> An outsider may not understand the vision and direction of your project. They may not have the same habits and norms for testing, code quality, and style. However, your own team members should share these norms.
Yeah, no. The challenge of running a company is that even within your company, everyone becomes misaligned the moment you’ve aligned. Have you hired anyone new? They don’t understand the company’s vision, nor have internalized your engineering standards. And even once they’ve worked with you for years, they still won’t have aligned on the things that they never ran into, AND they’ll misalign on all the things they've developed opinions on.
> Doing this means that a human only spends time reviewing code that has already been proven technically correct.
Yes, CI is useful for ensuring technical correctness. But technical correctness is one very small aspect of reviewing, and usually not where the big issues come in day to day.
Getting eyes on code is not about preempting things that would fail CI anyway. You probably should be running your full CI suite on PRs, too. To me it’s about catching misalignments on what the product should do, broader misalignments on how it should be engineered, catching misalignments that someone who’s too deep in the weeds would miss, talking through issues realized after the requirements had already been set and discussed, and more. A CI pipeline can’t fix that, pair programming will only address the things that an engineering peer who is as equally in the weeds as the contributor is, and integrating the newest things to production without human review is nice if your product is sufficiently low stakes that subjecting users to the effects of these people problems is okay; but that’s not for me and my team.
Pushing faster is not a pure good without trade offs.