AI Pulse by Inblix

OpenAI's first hackathon draws 500 RSVPs for 100 seats

OpenAI Blog · Jul 20, 2026 · 2 min read · Read original article →

Curated by the Inblix editorial team


Featured image for article: OpenAI's first hackathon draws 500 RSVPs for 100 seats

OpenAI held its first public hackathon on March 3rd, pulling together a hundred developers, researchers, and students from a pool of over 500 applicants who responded within two days. The speed of that response tells you something about the gravitational pull OpenAI now has in the broader AI community—and the hunger for spaces where people can build alongside the researchers pushing the field forward.

The day began with talks that felt less like corporate keynotes and more like a research lab opening its doors. Sam Altman took questions on AGI timelines and safety, describing his personal pivot to the field because it was, in his view, an underfunded area with the potential to affect everyone. Ilya Sutskever covered self-play with RL agents, while Alec Radford walked through the ever-expanding taxonomy of GANs. Josh Achiam gave an RL introduction; the code and slides are now public.

Then came the eight-hour code sprint. The projects were refreshingly concrete, sidestepping the breathless futurism that often clings to these events. One team built a classifier to identify purine-rich seafoods that gout patients should avoid. Another implemented PPO with curiosity-based intrinsic rewards to train an agent that smashes block towers. A pair of high schoolers tackled Natural Q-Learning based on one of OpenAI’s own research requests. An attendee called Andy Matuschak developed the “scrying pen,” a machine-assisted sketching tool built on SketchRNNs. Cirrascale provided the GPU machines that made the sprint possible.

What stood out wasn’t just the output but the mix of people producing it. OpenAI explicitly sought gender balance and pulled from a range that included high schoolers, nonprofit engineers, and industry practitioners. Several attendees said that balance changed the texture of the day. With the Scholars and Fellows programs now in the background, the hackathon reads less like a one-off and more like an early move in a longer recruitment strategy—one that bets on giving people a weekend with the tools and seeing what they build.

💡 Key Takeaways

  1. OpenAI's first hackathon attracted 500 RSVPs in 48 hours, signaling intense demand for hands-on access to the organization's research community.
  2. Sam Altman described AI safety as an underfunded area with broad impact, offering a rare personal window into how he prioritizes the lab's direction.
  3. The eight-hour sprint produced projects that were strikingly practical, including a gout-related food classifier and a curiosity-driven RL agent, rather than abstract demos.
  4. Attendees spanned high schoolers to industry engineers, and the deliberate push for gender balance was called out by participants as shaping a better event experience.

Keep reading: See related articles below for more coverage on this topic.

Get smarter about AI

The sharpest AI news, curated daily. Delivered free to your inbox.

Learn more

Glossary terms

← Back to all articles