumer pressure helps drive cost and performance improvements beyond what government purchasing typically demands.
- •Erin recounts DJI’s evolution from consumer drones into public safety and defense markets
- •DJI’s drone motor expertise translated into robot actuators, described as a hard robotics problem
- •Process and technology transfer between military and non-military use cases is presented as longstanding
- •Erin describes the federal government as helping subsidize early process learning, while consumers drive capability and price pressure
- 6:51 – 8:38
First-of-a-kind manufacturing economics and defense as early CapEx support
Connor and Erin discuss why first-of-a-kind physical products are expensive and how early buyers can change unit economics. They highlight defense as a customer willing to pay initial costs that help finance factories and equipment, which later benefits commercial markets.
- •Building bespoke, first-of-a-kind systems is inherently costly at the start
- •Government willingness to pay early can underwrite factory CapEx and accelerate scale
- •Connor emphasizes how early demand enables more machines, more production capacity, and eventual consumer viability
- •Erin gives PCBs as an example: defense-backed PCB fabs can later supply robots and medical devices like hearing aids
- 8:38 – 10:16
The SpaceX/Tesla lesson: vertical integration when the industrial base isn’t there
Asked how startups drive down costs and move faster, Connor points to SpaceX and Tesla as rare modern examples of building at scale. He and Erin explain that vertical integration can be necessary when suppliers don’t exist, while also noting founders should use high-quality suppliers when available.
- •Connor cites SpaceX and Tesla as key models for building physical systems at scale
- •Vertical integration is framed as situational: necessary when supply chains are missing or exquisiteness/cost must be controlled
- •Erin argues SpaceX and Tesla had to build missing industrial capacity themselves
- •They discuss Tesla’s need to build battery supply chain and power electronics/drivetrain capabilities
- 10:16 – 13:18
Mariana Minerals: owning the mine to reinvent the entire process with software and AI
Elena brings up Mariana Minerals, and Erin explains why the company bought a copper mine rather than relying on incumbents. Connor adds that in the physical world, owning the asset can be required to change it, and both emphasize rethinking the full workflow while capturing data to improve the system over time.
- •Erin says Mariana believed incumbents wouldn’t adopt technology fast enough to “move the needle”
- •Mariana is described as owning the value chain: acquire, develop, mine, operate, refine, and sell the commodity
- •Software and AI are presented as major components across stages, enabled by owning the process
- •Connor stresses “showing is believing” in physical industries and notes end-to-end ownership enables data feedback loops
- 13:18 – 14:11
Why AI still can’t ‘turn a wrench’: many small improvements beat a silver bullet
Connor argues AI will be powerful in the physical world but cannot yet replace hands-on work, while Erin notes progress is happening. They discuss how modernization comes from deep process understanding and many incremental changes across quoting, production, inspection, and logistics rather than a single AI feature.
- •Connor states AI will be a phenomenal tool in the physical world but “cannot turn a wrench yet”
- •Modern physical operations improve through many steps, not one “shiny” AI product
- •Examples include factories like Hadrian optimizing multiple nodes: quoting, making parts, inspection, transfer
- •Both emphasize first-principles review of processes that have been done the same way for decades
- 14:11 – 18:04
AI-enabled flexibility in factories: making high-mix, low-volume work viable
Erin explains how AI tools are changing the economics of automation, especially for high-mix, low-volume manufacturing that’s traditionally expensive to automate. She describes technicians using tools to reconfigure lines rapidly, reducing reliance on slow, costly system integrators; Connor ties this to faster prototyping and quicker rate production.
- •Erin contrasts low-mix/high-volume vs high-mix/low-volume manufacturing and automation challenges
- •Manufacturing technicians can use new software tools to retool lines faster and cheaper
- •Erin gives an example of using Claude to reconfigure a line in minutes instead of months via integrators
- •Connor connects flexible lines to faster iteration, prototyping, and scaling to rate production; both highlight “design for manufacture”
- 18:04 – 20:37
Building before the government asks: private capital changes defense procurement dynamics
Connor explains why defense has historically been difficult for startups, citing misaligned incentives where the buyer and user differ. He argues a major change is private capital funding products ahead of formal requirements, enabling companies to build and prove systems sooner and deliver more current technology to soldiers.
- •Connor describes defense’s disjointed decision-making: buyer, user, and appropriator are separated
- •Historically, primes waited for government money and delivered slow, backward-looking systems
- •Private capital now funds building in advance of government-directed requirements
- •Connor says this can yield better products aligned with current technology and battlefield needs
- 20:37 – 24:47
Why founders choose brutal physical-world problems—and why they can be highly valuable
Erin describes the demanding reality of building mining and manufacturing companies—remote travel, hard negotiations, and time away from family. Connor and Erin argue the big shift is that these companies are not only important but can also be among the most valuable, attracting ambitious founders to work on hard, consequential problems.
- •Erin details the rough, physical realities of building companies like Mariana, including remote sites and difficult supplier culture
- •She frames many builders as “missionary founders” drawn to enormous, worth-it problems
- •Connor says AD companies can now be both important to national life and highly valuable businesses
- •Erin notes a “full circle” return to physical products and says many top valuable companies are hardware-focused
- 24:47 – 28:28
What site visits reveal: steel-toed boots, factory time-lapses, and visible momentum
Elena asks what it’s like visiting industrial sites, and Connor and Erin describe the grounded, sometimes unglamorous experience. They talk about the satisfaction of repeat visits as warehouses turn into busy production spaces, along with cultural markers like giant flags and on-site gyms that reflect team identity and purpose.
- •Connor describes needing steel-toed boots and visiting remote, physically hazardous sites
- •He notes a playful competition over who has the largest American flag at a site
- •Erin highlights the value of revisiting facilities to see rapid build-outs and operational chaos as capacity ramps
- •They mention time-lapse filmmaking as a common way companies document factory build progress
- 28:28 – 32:05
Defense manufacturing as the on-ramp to a robotics wave—and better domestic suppliers
Elena asks whether making components in the U.S. creates a compounding effect; Connor and Erin emphasize cost, quality, and iteration speed improvements. Connor argues defense systems require subsystems similar to future robotics, and Erin gives a PCB quality story to show how a domestic supplier ecosystem has improved quickly through new startups.
- •Connor and Erin say onshoring creates compounding gains in cost, quality, and speed of iteration
- •Connor links defense subsystems (actuators, motors, valves, switches) to future robotics needs
- •Connor claims robotics firms today often source heavily from China and want to onshore for multiple reasons
- •Erin describes improved PCB yield as new suppliers applied software/modern methods, reducing waste from quality failures
- 32:05 – 41:05
The ‘manufacturing middle’: co-engineering a new supply chain for new architectures
Connor argues that new end systems (military, planes, cars) force changes in the inputs and subsystems beneath them. He describes an opportunity for a new middle layer of suppliers to become co-engineers—rethinking and rebuilding components rather than simply fulfilling old specs—while Erin parallels the same shift happening in the power grid as it moves toward software-native power electronics and storage.
- •Connor describes legacy procurement patterns where suppliers repeatedly deliver old designs because incentives favor reuse
- •He argues new end-system requirements will force re-engineering of subsystems using new materials and approaches
- •Connor predicts a new “manufacturing middle” that co-engineers with OEMs instead of staying in receive mode
- •Erin applies the idea to the grid: moving from static/mechanical systems to modern, software-native power electronics creates a chance to rethink the entire supply chain
- 41:05 – 49:17
America’s edge: capital markets, community participation, and the will to build
Erin and Connor discuss how U.S. capital markets fund expensive physical build-outs and can finance development ahead of certainty, contrasting this with top-down state direction. They close by addressing the U.S.–China power-production gap as a question of national will, and they argue physical projects must bring communities along so more people share in the upside across regions beyond coastal tech hubs.
- •Connor says U.S. capital markets can follow entrepreneurial success and fund ahead of obvious demand
- •They note more financial players are participating (venture, private equity, credit), which matters because AD companies need large capital
- •Erin frames the U.S.–China power gap as not primarily a technology problem but a question of national will to build
- •Both emphasize physical projects take time and must include communities so economic growth is widely shared, reducing wealth concentration and expanding prosperity
