‹ BackHN Continuity

Thread

What even is an OS now?

310 points · 479 comments · fratellobigio

  1. Xirdus · · focus · HN ↗
    In my opinion, this post misses the forest for the trees. Yes, the model of an OS running various off-the-shelf apps for various tasks will soon become obsolete. But it's not the idea of the OS that's outdated - it's the idea of apps. Why would you ever want to ask an AI to make an app for you to complete some task, if you can instead ask the AI to complete the task directly? The future isn't a bunch of extremely personalized apps - the future is a single AI assistant doing all the work.
    1. zahlman · · focus · HN ↗
      > Why would you ever want to ask an AI to make an app for you to complete some task, if you can instead ask the AI to complete the task directly?

      Because there are limits to the ease that "asking the AI" permits. As a trivial example, no matter how good the AI is at doing stuff, I will always be able to type `ls` faster than I can ask it "what's in the current folder?".

      Because it's deterministic once written.

      Because it's vastly more efficient for my use cases, and I actually value that.

      Because I like the idea of making things. (It's frankly bizarre to me that this needs to be explained on a site called "Hacker News".)

      1. Xirdus · · focus · HN ↗
        > I will always be able to type `ls` faster than I can ask it "what's in the current folder?".

        A folder is itself becoming a relic of the past. Most phone users nowadays aren't even aware of the concept.

        > Because it's vastly more efficient for my use cases

        Care to elaborate which use cases they are that benefit from AI code generation but not from AI agent? I assume it's something you do several times per day?

        > Because I like the idea of making things.

        I'd think you wouldn't ask the AI to do it for you then?

        1. zahlman · · focus · HN ↗
          > A folder is itself becoming a relic of the past. Most phone users nowadays aren't even aware of the concept.

          And this is an actively bad thing.

          > Care to elaborate which use cases they are that benefit from AI code generation but not from AI agent?

          Making reusable things is a goal in itself.

          > I'd think you wouldn't ask the AI to do it for you then?

          Design is also an interesting task.

          1. Xirdus · · focus · HN ↗
            It feels like you want to generate code for the sake of generating code, with code generation being the goal in and of itself regardless of applicability. In that case, yeah, sticking with code generation is the best way to get code generation. Kinda like crotcheting, which is a pretty bad way of getting clothes in 21st century, but a lot of people do it anyway simply because they like crotcheting.

            Limits of asking AI, `ls` being faster, determinism, efficiency, and the idea of making things are all pointless and irrelevant. None of these matter. What matters is that you get to use code generators. And to that I say - fair enough, but please don't muddy up the discussion by pretending there's anything more to it.

            1. zahlman · · focus · HN ↗
              This is bizarrely dismissive. Why don't these things matter? And how did you come to that conclusion about what I want? That is not at all what I want.

              There is no "pretending" going on here and it is insulting for you to suggest it.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.