I think technical debt comes from not cleaning up/refactoring once you’re done. Also, doing this work requires to have a good understanding of the codebase as w

I think technical debt comes from not cleaning up/refactoring once you’re done. Also, doing this work requires to have a good understanding of the codebase as well as its style/cohesion. I’m wondering if ML will get good enough to produce PR reviews that comment on style and refactoring opportunities to match the codebase’s spirit.