A terminal — basically a text window developers use to talk to their computers — called Warp just opened up all its code for anyone to see and use. That's mildly interesting on its own. What's actually fascinating is how they build it.
Instead of hiring more programmers to send in code changes, Warp asks people to describe ideas and write out what they want. Then an AI system they built — called Oz — does the actual coding, planning, and testing. Humans come back at the end to check whether it worked.
Nearly a million developers use Warp already. And now they're offering this same "humans describe, AI builds, humans verify" model to other open-source projects too.
This is one of the first real examples of a well-known product being maintained primarily by AI agents, not human programmers typing away.
We're moving toward a world where the bottleneck isn't "do we have enough developers" — it's "do we have people who can describe what they want clearly enough for an agent to build it." That's a different skill. And it's one your team might already have.
Open-source — When a company makes all its code public, free for anyone to read, use, or improve. Like publishing your recipe instead of keeping it secret.
AI agent — An AI that doesn't just answer questions, but takes actions — searching, writing, testing — on its own, step by step.
Terminal — A plain text window developers use to give instructions to a computer directly, without clicking around a visual interface.
AGPL — A type of open-source license. It lets you use the code freely, but if you build something with it, you have to share your changes publicly too.
Something worth sitting with: if an AI can take a written description and turn it into working software — what kinds of things in your business could you describe clearly enough to hand off?