Jordan Schneider's conversation with Jen Pahlka cuts through the usual technocratic fog to reveal a startling truth: the American government isn't failing because it lacks brilliant ideas, but because its operating system is stuck in the industrial age while the rest of the world has moved to AI. While most discussions focus on specific policy wins or losses, Pahlka argues that the real bottleneck is a 75-year accumulation of "regulatory cruft" that no human team can untangle, yet which AI might finally help us prune.
The Architecture of Failure
Pahlka doesn't mince words about the state of the administrative state. She notes that while the mid-20th century boasted an effective bureaucracy, it lost its ability to self-renew. "We got lazy and let policy and process accumulate like layers of cruft — archaeological layers you can dig back through," she explains. This is a powerful reframing of government dysfunction. It's not necessarily that the system was intentionally sabotaged, but rather that it was allowed to ossify. The author draws a sharp parallel to the "work simplification" practices of the Eisenhower era, where agencies constantly streamlined their own processes, a discipline that has since vanished.
The evidence for this decay is staggering. Pahlka points to the New Jersey unemployment insurance system, which operates under "7,119 pages of active UI regulations." This isn't just bureaucracy; it's a structural impossibility for human operators to manage during a crisis. As Pahlka puts it, "That brittleness is especially dangerous for a program that operates at low volumes day-to-day but needs to scale 10x or 20x in claims during a crisis." The sheer volume of text makes the system unadministrable, a fact that only becomes clear when one considers the historical context of the Progressive Era, where the goal was efficiency, not the accumulation of constraints.
"The AI cannot do anything about the political will required to unwind the memos, guidance, policy, regulations, and statutes that need to be unwound."
This distinction is crucial. Pahlka argues that while AI can map and rewrite the code of government, it cannot generate the political capital to pass the laws that would actually delete the old rules. Critics might argue that this places too much faith in a political system that has shown little appetite for such radical simplification. However, the argument holds weight because it identifies the specific missing variable: a clear target. As Pahlka notes, "Until we put forward what we think that should look like, we haven't tested the will of our political leaders to get us there."
The Procurement Trap
The conversation shifts to the economic reality of modern software. Pahlka highlights a bizarre disconnect: while the private sector sees software engineering productivity jump 10x or 100x, the government is still paying for legacy contracting models. "It's going to be decades before government actually pays less for software — and right now we're probably going to start paying more," she warns. This is a five-alarm fire that the executive branch has largely ignored. The procurement systems, legal reviews, and contracting rules are the "bottom of the Maslow's hierarchy of government needs," yet they are the very things preventing the adoption of modern tools.
The author suggests that the solution lies in a cross-ideological coalition to modernize the operating model, a mission central to the Recoding America Fund. The goal isn't just to add AI on top of broken processes, but to fundamentally restructure how the government hires and manages its workforce. "You can't iterate meaningfully on policy when the basics aren't covered," Pahlka asserts. This echoes the logic of Pareto efficiency, where resources are allocated to maximize output; currently, the government is spending vast resources on maintaining a system that produces minimal value.
The Political Ceiling
Despite the technological promise, the ultimate barrier remains human. Pahlka is clear that AI is a tool, not a savior. "The binding constraint isn't the AI. It's our political system," she states. The fear of AI is often overstated; the real fear is that we will lock in new, rigid rules before we understand what is possible. The "cascade of rigidity" she describes means that well-intentioned guardrails in a risk-averse bureaucracy become barriers that cannot be overcome.
Yet, there is a path forward through state-level experimentation. "States are valuable because you have more opportunities to find where the energy is, prove it works, and let other states and cities adopt it," Pahlka argues. This federalist approach allows for a test-and-learn framework that the federal government, with its massive inertia, cannot easily replicate. The hope is that successful models in places like New Jersey or California can create a proof of concept that forces the federal level to adapt.
"We need to start thinking in terms of actually meeting the moment rather than moving slightly ahead from where we are today."
Bottom Line
Pahlka's most compelling argument is that the gap between private-sector expectations and government delivery is now a threat to democracy itself, and AI is the only tool capable of bridging it at scale. The piece's greatest vulnerability, however, is its reliance on political will to execute the very reforms that the current system is designed to resist. The reader should watch for whether state-level pilots can actually generate the momentum needed to force a federal overhaul of the civil service and procurement codes.