How to Kill the Code Review

(latent.space)

14 points | by tonkkatonka 19 hours ago ago

4 comments

  • compacct27 13 hours ago ago

    This isn't it. After spending my morning on a code review and catching several major issues despite it being a +7kloc diff, the real question is--how do we make the tooling around code reviews better? Intentions aren't enough, they're vague and the implementation details matter. They help, though, and this is closer to the answer to code reviews: check out the intentions, get an AI-assisted overview of the architecture of the code, then dive into the more important parts of the architecture.

    Maybe it's because I'm in UI dev, but intentions aren't enough at all

  • thomascountz 11 hours ago ago

    A code review is an engineering practice and a knowledge management process; it should not be confused with a merge or pull request, which is a change management and version control process. When you say, “code review wasn’t even ubiquitous until around 2012–2014…,” I think you’re referring to the approval tooling built into merge requests, not to code review itself. Engineers were having their code reviewed long before then.

    By the time you genuinely expect a PR to be merged, it should be essentially rubber-stampable, in my opinion. It shouldn’t be the first time someone else is looking at your code—let alone the first time anyone is reviewing your design.

  • swyx 19 hours ago ago

    just got pinged by f5bot - hi editor here! Personal take: at this point multiple people are already weighing how to remove the human code review bottleneck from agents becoming fully productive. Ankit was brave enough to map out how he sees SDLC being turned on its head and wrote this.

    i'm not personally there yet, but I tend to be 3-6 months behind these people and yeah its definitely coming.

  • tonkkatonka 19 hours ago ago

    Human-written code died in 2025. Code reviews will die in 2026.