git blame --ai: The Commit That Shaped (or Broke) Open Source

A year ago, AI in open source was mostly a cool side tool. Now it’s everywhere, and the community is split on it.

On one side, people write code twice as fast, beginners are sending their first pull requests, and AI is finding bugs that sat hidden for decades. On the other side, maintainers are buried under junk PRs, one AI agent even wrote an attack post about a maintainer who rejected its code, and people are rebuilding whole apps and changing their licenses in days.

In this talk I’ll go through what’s actually happening: the good, the messy and the legally weird. We’ll look at how projects like Linux, Gentoo and QEMU are dealing with it, and what happens to copyright when anyone can clone anything.

This is more of a community discussion so everyone will have different takeaways from it . But here are things we will talk about

The Upside: AI lowers barriers, speeds up coding.

The down side : Junk PRs are drowning maintainers.

Morals : AI clones and relicenses code, legally murky.

View on it : Projects range from banning AI to embracing it.

What to do with it : Finding bugs is cheap, fixing isn’t.