RIP software hackathons. Long live the hardware hackathon
Hackathons are increasingly seen as polished pitch contests dominated by AI-generated “slop” and mock UIs, rather than intense sessions of hands-on coding and experimentation. Commenters contrast today’s corporate, presentation-first events with earlier, more collaborative or technically driven gatherings, and argue that hardware-focused projects now offer more tangible challenge, authenticity, and resistance to AI automation. Others counter that both software and hardware hackathons still have value for learning, creativity, and rapid prototyping, but only if incentives move away from free labor for sponsors and toward genuine making.
State of Software Hackathons
- Many describe modern software hackathons as pitch or UI competitions: polished mockups, landing pages, or even pure PowerPoints often beat technically difficult projects.
- Judging is seen as biased toward slick design, storytelling, and “vision” rather than implementation depth or robustness.
- Several participants report internal corporate hackathons where half-finished or pre-existing projects win, and management then tries to productize them quickly or add them to teams’ workloads.
- Some view this as exploitation or “mandatory fun”: free ideas and free labor under social pressure to participate.
- Others still find value: practicing pitching, communication, and storytelling, or enjoying the social and learning aspects.
Hardware Hackathons and Tangibility
- Many are enthusiastic about hardware-focused projects: they’re tangible, harder to fake, and satisfying to hold and demo.
- Typical “hardware hackathon” work is described as wiring sensors/actuators to Arduinos/Raspberry Pis with simple circuits, not full custom PCBs.
- Some argue that calling this “hardware” is misleading, as it’s primarily software on resource-constrained devices with light assembly.
- There’s optimism that cheap microcontrollers, 3D printers, and accessible tooling will enable more weekend hardware projects that become real products.
AI, “Vibecoding,” and Quality
- Several comments agree that AI-assisted “vibecoding” now dominates: LLMs can quickly generate plausible code and UIs, so hackathon difficulty shifts from coding to idea selection and presentation.
- Some celebrate this, predicting a “golden age” where weekend hackathon projects can be nearly production-ready.
- Others push back: software is not “solved”; robustness, safety, and elegance remain hard, and AI-generated output is often “slop” with broken backends or shallow functionality.
- AI-generated branding and food photos for small businesses are widely criticized as soulless and off-putting compared to imperfect human work.
AI and Hardware
- One side claims hardware is the new safe frontier because AI “can’t solder, cut enclosures, or debug analog issues.”
- Others counter that, combined with existing assembly machines, cameras, test equipment, and 3D printers, AI can increasingly participate in hardware workflows too—though not yet at hackathon-level cost.
- Safety concerns are raised about AI-designed physical systems (e.g., drones), where failures can have serious real-world consequences.
Nostalgia and Alternatives
- Multiple comments express nostalgia for early/grassroots hackathons and dev sprints focused on collaborative free software, game jams, or civic tech with no judging or prizes.
- Some suggest returning to non-competitive “build together” events, where the only goal is progress on a project, not winning.