Alan Brouilette: the evolution of 18F, & the future of civic tech

The Long Game:

What Government Innovation Can Teach Silicon Valley

The swagger has faded. After months of breathless promises about artificial intelligence revolutionizing everything from customer service to drug discovery, technology firms now confront a sobering reality: massive investments, modest returns, and a market that has lost patience with vaporware. The parallels to government's own experiments with digital transformation are striking—and instructive.

Alan Brouilette, who spent almost a decade as chief of staff at 18F, the federal government's in-house digital consultancy, has watched this pattern before. His insights from building civic technology teams offer an unexpected corrective to Silicon Valley's current predicament. The lesson is not about technology at all. It is about humility, patience, and understanding that real change happens in systems, not press releases.

Brouilette’s first principle sounds almost radical in today's climate: enter with humility. When 18F launched under President Obama, it arrived with the confidence typical of Silicon Valley's best and brightest, convinced that superior technical skill could overcome bureaucratic inertia. That assumption proved costly. The real work of transformation, Brouilette and 18F learned, meant recognizing that existing systems—however frustrating—existed for reasons. Even bad reasons represented someone's logic, someone's approach tacitics, someone's risk mitigation strategy. Progress required understanding these motivations, not dismissing them.

This connects to a second, counterintuitive insight: technical prowess matters less than the ability to navigate human complexity. 18F eventually began hiring for "soft skills"—the unmeasurable qualities that make teams function smoothly under stress. They discovered what basketball analytics had already revealed: that certain players made everyone around them better, even when their individual statistics looked unremarkable. In technology, as proved once again by recent AI deployments, computational power as a constraint is negligible when considering the constraint that trust, coordination, and the grinding work of aligning incentives across organizational boundaries present.

The bureaucracy itself deserves more sympathy than it typically receives. Those overlapping rules and ponderous processes represent an evolved defense against political whiplash. They slow change deliberately, ensuring that enthusiasm alone cannot capsize the ship. Silicon Valley executives now discovering that their AI tools must navigate compliance regimes, liability concerns, and stakeholder skepticism are encountering the same friction. The mistake lies in viewing these obstacles as problems to overcome rather than signals to interpret.

Perhaps most relevant to today's AI disappointments is Alan's final principle: avoid premature fanfare. 18F's high-profile launch created expectations that poisoned future relationships. Every partner agency wondered: "Weren't you going to solve everything? Is it all fixed yet?" Better to build credibility through small victories before announcing grand ambitions. AI firms that promised to revolutionize industries before demonstrating basic competence now face similar skepticism.

The irony is rich: government's supposedly calcified bureaucracy recognized these lessons while Silicon Valley's ostensibly agile innovators repeated old mistakes. 18F was ultimately dismantled in early 2025, but its alumni network persists—a distributed knowledge base more resilient than any single organization. That long-term perspective, that emphasis on building capability rather than dependency, represents precisely what technology deployment needs now. Tech transforms nothing by itself. People, given time and respect, transform systems. 

Article cited in interview: The No-Stats All-Star. By Michael Lewis

What we’re into this week

Aaron

Talking with Alan brought up so many topics, but for me it boiled down to one thing: being able to talk to and relate to people is really the skillset he brought to 18F and tried to impart to everyone there. And I think this particular niche is very ascendant- Business Insider ran a piece back in November with the Chief Innovation Officer of CISCO, a technology company, where he said that knowledge and comfort with the humanities are the most important skill he’s seeing in the landscape currently. 

Ana

I completely agree with this. The Economist just released the third season of Boss Class, their podcast about the working world, focusing on AI and its use in the workplace. They conclude that it will change many things, but most of what it will change is the focus managers and teams can spend on more difficult problems. This, of course, surprises no one in the Ishmael Interactive world, since we talked about it on the CX Pod back in October. 

Scott

Yeah, Alan’s perspective is so valuable. From the other side of the coin, his talk reminded me of this piece from Jack Conte, CEO of Patreon, where he breaks down how “the algorithm” is a distillation of a business’s goals. He shows that, by and large, those goals prioritize ad revenue, which is, to Alan’s point, so short-sighted. In response, Conte presents a proposal for prioritizing creators and connection over advertiser engagement.

Credits

Guest: Alan Brouilette
Host: Aaron Meyers
Text: A. Adams
Producer: Ana Monroe
Artwork: Group of Human Forms; A Man Seated [recto] 1884-1888 Paul Gauguin. Via the National Gallery
Why this artwork? At its core, Ishmael Interactive is a human-centered company. Alan Brouilette directed 18F in the same way, pivoting the group from focusing on technology fixes to the deeper and arguably more interesting human problems that they could tackle. Gauguin did the same, sketching the human form over and over again throughout his life, experimenting with media, scale, and granularity, but never tiring of his subject.