From Competition to Collaboration: how hackathons are evolving
I’ve always enjoyed hackathons. There’s something energizing about the compressed timeline, the scramble of creativity, the way strangers become teammates or friendships deepen under pressure. In their original form, hackathons were often competitions: a race to build an app, solve a business problem, or make something faster or more efficient. That model still has its place, but for me, it was always the spirit of the event, not the code or the prize, that mattered most.
In recent years, though, I’ve noticed a shift. At Carnegie Mellon, the Libraries have hosted several hackathons that move in a different direction, where it’s less about winning and more about collaborating, learning, and exploring. These events bring together people from across disciplines to share tools, surface questions, and build something meaningful in a short amount of time.
Hacking Open Science
I began reflecting on this shift a few years ago when I attended a bioinformatics hackathon hosted in 2023 by the CMU Libraries. It brought together interdisciplinary teams to build open-source tools and data pipelines, connecting students from computer science, machine learning, and biology in ways that felt very fluid.
I had several conversations at the event with Melanie Gainey, Director of our Open Science Program. She mentioned that so much of her work involves talking about open science in the abstract — through slide decks, presentations, and conversations across campus. What stood out to her was the realization that “this is what it actually looks like” when open science becomes practice, not just principle.
That comment stuck with me. We can offer workshops, LibGuides, modules, and demos. We can talk about tools, processes, and platforms that support open science — and why it matters. But this event felt different. It was more like an interactive case study in open science unfolding in real time: collaborative work across domains, open-source code pushed to GitHub, outputs shared via preprints, and active consideration of licensing, reuse, and impact.
Photo courtesy of CMU Libraries
Participants left with more than just a hands-on experience. They walked away with something tangible to add to their CVs and portfolios — code they wrote, a preprint with their authorship, and a deeper, more personal understanding of what open science and interdisciplinary engagement truly requires. For some, it was a first step into research.
That event stayed with me because it captured something that’s hard to explain: how the practice of knowledge creation is changing. And how a format like this — time-boxed, focused, and collaborative — can offer a kind of experiential learning that students might not typically encounter in a classroom or lab environment.
Expanding the Hackathon Model
In recent years, our library has broadened the scope of the hackathon model, applying it to new questions and communities. One example from 2024 was a reproducibility hackathon focused on digital humanities, organized by Chasz Griego in collaboration with faculty from CMU’s Department of English. Participants weren’t dealing with abstract scenarios, they were working with real data and code from a published article, trying to replicate and extend the original research. The event sparked important conversations about infrastructure, bias, and the durability of scholarly methods. It wasn’t about inventing something new so much as stress-testing the systems we already rely on. What holds up? What’s missing? What needs to be rethought?
That same ethos shaped the 2024 AI Literacy Hackathon, which brought together participants from 17 institutions and was led by Emily Bongiovanni, Emma Slayton, and Nicky Agate. The focus was on co-creating open educational resources around AI ethics, usability, and societal implications. What stood out to me was the absence of competition. Instead of racing toward a finish line, participants worked together to develop infographics, exercises, and teaching tools — all openly licensed, so others could adapt and build on them. Check out the project site via OSF.
Photo courtesy of CMU Libraries
The CMU Libraries’ 2025 Reproducibility Hackathon pushed the model even further, inviting students to engage directly with research on large language models. Using real data and code from a study examining whether LLMs write like humans, participants didn’t just test reproducibility — they explored how style, syntax, and voice shift across different models and contexts.
In an article about the event, one doctoral student reflected on how much of his own statistical training wasn’t intuitive to others and how that realization reshaped his thinking around accessibility and reproducible design. Another participant, an alumnus working in industry, described the experience as a reconnection to something he missed: an academic community rooted in shared questions and open dialogue.
There’s an interesting throughline that runs across these events, one that reveals itself in the conversations. They unfold across disciplines, between critique and code, among people who might not otherwise find themselves in the same room. And they often go deeper than the task at hand. These hackathons aren’t just about tools or methods; they become moments of insight, and invitations to reconsider how we understand, evaluate, and communicate the systems we’re increasingly building our future on.
What ties all of this together isn’t just the subject matter — whether bioinformatics, digital humanities, or AI — or even the hackathon format itself. It is also the moment we are in: a world shaped by data, algorithms, and accelerating complexity. These events reflect that through time-bound collaboration, purposefully open-ended goals, and a shared sense of inquiry. It is not about speed. It’s about momentum — and what can happen when people come together to explore something that matters to them, or at least sparks their curiosity.
Libraries as Connectors & Conveners
Perhaps this post isn’t really about hackathons but rather part of a deeper, ongoing reflection, one that is central to how I see the role of libraries. Both practically and symbolically, libraries serve as connectors and conveners. We connect people to ideas, to each other, and to the challenges and opportunities that define our time.
From a library point of view, we are not just hosting these events — we are shaping them. Facilitating connections. Supporting the infrastructure. Creating the conditions for interdisciplinary work that doesn’t always happen on its own.
Here’s another example that shows how committed we are to this model: through our Open Source Program Office, the Libraries support student-led hackathons by offering microgrants, mentorship, and encouragement for events that explore open knowledge, civic tech, ethics, and more.
Despite the different topics, the hackathons we’re hosting all share a common approach: collaborative, time-bound, and deeply engaged, with an eye toward openness. Even zine-making workshops, with their hands-on, improvisational nature, feel like close cousins to this model. I wanted to bring them together here not simply as a recap, but as a reflection on what they represent.
Because this feels like yet another example of how the library is being meaningfully redefined — not just as a service point but as a catalyst, not just as a collection but as a platform. A space where knowledge is not only accessed, but questioned, shaped, shared, and practiced in real time with real collaborators.
I’d love to hear how others are evolving the hackathon or research sprint models — especially in ways that reflect your community’s needs, questions, and energy.
