What's Still Worth Sharing as Open Source?

Before coding agents, there were strong incentives to build software in the open. Execution was the bottleneck, and ambitious projects were only feasible if many people worked on them together. Open source was the mechanism to share ideas and get others involved. And because the software was meant to be shared, it had to work for many people across a wide range of use cases.

With agents, this has changed. At Aonyx, we're often building our own tools instead of adapting to existing open source ones. Agents can execute as fast as we can steer them, and a single developer can now build applications and tools that used to take a team or a community. Because we need to collaborate less on creating software, we can tailor it to our own needs. That, in turn, makes it less valuable to share, because it’s purpose-built for our own use case.

I’m wondering what this means for open source. If there’s less reason to build software in the open, and less value in adopting what others have built, what are we left with?

I’d love to hear where you’ve stopped reaching for existing projects, and what you would never build yourself.