Hot take: good teams have always had this problem. Great teams learn to solve it. Back in the days before AI, I would regularly spend weekends on yolo PRs changing thousands of lines of code. The goal wasn’t to merge. I just wanted to get a feel for what the changes would look like. Tons of reasons for this: - exploring architecture - figuring out blast radius of changes - getting early product feedback - feeling out different flows, APIs etc The key to making this work: assume all PRs will…