On Wednesday, September 17, several hundred technologists, legislative staff, software developers, open government advocates, and so-called “civic hackers” streamed into an auditorium beneath Capitol Hill for the 7th Annual Congressional Hackathon. Despite a looming shutdown and rising political violence, House Speaker Mike Johnson made an appearance.
The ostensible focus of the 2025 event was how to leverage artificial intelligence tools in government. But as Speaker Johnson strutted onto the stage to give opening remarks, it was immediately evident that this year’s vibe was off. The second Trump administration’s ruthless approach to civil service reform, epitomized by images of a chainsaw-wielding Elon Musk, had dashed public sector technologists’ modest hopes for genuine reform. Minutes into Johnson’s speech, a heckler shouted “You’re firing everyone!” As Johnson invoked “efficiency” as the paradigm of government reform, attendees whispered, snickered, or shook their heads.
Johnson had an answer for these skeptics. “In many ways,” he said, “the hackathon pioneered the DOGE model.” He was onto something, but perhaps not quite in the way he thought. The hackathon skyrocketed to prominence in the 2010s, propelled by Obama-era aspirations for digital democracy. Far from a development out of left field, DOGE was an outgrowth of the faith in technological fixes that was nurtured during those years. Given the disappointment of those earlier hopes, DOGE’s failure to live up to its own hype was entirely predictable.
‘Hackathons” are intensive, often multi-day events that bring technologists and other experts together to brainstorm and build new tools. This genre of gathering was inspired by the “hacker conferences” that spread alongside personal computing in the 1980s and ’90s. The hackathon model prioritizes openness, decentralization, hands-on cooperation, and the free exchange of ideas. Participants roll up their sleeves and troubleshoot technological solutions. Today, there are tens of thousands of hackathons per year.
Despite the regular demonization of hackers as criminals in American media, by the end of the 20th century they came to embody the freedom afforded by technological prowess in computerized society. They were the antiheroes of the internet age—gunslingers in cyberspace, homesteaders on the electronic frontier. Futurists cast the hacker as the embodiment of the entrepreneurial spirit and the torch-bearer of the American dream, and Silicon Valley gurus like Paul Graham described them as artisan-technicians to rival Thomas Edison. The influence wielded by the tech industry site Hacker News is a testament to the hacker’s cultural capital.
The 2000s saw the stock of Silicon Valley soar in America. Glossy magazine profiles and popular shows like Silicon Valley put the place and the people who lived there on the country’s cultural map. Meanwhile, initiatives like “One Laptop Per Child” promised that cheap digital devices would democratize education across the world. From Arab Spring protesters to hacktivist groups like Anonymous, tech-savvy activists seemed poised to establish a vanguard of democracy in the backyards of dictators. (Even though those hopes were quickly dashed.) Hopes for a future of far-flung supply chains, high-tech manufacturing, and novel forms of knowledge work for college graduates were buoyed by a belief in progress driven by technology.
“For a brief moment, it seemed that the global village was finally within reach.”
It was in this context that hacking became a metaphor for reforming America’s political order. This was largely thanks to Barack Obama, the “social media president.” It was in the early years of Obama’s presidency that the Congressional Hackathon was born. First held in 2011, the idea for it was hatched at the Facebook headquarters. The first Hackathon, overseen by then-Majority Leader Eric Cantor and Democratic Whip Steny Hoyer, channeled hopes that technologists had something positive to contribute to the public good. “Programmers and software developers are uniquely positioned to improve our nation’s legislative institutions because it will be their creations that will empower citizens to more meaningfully engage in the legislative process,” stated the report from the first gathering. The subsequent edition proposed solutions to make Congressional data readily available and provide easier access for constituents.
The Congressional Hackathon was followed by other gatherings in a similar vein. In 2013, the White House held an “Open Data Day Hackathon,” where 16 hackers used government data to generate visual tools, widgets, and software packages that allowed researchers to use government data. The tool “Widget the People,” for instance, allowed local petition gatherers to create a digital “thermometer showing how many signatures your petition needs before it reaches the response threshold.” But the event’s signature production was “We the People,” a digital platform that made petitioning the government easier for Americans. By 2014, the platform had registered 15 million users and collected more than 22 million signatures on more than 360,000 petitions. For a brief moment, it seemed that the global village was finally within reach.
The actual political change hackathons produce is hard to measure. It is piecemeal and decentralized, conjured by a confluence of technical experts, nonprofit networks, industry partners and legislative staff in short doses year over year, and delivered in the form of app prototypes or project code dumped onto GitHub.
What we do know, however, isn’t encouraging. In a 2021 analysis of hackathons, researchers surveyed repositories of code from hackathons on GitHub and found a graveyard of techno-optimistic dreams. The study examined the use of all 11,889 hackathon projects from Major League Hacking (MLH) events in 2018-2019 which had GitHub repositories. Of these projects, 85 percent of updates were made within the first month and roughly 77 percent of the total updates occurred within a week of the event. After a month, only 7 percent of projects showed any signs of activity. In other words, the lifecycle of hackathon code is about the same as a once-worn item from H&M.
Despite these less than impressive results, the hope that creative geniuses will totally rethink processes and modernize organizations from within, sleuthing through code in caffeine-crazed sprints is unlikely to disappear. Hacking appeals to the seductive idea that technical fixes can solve our problems—political, personal, or otherwise. From the body politic in Washington, DC, to our very flesh and bone—which transhumanist “biohackers” like Bryan Johnson seek to manipulate through emerging medical technologies, the concept of hacking is no longer the mere label many for computerized subcultures. It is an overriding concept, a metaphor for modern living, continually creeping into new corners of experience in the 21st century.
Considering this history, perhaps the long arc of hacking always bent toward something as haphazard as DOGE. The logic of rogue tech-savvy experts calling all the shots is merely the latest permutation of Obama era techno-optimism. For two decades, our political culture empowered programmers to solve the problems our institutions could not. This maneuver was born of genuine hope, but also betrayed a bottomless despair: Our institutions are so inert, our systems so outdated, and the incentives for meaningful reform so absent that lawmakers’ best recourse was to devolve procedural reform to cohorts of volunteers. Government itself has become a cascade of “accountability sinks,” to use the writer Dan Davies’s term—that is, systems so complex that assigning any responsibility for the mess is a guessing game.
As large language models continue to improve and tech companies compete to hyperscale their computing infrastructure, tech solutionism has gained momentum again. A recent executive order from the Trump White House suggests AI will allow us to cure pediatric cancer; nonprofits and startups promise that AI can democratize access to legal services, an end-run solution to our overwhelmed courts; the World Economic Forum outlines how AI can help fight climate change, even as data centers contribute to rising carbon emissions. But these problems will require more than a room full of programmers and new promises of a technical fix.