I liked programming a lot, but with AI and how my company is basically requiring us to use it for AI-first development I’m not enjoying my job anymore…

It feels like forced effort on my part to write design documentation and refine it so the LLM does it the way I want and having to review the prompt continuously opposed to actually solving problems the way I would. Plus things aren’t even supposed to be reviewed until an AI review of the AI code is done… I just feel like I’m supposed to be a rubber stamp on the output. They say that dev/review skills won’t atrophy because you’re always reviewing the code, but I can’t accept that conclusion. I became a better developer by actually problem solving, not reading code to see if it makes sense. It’s all about getting more done and trusting it at a conceptual level now, but not really understanding what’s really going on underneath the AI-gen docs because it’s “so much better than human-generated code”.

I don’t understand or really handle that approach well. I don’t really program outside of work because I want to just hang out with my family after it and it feels like a chore to do my job anymore. I just don’t know how to make myself enjoy that type of workflow they are trying to force onto everyone. Any ideas on how to find something to enjoy with that?

  • Serinus@lemmy.world
    link
    fedilink
    arrow-up
    3
    arrow-down
    1
    ·
    19 hours ago

    I was never particularly proud of how much Stack Overflow I had memorized anyway. That wasn’t the fun part of my job.

    Most of these companies pushing AI so they can show half their devs the door are insane. You can’t just vibe code shit and expect results other than a full airline ground stop and Microslop Office being down worldwide.

    The AI is pretty worthless without me. I have zero concerns about being replaced by a project manager. Companies are being sold something else, but they’re gonna figure it out pretty quickly. (Though not quickly enough for some of us. Negotiate to keep your tenure and severance when they bring you back .)

    But AI as a tool is here to stay, and it’s better to build those skills earlier than later. The Riemann hypothesis shows a bit of a model for future work, both good and bad. You can’t just tell AI “solve the Riemann hypothesis”. It would have gotten nowhere without the expert guiding it and using it as a tool. Fortunately most of the stuff we’re doing isn’t cutting edge research, so hopefully we don’t have to deal with a pile of slop that we can prove is correct, but nobody understands how.

      • rwdf@lemmy.world
        link
        fedilink
        arrow-up
        1
        ·
        2 hours ago

        Use it to fill in the blanks or draw by numbers. Set up your own scaffolding or better yet, a model implementation. Have the model define skills and guidelines based on what you have already programmed. Then use the AI to write new implementations of similar things. See what it can and can’t do. See where it goes off the rails and watch it introduce unnecessary complexity every time the instructions are ambiguous/imprecise.

        Using AI with carte blanche on a greenfield project without any preexisting docs, architecture definitions, established code quality and style guides, etc is a recipe for disaster.

      • Beej Jorgensen@lemmy.sdf.org
        link
        fedilink
        arrow-up
        4
        ·
        16 hours ago

        A couple ideas here:

        • Keep coding your own stuff on the side.
        • Don’t let the AI give you code that you don’t understand.
        • Understand the greater architecture of what it’s building.
        • Be specific in your instructions.

        Yesterday I was using Copilot and Claude to write a clone of the old trn Usenet newsreader in Rust. Got pretty far in about 7 hours. 8300 lines of code and 8300 words of design docs in Markdown. The architecture is pretty good, and the thing works. (Still missing is the thread graph display and posting, but I know those can be added without breaking anything.)

        Don’t get me wrong: if there were a button that destroyed LLMs forever, I’d have already pressed it. I have other projects that I write by hand even though LLMs could definitely do it because I like the challenge and I want to stay sharp. Reading code doesn’t do it.

      • Serinus@lemmy.world
        link
        fedilink
        arrow-up
        4
        ·
        19 hours ago

        Yeah, I don’t have a good answer for that. I’ve seen how easy it is to get lazy and just vibe with whatever it’s putting out, in others and myself (sometimes).

        The worst is when it’s clearly not working and the person just spins the AI in circles with “well, fix it” over and over again.

        I do think we have a history of looking for black and white answers that has bitten us in the ass. “Not invented here syndrome” led to the npm dependency hell, for example. There’s a middle ground. log4net and whatnot are great third party tools that should absolutely not be reinvented yourself. But maybe you should have handled Leftpad on your own.

        • Supercrunchy@programming.dev
          link
          fedilink
          arrow-up
          1
          ·
          4 hours ago

          The worst is when someone hands you some code or some document or reccomendation and they can’t answer any question about it, because they haven’t even read what they sent you. It just infuriates me beyond anything.

          I fully lost confidence in some of my colleagues that way. I feel they can’t think by themselves anymore. Like why do they think someone would pay them to just fully delegate the tasks to claude? Do I look so stupid that I can’t ask claude for the same thing? At least I can ask claude why it did some of the choices and get an half-correct answer back, unlike them.

          Why am I doing a code review if you didn’t bother even reading your own fucking PR???

          </rant>