The Deus Ex Machina Delusion
All views in this newsletter are my own and do not represent the views of The R Street Institute, the US Navy, or any other organization I am affiliated with.
Quick note: I'm on travel this week, so this newsletter is shorter than usual—consider it a trip down memory lane that reveals a pattern I'll explore more deeply in the coming weeks.
In the 1930s, France faced a strategic problem. They had to figure out how to defend against German aggression while maintaining political stability and economic prosperity. French military planners responded with an engineering marvel: the Maginot Line. This elaborate system of fortifications along the German border represented the most advanced defensive technology of its era, complete with underground railways, air conditioning, and supplies for extended sieges.
The Maginot Line failed catastrophically not because the technology was inadequate, but because France had made a categorical error. They treated a strategic problem as an engineering challenge. When political constraints prevented extending fortifications along the Belgian border, Germany simply went around them in 1940. France had invested billions in a technological solution that couldn't address the actual strategic question.
We're Making the Same Mistake
I touched on this a few weeks ago, when I critiqued the FCC's decision to ban Chinese government-controlled certification laboratories. The agency identified a real vulnerability: labs with access to product designs, source code, and technical specifications months before market launch, operating under China's 2017 National Intelligence Law that compels cooperation with state intelligence agencies.
The FCC's solution was to remove Chinese labs from the approved list. This satisfies the need to take action, but failed to solve the problem. Products tested in Chinese labs can still reach US markets by routing through European, Japanese, or Korean certification first, then using Mutual Recognition Agreements for US market access. We've built an elaborate fortification along one border while leaving others wide open.
This isn't an isolated example. American technology policy repeatedly commits what we might call the "deus ex machina fallacy" - The belief that the right technology, properly deployed, will resolve strategic challenges that have no technical solutions.
The Pattern Repeats
AI safety frameworks treat risk as primarily a technical problem requiring technical solutions: capability evaluations, red-teaming protocols, compute thresholds. These aren't wrong, but they miss the strategic question: how do we maintain competitive advantage in AI while managing risks when adversaries face no equivalent constraints?
Cybersecurity policy deploys increasingly sophisticated defensive technologies while ignoring that network defense is fundamentally about risk management in an environment of persistent competition. CISA's expanding critical infrastructure definitions treat cybersecurity as a technology deployment problem rather than strategic resource allocation.
Export controls on semiconductors represent sophisticated technical policy. They identify specific chip capabilities, track manufacturing equipment, and monitor research collaboration. But treating technological competition as an equipment denial problem misses that China's strategic objective is technological sovereignty, which our restrictions actively strengthen.
In each case: identify a strategic challenge, deploy a technological response, express surprise when the technology doesn't resolve the underlying problem, then call for more sophisticated technology to fix the first technology's limitations.
We're building Maginot Lines in every domain.
What I'm Working On
Next week's newsletter will dig into this pattern more deeply. Why do we keep treating strategic problems as engineering challenges? What does Clausewitz's concept of "friction" tell us about the gap between theoretical plans and operational reality? And how can we develop better frameworks for distinguishing between problems that technology can solve and problems that require strategic thinking?
For now, the question to ask about any technology policy proposal: "What happens after we deploy this technology?" If the answer doesn't address how adversaries will adapt, what second-order effects will emerge, and how the technology advances strategic objectives beyond just technical capabilities, you're looking at another Maginot Line.
What This Means For...
Policymakers: Before deploying technical solutions, ask whether you're addressing a technical problem or a strategic one. The FCC lab bans, AI safety frameworks, and export controls all represent sophisticated technical policy that can't resolve strategic competition. Technology serves strategy—not the other way around.
Tech companies: Recognize the pattern when you see it. Policy that treats strategic competition as a technical problem creates compliance costs without addressing competitive dynamics. Companies that understand the strategic frameworks driving policy can better anticipate where regulations will fail and require revision.
Aspiring strategic thinkers: Start noticing how often we confuse technical capabilities with strategic solutions. The Maginot Line failed because excellent engineering can't compensate for strategic errors. Neither can excellent AI safety protocols, sophisticated export controls, or advanced cybersecurity tools.
Next week: A deeper dive into the deus ex machina fallacy and what military strategic theory tells us about distinguishing technical problems from strategic ones.