Three Startups Reinventing Critical Infrastructure

Episode Summary

Executive Summary: The episode spotlights three A16Z American Dynamism portfolio founders and their missions: Ulysses’s ocean robotics platform for conservation, infrastructure, and defense; Mariana Minerals’s software-enabled, vertically integrated critical-minerals company; and Radiant’s mass-producible micro-reactors for off-grid power. Across all three, the core theme is rebuilding hard infrastructure with automation, better software, and domestic manufacturing to strengthen U.S. industrial and strategic capacity.

Main Topics: Ulysses and the ocean robotics stack (Priority: 5/5): Will O'Brien explains Ulysses as an 'ocean company' building autonomous surface and underwater robots for conservation, commercial inspection/repair, and defense applications, with a full-stack approach from sensors to recovery vessels. Why the ocean is an underbuilt frontier (Priority: 4/5): The discussion argues that oceans have been culturally neglected compared with space, despite being a historic frontier and a strategically important domain now being re-energized by better robotics, connectivity, and logistics. Mariana Minerals and the critical minerals value chain (Priority: 5/5): Turner Caldwell describes the complex path from ore discovery to refined metal and engineered material, arguing that the industry needs software, automation, and integrated operations to overcome low productivity and supply-chain constraints. China, geopolitics, and industrial capacity (Priority: 5/5): Both ocean and minerals segments frame China as a major competitor with scale advantages, but also as a signal that the U.S. can still build decisive advantage through domestic production, permitting reform, and industrial policy. Radiant and mass-producible micro-reactors (Priority: 5/5): Doug Birdauer explains Radiant’s plan for trailer-sized nuclear reactors that can be deployed quickly for off-grid customers, positioning nuclear as a clean alternative to diesel in high-cost, remote, or island markets. Permitting, regulation, and talent bottlenecks (Priority: 4/5): Across mining and nuclear, founders emphasize that slow permitting, conservative incumbents, and fragmented bureaucracy slow deployment more than technical feasibility does, making policy and execution as important as innovation.

Key Arguments: Ocean robotics is now feasible at scale because cheaper autonomous surface and underwater vehicles can be networked to solve conservation, commercial, and defense tasks. The ocean has been under-invested in due to cultural narrative shifts, classification, and lost institutional ambition, not because it lacks strategic importance. U.S.-China competition in the undersea domain is still winnable for the U.S. because the market is smaller and domestic manufacturing can scale quickly. Critical minerals are essential to nearly every modern product and system, so control of refining and processing matters as much as mining. Mining and refining are bespoke, high-latency, high-risk businesses where a small process change can cause multi-million-dollar losses, which makes software and automation unusually valuable. Mariana’s thesis is that the real opportunity is not just a point solution but end-to-end control of construction, commissioning, and plant operations. Nuclear micro-reactors can compete with diesel in expensive, remote markets without competing with large grid-scale reactors, making them a pragmatic near-term use case. The nuclear sector still lacks mature fuel, waste, and regulatory infrastructure, so building products and ecosystems together is necessary for the industry to flourish.

Data Points: Ulysses portfolio films: 3 - A16Z American Dynamism screened short films about three portfolio companies. Mariana Minerals fundraising: $85 million - Turner Caldwell is introduced as coming out of stealth with $85M raised. Ocean drone production: single-digit thousands per year - Will O'Brien says global autonomous underwater vehicle production is in the single-digit thousands annually. China and U.S. AUV production: hundreds each per year - He says both China and the U.S. make hundreds of underwater drones annually, suggesting no massive disparity yet. China submarine-sized AUV range: 18,000 kilometers - Referenced as a serious undersea threat capable of reaching near U.S. coastline. China AUV operating depth: 4,000 meters - Used to illustrate undersea capability and asymmetric threat. International supply impact: 2–3x - A planned U.S. facility alone could reportedly double or triple international autonomous underwater vehicle supply. Hamilton Society debate attendance: 600 attendees - Turner cites a billionaire-tax debate as an example of cultural impact. Hamilton Society vote count: 400 votes against the billionaire tax - Used to illustrate influence of debate society discussions. ore concentration: less than 1% to less than 5% - Mariana describes typical ore concentration before processing, unless the deposit is world-class. Construction labor example: 13,000 people on site - Turner cites a Chinese nickel refining operation in Indonesia during construction and commissioning. Construction lag: ~3 weeks - Mariana describes a typical lag between field activity and executive visibility on large construction projects. Data center energy reduction: 30% - Referenced from Google/DeepMind optimization of thermal systems as a model for RL-driven industrial optimization. Refinery commissioning time: 2–4 years - Turner says many Western refineries take this long to commission. Chinese refinery commissioning time: 6 months - Contrasted with faster execution in China. Lithium growth need: 4x - Turner says lithium production capacity needs roughly to quadruple over the next decade. Tennessee site size: 80 acres - Radiant just signed an 80-acre Tennessee site for manufacturing. Radiant deployment time: 48 hours - Micro-reactor can be on-site and powering customers within about 48 hours after arrival. Radiant reactor lifespan: 5 years - Each unit is described as lasting roughly five years. Diesel equivalent: 2 million gallons - One reactor is said to equal about this much diesel energy. Initial diesel cost target: $6.50/gallon - Radiant says it beats diesel at roughly this price level, not cheap baseline diesel. Nuclear waste storage: 10 acres - Radiant says its small reactors could be stored in dry cask on about 10 acres of its 80-acre site.

Pivotal Quotes: "“We’re building the ocean company.”" — Will O'Brien: Defines Ulysses’s mission and positions the firm as a broad platform for ocean operations. "“We’re a vertically integrated software-first minerals project developer and operator.”" — Turner Caldwell: Summarizes Mariana Minerals’s business model and full-stack thesis. "“There is no nuclear industry.”" — Doug Birdauer: Used to argue that the sector is still pre-industrial and lacks mature fuel, waste, and product infrastructure.

Implications: The episode argues that strategic industries will be rebuilt through full-stack execution, not point solutions. For listeners, the takeaway is that robotics, software, and policy are converging to make oceans, minerals, and nuclear power newly investable and geopolitically critical.