‹ BackHN Continuity

Thread

Will Open Source Survive the Agents That Replaced It?

28 points · 61 comments · albertoarena

Loading the complete thread in the background. This saved snapshot is available now. Refresh

  1. ed_elliott_asc · · focus · HN ↗
    AI is causing everything to change, small basically trivial oss projects are no longer really needed - that doesn’t seem too bad, we can focus on the more exciting things now!
    1. bcjdjsndon · · focus · HN ↗
      Why is this downvoted?
  2. someguyiguess · · focus · HN ↗
    I’ve often wondered the opposite. The number of pieces of software I used to pay for that I was able to replace with software I’ve built using agents is nearing double digits. I wonder how much longer people will need to pay huge software companies instead of just collaborate with a few other folks using AI and build their own better version.
    1. piker · · focus · HN ↗
      As a developer (it sounds like), you were probably never even a drop in the bucket dumped into the ocean of the proprietary software market. Most people don’t want to change their own oil.
      1. ChrisLTD · · focus · HN ↗
        Developers don’t even like to pay for their own code editors. Something they used to work in all day.
        1. bayindirh · · focus · HN ↗
          I'll personally pay (and do pay) for good development tools. I prefer free software tools primarily due to ethical and philosophical reasons (and I also want my code doesn't depend on any tool which can't be obtained in the future), but I have a couple of non-blocker, high quality, yet closed source tools in my arsenal.
      2. le-mark · · focus · HN ↗
        > Most people don’t want to change their own oil.

        Changing oil is easy, it’s disposing of the oil that’s hard and why I don’t change my own oil and haven’t in years.

        1. piker · · focus · HN ↗
          Sort of like maintaining your own text editor, yes
    2. glimshe · · focus · HN ↗
      Roughly zero.

      Not only you're making huge assumptions on AI's ability to build software useable by non-software engineers without professional input, you're also underestimating people's desire to pay just for convenience ("gee, I need a diet tracker, let me find a few people to help me vibe code it instead of paying $10")

      1. intrasight · · focus · HN ↗
        It's a valid assumption, and it is going to have a huge impact on the SaaS marketplace. "I need a resource tracking application. Let me find a consultant to help me vibe code it rather than pay $10,000 a month."
        1. bcjdjsndon · · focus · HN ↗
          Hiring a consultant to vibe code isnt the same as a non technical vibe coder, the whole point he's making.
    3. bayindirh · · focus · HN ↗
      I have paid a relatively big price for a software which does a simple thing: Drawing diagrams. The application? It's called OmniGraffle Pro. That simple looking program drawing diagrams can do flowcharts to scaled architectural or industrial diagrams and everything in between.

      Try replicating it with a couple of agents. Not the capabilities, but the performance, stability and dependability and sustainability as well.

      Same is true for my favorite text editor on macOS, BBEdit, or IDE of choice, Eclipse. I can increase the number of examples.

      These tools are valuable not because they are expensive. They are valuable, or indeed priceless, because of the polish they have. The small features make no sense for a beginner or how they cope when things get absurdly haywire.

      Immediate features make us excited and they are easier to replicate because they are visible, but things like performance, dependability, error recovery or the small tweaks they have gone through in the last decade are not, and this is what matters.

      We're exploring computing via AI again, and we act as that we did nothing in the last 60 years or so. Software and computing is much deeper than it looks at the surface, and we'll learn this in a very very painful way, again.

      1. jlkuester7 · · focus · HN ↗
        This. I happily pay for IntelliJ software (even though basically everything else I use is FLOSS). It is not really this feature or that feature of WebStorm that makes it worth while, but just that 40+ hours of the weeks it just works. For normal stuff, for edge cases, small projects, big repos. When I need something, it just works (with out me needing to fiddle with an agent to build me some brittle scripting that will kinda work today, but break tomorrow....).
    4. Forgeties79 · · focus · HN ↗
      You say this but can that software work on other machines that aren’t configured the exact way yours is?

      Will it survive the next OS update? How about 3?

  3. mikodin · · focus · HN ↗
    As I've been vibe coding a side project, I've been curious on just how much cruft and slop the agents are training on now. A part of me feels like I should actually have it be in a private repo.
  4. wongarsu · · focus · HN ↗
    I've noticed that for smaller projects (<=10k lines of code) the valuable thing is no longer source or binary artifacts, it's the encoded ideas, architecture and domain knowledge. A folder of plan files is about as valuable as the code itself, and easier to adapt to new domains

    I wonder if that will spawn a new type of open source project. Where we have free exchange of ideas, without necessarily requiring the free exchange of code (which would violate a lot of Stallmann's ideas about open source, but might ironically be attractive to people who dislike AI training on their code)

  5. cc62cf4a4f20 · · focus · HN ↗
    For one off tools, sure. For ongoing tools, either they are right after the first implementation and you use them for years. Alternatively, you need to maintain them and they decay until you discard them.

    You also need to be ok with your tool to do task A (you had an agent build) to work different from my tool to do task A (I had an agent build). Meaning we can’t reliably compare the output from those tools.

  6. strogonoff · · focus · HN ↗
    The ecosystem of open-source reusable components will remain a good idea as long as architectural principles like low coupling and high cohesion still apply—otherwise it’s wasting resources on reinventing wheels.
  7. 0xbadcafebee · · focus · HN ↗
    [delayed]
  8. zkmon · · focus · HN ↗
    Agents replaced open source? An Agent is an actor and open source is an artifact. An actor can't replace an artifact.

    Opensource is defined by it's licensing terms and the community around it.

    1. bakugo · · focus · HN ↗
      [delayed]
      1. bcjdjsndon · · focus · HN ↗
        > Open source communities are dying due to AI spam

        They're luddites refusing to get with the times

        1. 12ahs6 · · focus · HN ↗
          You have a history of attacking those that generously gave to the trillion dollar industries and depict them as selfish.

          You are no better than Goebbels or Stalin.

          1. bcjdjsndon · · focus · HN ↗
            You have a green username, you have no life
        2. canelonesdeverd · · focus · HN ↗
          What does "getting with the times even mean" at this point? Just bending over and taking on all the griefing and bullying?

          Btw luddites are the reason you're not slaving away at a factory.

    2. matheusmoreira · · focus · HN ↗
      Free and open source software communities were not always friendly, even at their peak... And now there are far too many projects that are outright hostile towards anyone who uses AI for any purpose. Very discouraging.
      1. bakugo · · focus · HN ↗
        [delayed]
        1. matheusmoreira · · focus · HN ↗
          I understand the hostility against AI itself, but people are reacting with prejudice against AI users. People are building blacklists of AI pushers and their tainted projects. This is not fine.
          1. Henchman21 · · focus · HN ↗
            It seems to me that when you choose to outsource your thinking to a machine, you cease to be a user and turn into a bit of a dreamer. The implication over time is that your cognitive skills will significantly deteriorate because you’ve decided to outsource your thinking. This is also not fine.

            I don’t like the whole “lists” thing, that sounds like literal insanity. But I feel like I can understand the reasoning behind it, even if I think it’s a step too far. “No AI contributions” is easily enough IMO. But then all this is just IMO :)

            1. matheusmoreira · · focus · HN ↗
              Yeah I agree. I can understand the reasons behind it, I just feel like there should be more nuance. Some people on lobsters called my hand written projects slop because I allowed AI to touch them and repeatedly told me to get out, and it only spiraled from there. At this point I'm not even sure if I should share my work anymore.

              > your cognitive skills will significantly deteriorate

              This is a fair concern. I for one am trying my best to distill the AI's weights into my own brain. My human limitations make this process slow, but I am trying.

              At the same time, I'm also vibecoding things I don't fully understand, simply because it'd be foolish to not spend the tokens I paid money for.

  9. addag · · focus · HN ↗
    Putting aside the bond I have towards with the FOSS community, I am curious of what will be the new equilibrium of this ecosystem.
  10. brightball · · focus · HN ↗
    It seems like the opportunity is for open source to explode thanks to AI. This week I replaced my home security camera system with open source Frigate, because I was able to use AI to take the source code, replace the object detection models that were included with the code with something that was a better fit for my machine, customize how the videos were stored and under what circumstances. There's a lot more that I can do with it that I haven't yet as well.

    I couldn't do that with the proprietary system.

    It seems like we'll see a lot more customized open source options or rapidly accelerated open source projects thanks to this type of thing. I'm really curious to see if something like LibreOffice stops playing catchup with Microsoft and actually passes it.

    If the improvement rate of Omarchy for Linux desktop experience can be a barometer for anything, I'm extremely optimistic.

    The #1 barrier for me to do 90% of the things that I've done for myself over the last 3 months for the prior 10-15 years has been time. I'm more than capable of downloading Frigate and figuring out how to customize everything, modify and compile the code to do what I want, but it would have taken me hours to days to do it. Now I can setup a couple of stacked /goal prompts overnight and get there. It's fantastic.

    1. SkiFire13 · · focus · HN ↗
      > This week I replaced my home security camera system with open source Frigate, because I was able to use AI to take the source code, replace the object detection models that were included with the code with something that was a better fit for my machine, customize how the videos were stored and under what circumstances. There's a lot more that I can do with it that I haven't yet as well. > > I couldn't do that with the proprietary system.

      You also didn't give anything to opensource. This shows that AI can allow open source to be used in more cases and in more flexible ways, but in no way does this mean that open source (and especially the good high quality one) will explode.

      If anything we saw an explosion of open source projects, most of them low quality slop, and even the ones that used to be high quality are slowly degrading.

      1. FabCH · · focus · HN ↗
        It shows opensource _usage_ will increase.
      2. jordanb · · focus · HN ↗
        This is exactly right. The gift economy that Open Source is part of is incompatible with the extraction economy AI represents.
        1. engineer_22 · · focus · HN ↗
          "Invert, always invert." Charlie Munger

          Open source is a side effect of an industry so extractive that it can dedicate some of its most talented workers to volunteer in non-competitive ventures. Open source culture is a tacit acknowledgement of a privileged social class that needs not lower itself to brutish competition for its dues, this class instead enjoys the patronage of wealthy benefactors, reminiscent of the enlightenment system of arts which was enabled by the pillaging of a continent. For all the treasures of Europe, hundreds of years later a hole of human deprivation remains bored through the heart of South America. The extremes of noble beneficence and savage exploitation are often found happy bed partners, once the covers are removed.

          AI is a reversal of those fortunes, it's democratization of compute. Now anyone literate can write instructions for a computer to do useful work. Certain US labs limiting foreign AI access wasn't a pause to build sufficient security tools to prevent proliferation of powerful op sec weapons (widely reported in government, industry and media), it was about throttling capabilities in poor countries where such software promises the biggest threat to western tech. It's the same reason the giants now murmer about slowing the frontier, they realize they've passed the optimum, any advancement now works against their power and advantage. The problem is, they're spurred forward by naive eastern labs that have never enjoyed the same global dominance. The giants realize the technology is so revolutionary that it will literally devours its masters, but they've made a Faustian bargain and can't help but play to the end.

        2. engineer_22 · · focus · HN ↗
          The gift economy is a patronage system akin to the Renaissance arts, enabled by maximal extractive leverage. AI represents a democratization of the human machine interface
      3. brightball · · focus · HN ↗
        That's valid. I didn't contribute anything back yet but I certainly could at the point that I'm convinced the refinements are polished enough.

        At the same time, how different is that from much of open source today. Usage is dramatically higher than contributions. How many open source packages are included in every software project we build with. How many of those are any of us contributing back to?

      4. potsandpans · · focus · HN ↗
        > You also didn't give anything to opensource.

        The prevailing moral panic prevents anyone in the space from sharing in communities where open source is generally discussed. People flag and openly disparage (against community guidelines) and have histrionic reactions.

        We'll just have to wait until the topic cools off before we can see how llm usage impacts open source.

  11. timr · · focus · HN ↗
    Ironically, this article has so many signs of being written by an LLM that I couldn’t finish it.
    1. Krei-se · · focus · HN ↗
      100% the smoking gun. It's not the thought that counts, it's the typing you save. You don`t need to read the text, it's not enough. You need to digest the slop.
    2. pbronez · · focus · HN ↗
      I generally don’t bother to complain about claudism but YES. There’s a special irony to using a tool to complain about how the tool is breaking things.
  12. Krei-se · · focus · HN ↗
    "Maybe what’s actually happening here isn’t death or survival, it’s a change of job"

    AI sentence structure spills everywhere

    1. bcjdjsndon · · focus · HN ↗
      Surely this is an artefact of the training data?
      1. Krei-se · · focus · HN ↗
        it's not an artifact, it's thinking out loud guardrails
        1. albertoarena · · focus · HN ↗
          Exactly
  13. enclave402 · · focus · HN ↗

    [dead]

  14. 1298ahg · · focus · HN ↗
    Nothing has been replaced. This is just propaganda by a consultant who explicitly is for hire for AI integrations. He has probably never done anything in open source.

    He uses open source and sells it out for cheap marketing. Clients take note.

    1. albertoarena · · focus · HN ↗
      Read author's profile linked in the blog. I do contribute to open source.
  15. JV00 · · focus · HN ↗
    I think there will be a difference between foundational software and application software. The more you go to the outer layer the more code will be generated on the spot, throw-away, single usage. On the other extreme, foundational software that is battle tested, security hardened, has survived waves of AI-hacking, will become irreplaceable and crystallise as a semi-immutable layer. Open source is in the best position to win foundational software while agents will dominate the outer layer of disposable code.
  16. jchw · · focus · HN ↗
    Why is this on the frontpage again? It's slop writing with a blatantly untrue borderline ragebait premise. Can we please stop this madness? Hacker News doesn't have a downvote feature (for stories), and that absolutely does have its pros and cons, but this needs to stop. This is basically SPAM.
    1. 129sjhf · · focus · HN ↗
      The sheep here are trained daily to automatically bleat the talking points in their elevator pitches for the 12875th vibe coding startup.
  17. VCFundedGenYer · · focus · HN ↗
    What?

    Author has no idea what open source even means.

    1. albertoarena · · focus · HN ↗
      Have you checked my Github profile as author of the blog?
  18. sehw · · focus · HN ↗
    Fuck intellectual property and shit. LLMs fuck yeah!
  19. dwheeler · · focus · HN ↗
    False premise. AIs dont generate anything in 30 seconds that has been widely reviewed and tested.

    If you want it to be corrrct, it's always easier to install some existing OSS and instead focus on whatever is unique to you.

  20. mohd_rafay · · focus · HN ↗

    [dead]

  21. eggbrain · · focus · HN ↗
    This is the biggest opportunity in open source I've seen in my lifetime. I bet you in three years we'll have more and better open source libraries than ever before.

    For example, a few months ago I was looking to build a license detection tool, and within a few weeks with the help of Claude built a tool that was 20x better than the best open source one I found (I also open sourced mine as well: <a href="https:&#x2F;&#x2F;github.com&#x2F;licensedetector&#x2F;cli" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;licensedetector&#x2F;cli).

    The biggest problem will be navigating through the glut of libraries that have been written in my mind -- 100 tools for the same use case, all with &lt; 100 stars will make it hard to decide between. Also, if the user didn&#x27;t write them thinking about anything other than their own use cases when releasing their library, the standard response will probably be for the agent or programmer to recreate the wheel yet again, only for their specific use case, making it now 101 tools for the same use case.

    1. albertoarena · · focus · HN ↗
      I&#x27;m also trying to build as much as I can in open source, and I feel productive. But, as I mentioned in the article I wrote, it&#x27;s not nice to see how AI is often reinventing the wheel.
  22. MiscIdeaMaker99 · · focus · HN ↗
    What does it mean for open source to survive? Exactly what software are we talking about here?

    I don&#x27;t know about everyone else&#x27;s experiences, but the AI agent that I use every day at work would rather install a tool or library than spend tokens recreating software that already exists out there. As such, the argument doesn&#x27;t make much sense to me, personally.

    1. mugul · · focus · HN ↗
      Agree with you on that one, and if the agent somehow tries to re-invent the wheel I will definitely remove the code and do the install myself. I have way more trust for OSS projects maintained by motivated humans (who can be using coding agents if they want, I don&#x27;t care, as long as project quality remains high).
  23. Kuyawa · · focus · HN ↗
    I told AI to create and editor just like Sublime (my editor for the last decade) and it did in less than 10 minutes. File tree, multi tabs, multi carets, syntax highlighting, code snippets, everything I need in an editor (except LSP for now, will be added in v2)

    I haven&#x27;t used Sublime this week, it works as expected and I can&#x27;t be happier

    Oh, and it cost only $2 to build using DeepSeek

    Then I told DeepSeek Harness to build a clone of itself in Swift, unbelievably it did!

    So, apps are in a downward spiral towards zero value, everybody can do their own. The real value is in market penetration and the race is going to be fierce

  24. alsanan · · focus · HN ↗
    Agents replacing today open-source doesn&#x27;t means they can do the same in the future. In first place we must all understand that AI is being feeded with existing projects. If we stop creating new projects, may be AI won&#x27;t have enough base to do the same and the open-source would necessary.
  25. skyzhao1223 · · focus · HN ↗

    [dead]

  26. tpoacher · · focus · HN ↗
    Your fallacy is the fallacy of the loaded question.

    If anything open source is proving to be the last bastion of hope against the trend of being mindlessly replaced by agents.

  27. flashgordon · · focus · HN ↗
    I would say the bigger worry should be for &quot;non open source&quot;. Where I work we buy so much &quot;vendor crap&quot; because they enter through one area of expertise (eg CRM) and upsell into a bunch of other crap they have no expertise in but purely via acquisitions they have done - I am talking about an imaginary vendor that rhymes with &quot;DalesPorce&quot;. We have even demonstrated we are able to replicate their meagre (but highly marketed) offerings with our own high quality alternatives with much less cruft by using AI and a combination of OSS (especially by improving the OSS in the process).

    The usual argument goes &quot;oh but now you have to maintain it&quot;. Guess what - we have dedicated &quot;DalesPorce&quot; engineering teams just to maintain the expensive and buggy&#x2F;unreliable DalesPorce setups - completely negating the so called &quot;you can be fully hands off&quot; value they promise. Even worse when we measure productivity of these dev teams - Teams on the DalesPorce stacks are 1&#x2F;5th to 1&#x2F;20th as productive as teams on the boring old stacks.

    My 2c is that this is a great opportunity for OSS. Instead of just building basic early stage offerings they should help help companies embrace the &quot;commodotize the complement&quot; strategy. Not just for the main offering but also in offering uber easy to setup and manage control plane services over them.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.