Building the Physical World: Why Software-Style Speed is the New Industrial Moat

The core idea here is that treating physical systems like code is the only way to reverse decades of industrial stagnation. By moving from manual, fragmented workflows to model-led, autonomous design, companies like Unlimited Industries and Diode Computers are trying to shrink project timelines from years to months. The result of this shift is not just faster construction, but a fundamental change in how capital and labor interact. For leaders and investors, the advantage lies in realizing that the bottleneck is no longer engineering talent, but a lack of digitized, manufacturable design data. Those who build the infrastructure to turn ideas into physical reality today will own the industrial capacity of tomorrow, creating a moat through sheer operational speed.

The Hidden Cost of Human-in-the-Loop Designs

Most industrial sectors use a stage-gated process that prioritizes risk reduction over speed. While this feels safe, it creates a trap: the design process becomes so slow and fragmented that by the time a project is ready for construction, the original assumptions are often obsolete. Alex Moden notes that in large-scale infrastructure, the industry has lost the ability to build ambitiously.

"If you spend six months designing something and say you have another three months before you lock everything and you fund the project and you build it, if you want to change something six months in, you start over. It's like a total nightmare."

-- Alex Moden

This creates a feedback loop where the difficulty of iteration discourages innovation. In contrast, an everything-as-code architecture allows for parametric adjustments. Instead of restarting a project, engineers simply update variables, turning a rigid, fragile process into a resilient, iterative one.

Why Obvious Fixes Fail to Scale

Conventional wisdom suggests that the path to industrial automation is through better robotics. However, Davide Asnaghi points out that the electronics industry already has the robotics; the failure lies in the 80/20 gap, or the 20 percent of components that do not fit standard automated processes.

The industry reliance on manual labor to bridge this gap is a terminal strategy. You cannot scale data center production by simply hiring more people, especially when the trades are facing an aging workforce and labor shortages. The competitive advantage does not come from building a better robot, but from redesigning the product itself to be manufacturable by the machines you already have.

"The biggest problem is that there is a 80, 20 robotic automation versus like manual labor. And so right now, nobody has really bridged that gap in the United States."

-- Davide Asnaghi

The insight here is that constraint-based design is superior to brute-force assembly. When a design is optimized for manufacturing from the start, the need for human intervention drops to near zero.

The 18-Month Payoff: Why Patience is a Competitive Moat

The most non-obvious dynamic is the shift in total cost of ownership. While most teams optimize for capital expenditure, the real payoff is in schedule compression. In project finance, time is the primary variable for internal rate of return. If a facility takes one year to build instead of five, the financial impact is massive, yet most firms remain stuck in the 1990s because their incentive structures punish the adoption of new technology.

This creates a unique opening for startups. By vertically integrating enough of the process to provide a clean interface to the industry, companies can bypass the need to convince traditional firms to change their ways. They do not sell software; they sell a finished, optimized physical product. This requires a level of patience most startups lack, specifically the willingness to do the hard work of manual integration today to earn the right to automate the system tomorrow.

Key Action Items

• Audit your design-to-build loop: Identify where your team is using a human-in-the-loop approach and ask if that step is a necessity or a failure of your design constraints. (Immediate)
• Adopt everything-as-code primitives: Move away from static, snowflake designs toward parametric models. If you cannot iterate on a variable, you are not building for speed. (Next 3-6 months)
• Prioritize manufacturability over sophistication: Stop designing for theoretical perfection. Design for the machines you already have. This creates immediate, measurable efficiency gains. (Next quarter)
• Invest in data generation: If you are in a hardware-adjacent field, treat your design artifacts as training data. The last frontier is the lack of digitized, manufacturable design data. (12-18 months)
• Focus on the total cost of ownership: Shift your optimization function from unit cost to schedule velocity. The payoff for shaving months off a project timeline far outweighs minor engineering efficiencies. (Ongoing)
• Build the infrastructure, not just the product: Focus on creating the rails, such as compilers, tools, and libraries, that others can build on. This creates network effects and generates the data needed for future AI models. (12-18 months)