• If you think vibe-coding clones of commercial application qualifies as open-source, you’ll be in for a rude awakening very soon.

    My prediction is: when enough high-value applications are cloned and working as well as the originals, the AI companies that made the coding agents will be buried under copyright lawsuits. Not the talentless monkeys who told agents to reverse engineer the applications, no no: the AI companies.

    And with good reason: AI doesn’t know how to reverse-engineer and code clones of applications out of the blue. It’s been trained to do that, and since it almost certainly was trained on stolen, proprietary data - or at least it’s impossible to prove that it wasn’t - none of the shit it makes is open-source by any stretch of the imagination.

    This will be challenged. You can take that to Kalshi and Polymarket.

          • Neither of us can prove AI code is clean. That’s precisely why it can’t be open-sourced. Or close-sourced under any restrictive license either for that matter, because it almost certainly contains GPL code too. It goes both ways.

            • AHemlocksLie ( AHemlocksLie@lemmy.zip ) 
              link
              fedilink
              English
              arrow-up
              5
              ·
              6 days ago

              Legally, I’m not sure that’s the case for that argument. My understanding is that if you want to win a copyright claim, you need to prove a violation occurred, not just point out that they can’t prove no violation occurred at all. AI has other issues for copyright, like I don’t think heavily vibe coded works are subject to copyright because there’s no real author to pin it on.

              • Plaintiffs can claim a violation occurred, and the fun will begin when discovery compels the defendants - the AI companies - to disclose what their models were trained on.

                As to “there’s no real author to pin it on”, that doesn’t hold water legally: I can’t put a brick on my car’s accelerator, let the car slam into a building, then claim I wasn’t responsible for what the car did. Similarly, Waymo is held responsible for what their autonomous vehicles do on public roads.

                • AHemlocksLie ( AHemlocksLie@lemmy.zip ) 
                  link
                  fedilink
                  English
                  arrow-up
                  3
                  ·
                  6 days ago

                  Even having been trained on copyrighted data won’t necessarily prove that something produced by AI actually used copyrighted code. That might get the AI companies in deep shit, but I’m not sure it’ll cause blanket problems for all AI generated code.

                  As for responsibility, that’s a very different thing from copyright ownership. I understand the idea as I proposed it isn’t intuitive, but this isn’t a fair comparison.

    • MrSoup ( MrSoup@lemmy.zip ) 
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      6 days ago

      If code is available to the open, to be read or copy, then it is open source. Even if it’s slop.

      No one will go after random dudes doing their apps with LLMs. Even if those are ruled that doesn’t apply to a copyleft license, it is still open source for providing code.

      If the code is really derived from a reverse engineering (so not what is branded as clean-room), then the original developer could guess if it is its code by the logic and structure (since stuff gets renamed) and demand a take down or a DMCA.

      But forks and mirrors would keep the code online.
      This is, even for slop, a win for open source since people can read code for research purposes or even reuse it for other projects.

      Mind how I use open source and not foss, as those are different things like you’re pointing out about licensing issues.

      • If code is available to then open, to be read or copy, then it is open source. Even if it’s slop.

        If the code violates copyright, or is stolen, or is otherwise sketchy, it’s not yours to slap an open-source license onto and offer, is my point.

        • MrSoup ( MrSoup@lemmy.zip ) 
          link
          fedilink
          English
          arrow-up
          3
          ·
          edit-2
          6 days ago

          That’s why we have plagiarism machines.
          Scramble and renaming stuff has never been so easy and poof, code is yours now.

          This can be seen (because it is) as immoral, but this is the reality we got. All we can do is using something immoral to achieve a public good which it could be a open source code available to all.

          Corpos are going to use it this way anyway, using open source code to build closed ones.

    • Laws around AI right now are such that anything AI produces is not copyrightable.

      Companies that push it out can be sued but the plaintiff won’t be able to prevail without a way to pierce the non-copyright framework.

      And since the company’s that would be suing benefit from the framework, I suspect it will be a settlement to purchase the code and then remove it from repositories.

      • Laws around AI right now are such that anything AI produces is not copyrightable.

        That’s not the issue. The problem here is that the article claims “everything is open-source”. But it isn’t, because AI code is tainted: it’s impossible to prove it wasn’t trained on copyrighted data (EXISTING, PRE-AI copyrighted data) and it didn’t use it, copy it or adapt it to create the vibe-coded slop. Tainted code can’t be released as open-source.

        • vagrancyand ( vagrancyand@sh.itjust.works ) 
          link
          fedilink
          English
          arrow-up
          2
          ·
          6 days ago

          Non-copyrighted code, which is all code produced by AI by law, is defacto open to being Open Sourced, since there’s no other option.

          It doesn’t matter what it was trained on. The law, right now, as has been proven repeatedly in court, is that all AI-produced products are not able to be copyrighted, but are legal to be generated, even when they are clearly trained on copyrighted data.

          It would require a new law to stop it, and if that happens then all AI goes away and all previously released ai-produced code would still be legal since you cannot punish people retroactively.

    • ImgurRefugee114 ( ImgurRefugee114@reddthat.com ) 
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      6 days ago

      AI doesn’t know how to reverse-engineer and code clones of applications out of the blue. It’s been trained to do that, and since it almost certainly was trained on stolen, proprietary data

      Absolutely not. That’s not how any of this works. An LLM can’t be trained on data they don’t have. If it worked how you’re saying it worked, they would have needed access to every proprietary source on the planet, including proprietary sources that don’t exist anymore because the original source has been lost. The vast majority of closed source proprietary programs, including cloud-based apps, can’t possibly have been trained on. While LLM fundamentals and have a basis in compression, they are not merely regurgitation machines.

      You can teach a model to generalize, extrapolate, and pattern-march reverse engineering practices using open source code and debugging work. Native compilation is a lossy process; the original source code is irrecoverable (in most cases) and is likely subject to copyright. But you can write new code that approximates the behavior; that generates similar enough or identical native machine code, while being novel and even separately copyrightable in some cases. Such is cleanroom implementation and black box reverse engineering (very different things, legally speaking; tho worth noting that RE is a protected right in most countries [complications come from DMCA digital locks]).

      You can test this yourself. Write completely novel code by hand that does something stupid; something no one else has ever written before. Move the binary to a new machine, give a frontier model some tools, and tell it to reverse engineer it to the same language. Layers of abstraction and higher level concepts will be lost, but it will be able to recreate the underlying behavior, to within some tolerance, with a new source code.


      That all addresses proprietary data cloning… But even “they were trained to clone apps” isn’t a given. I can’t prove that they weren’t, and perhaps they are (would be easy to train using synthetic and open data); but that needn’t be the case. If you’re good at debugging, you’re good at reverse engineering; at least to some degree competency. If you can read, you can write. It doesn’t mean you’re the best novelist in the world, but there is significant overlap inherent to the capability. And they’ve definitely been trained to debug.