alt Hacker NewsAI is destroying Open Source, and it's not even good yet
Comments
I feel like we are talking past each other.
1. I write hobby code all the time. I've basically stopped writing these by hand and now use an LLM for most of these tasks. I don't think anyone is opposed to it. I had zero users before and I still have zero users. And that is ok.
2. There are actual free and open source projects that I use. Sometimes I find a paper cut or something that I think could be done better. I usually have no clue where to begin. I am not sure if it even is a defect most of the time. Could it be intentional? I don't know. Best I can do is reach out and ask. This is where the friction begins. Nobody bangs out perfect code on first attempt but usually maintainers are kind to newcomers because who knows maybe one of those newcomers could become one of the maintainers one day. "Not everyone can become a great artist, but a great artist can come from anywhere."
LLM changed that. The newcomers are more like Linguini than Remy. What's the point in mentoring someone who doesn't read what you write and merely feeds it into a text box for a next token predictor to do the work. To continue the analogy from the Disney Pixar movie Ratatouille, we need enthusiastic contributors like Remy, who want to learn how things work and care about the details. Most people are not like that. There is too much going on every day and it is simply not possible to go in depth about everything. We must pick our battles.
I almost forgot what I was trying to say. The bottom line is, if you are doing your own thing like I am, LLM is great. However, I would request everyone to have empathy and not spread our diarrhea into other people's kitchens.
If it wasn't an LLM, you wouldn't simply open a pull request without checking first with the maintainers, right?
Reviewing code was also a big bottleneck. With lot more untested code where authors don't care about reviewing their own code it will take even more toll on open source maintainers. Code quality between side projects and open source projects are different. Ensuring good code quality enables long term maintenance for open source projects that have to support the feature through the years as a compatibility promise.
This weekend, I found an issue with Microsoft's new Golang version of sqlcmd. Ran Claude code, fixed the issue, which I wouldn't have done if agent stuff did not exist. The fix was contributed back to the project.
I think it is about who is contributing, intention, and various other nuances. I would still say it is net good for the ecosystem.
Isn’t it also destroying the internet with low-quality content and affecting content creation in general? Can LLMs still rely on data from the open internet for training?
> From what I've seen, models have hit a plateau where code generation is pretty good...
> But it's not improving like it did the past few years.
As opposed to... what? The past few months? Has AI progress so broken our minds as to make us stop believing in the concept of time?
Internet was a fun place … until they turned into s.. with ads all over. Social media destroyed it.
AI is killing creativity and human collaboration; those long nights spent having pizza and coffee while debugging that stubborn issue or implementing yet another 3D engine… now it is all extremely boring.
At least for me personal open source project[1], it has been >5x boost. In speed, motivation. Operating knowledge level etc. At some places, I even put inline comment, "this generated function is not understood completely" ! Or may be a question on specific syntex (c++20).
Are there maintainers of mature open source projects that can share their AI coding workflow?
The bias in AI coding discussions heavily skews greenfield. But I want to hear more from maintainers. By their nature they’re more conservative and care about balancing more varied constraints (security, performance, portability, code quality, etc etc) in a very specific vision based on the history of their project. They think of their project more like evolving some foundational thing gradually/safely than always inventing a new thing.
Many of these issues don’t yet matter to new projects. So it’s hard to really compare the greenfield with a 20 year old codebase.
Remember when projects were getting overwhelmed by PRs from students just editing a line in a README so they could win a t-shirt? That was 2020, and they weren't using AI. The open source community has been going downhill for a while. The new generation isn't getting mentored by the old generation, so stable, old-fogey methods established by Linux distributions are eschewed by the new kids. Technology advancement has made open source interactions a little too easy, and unnecessarily fragile. Some ecosystems focus way too much on crappy reusable components, and don't focus enough on supply chain security.
Here's the good news: AI cannot destroy open source. As long as there's somebody in their bedroom hacking out a project for themselves, that then decides to share it somehow on the internet, it's still alive. It wouldn't be a bad thing for us to standardize open source a bit more, like templates for contributors' guides, automation to help troubleshoot bug reports, and training for new maintainers (to help them understand they have choices and don't need to give up their life to maintain a small project). And it's fine to disable PRs and issues. You don't have to use GitHub, or any service at all.
I think AI is a huge boon as it reduces the human bottleneck.
AI is a tool that must to be used well and many people currently raising pull requests seem to think that they don't even need to read the changes which puts unnecessary burden on the maintainers.
The first review must be by the user who prompted the AI, and it must be thorough. Only then I would even consider raising a PR towards any open source project.
Open Source isn't going anywhere. Open Contribution might be on the way out. I built an open source command line tool (https://github.com/dtnewman/zev) that went very minorly viral for a few days last year.
What I found in the following week is a pattern of:
1) People reaching out with feature requests (useful) 2) People submitting minor patches that take up a few lines of code (useful) 3) People submitting larger PRs, that were mostly garbage
#1 above isn't going anywhere. #2 is helpful, especially since these are easy to check over. For #3, MOST of what people submitted wasn't AI slop per se, but just wasn't well thought out, or of poor quality. Or a feature that I just didn't want in the product. In most cases, I'd rather have a #1 and just implement it myself in the way that I want to code organized, rather than someone submitting a PR with poorly written code. What I found is that when I engaged with people in this group, I'd see them post on LinkedIn or X the next day bragging about how they contributed to a cool new open-source project. For me, the maintainer, it was just annoying, and I wasn't putting this project out there to gain the opportunity to mentor junior devs.
In general, I like the SQLite philosophy of we are open source, not open contribution. They are very explicit about this, but it's important for anyone putting out an open source project that you have ZERO obligation to accept any code or feature requests. None.
We all know the solution will be yet another AI agent reviewing the reputation of the pull requests from the public and rating them. This even seems like an easy win for Microsoft and GitHub. Just make it already.
Hard drives shortage is already old news, CPUs are next.
Prior to LLM the concept of "Open Source" could co-exist with "Free Software" - one was a more pragmatic view of how to develop software, the other a political activist position of how code powering our world should be.
AI has laid bare the difference.
Open Source is significantly impacted. Business models based on it are affected. And those who were not taking the political position find that they may not prefer the state of the world.
Free software finds itself, at worst, a bit annoyed (need to figure out the slop problem), and at best, an ally in AI - the amount of free software being built right now for people to use is very high.
This one is probably going to be controversial. But I feel highlighting the drawbacks are also important, not just the benefits.
Three patterns I've noticed on the open-source projects I've worked on:
1. AI slop PRs (sometimes giant). Author responds to feedback with LLM generated responses. Show little evidence they actually gave any thought of their own towards design decisions or implementation.
2. (1) often leads me to believe they probably haven't tested it properly or thought of edge cases. As reviewer you now have to be extra careful about it (or just reject it).
3. Rise in students looking for job/internship. The expectation is that LLM generated code which is untested will give them positive points as they have dug into the codebase now. (I've had cases where they said they haven't tested the code, but it should "just work").
4. People are now even more lazy to cleanup code.
Unfortunately, all of these issues come from humans. LLMs are fantastic tools and as almost everyone would agree they are incredibly useful when used appropriately.
I think I have seen more open source projects get released since LLMs came out and the rate seems to be increasing. The cost of making software and open sourcing it has gone down a lot. We see some slop but as the models get better, the quality will get better and from the pace I have seen we went from gpt-3.5 to now opus4.6 i dont think it will be long before the LLMs get much better than humans in coding!
My current position is that AI companies should be taxed and the money should be distributed to open source developers.
There is a strong legal basis for this to happen because if you read the MIT license, which is one of the most common and most permissive licenses, it clearly states that the code is made available for any "Person" to use and distribute. An AI agent is not a person so technically it was never given the right to use the code for itself... It was not even given permission to read the copyrighted code, let alone ingest it, modify it and redistribute it. Moreover, it is a requirement of the MIT license that the MIT copyright notice be included in all copies or substantial portions of the software... Which agents are not doing in spite of distributing substantial portions of open source code verbatim, especially when considered in aggregate.
Moreover, the fact that a lot of open source devs have changed their views on open source since AI reinforces the idea that they never consented to their works being consumed, transformed and redistributed by AI in the first place. So the violation applies both in terms of the literal wording of the licenses and also based on intent.
Moreover, the usage of code by AI goes beyond just a copyright violation of the code/text itself; they appropriated ideas and concepts, without giving due credit to their originators so there is a deeper ethical component involved that we don't have a system to protect human innovation from AI. Human IP is completely unprotected.
That said, I think most open source devs would support AI innovation, but just not at their expense with zero compensation.
From my POV (30 or so years working on the same FLOSS project), AI isn't "destroying Open Source" through an effect on contributions. It is, however, destroying open source through its ceaseless, relentless, unabatable crawling of open source git repositories, commit by commit rather than via git-clone(1).
Project after project reports wasted time, increased hosting/bandwidth bills, and all around general annoyance from this UTTER BULLSHIT. But every morning, we wake up, and its still there, no sign of it ever stopping.
>>... Crypto ... are [is] pretty much useless.
Other than by corrupt criminals and mafia types who have a need to covertly hide cash.
And then the current administration wants the government to 'protect' crypto investors against big losses. Gotta love it.
"If this was a problem already, OpenClaw's release, and this hiring by OpenAI to democratize agentic AI further, will only make it worse. Right now the AI craze feels the same as the crypto and NFT boom, with the same signs of insane behavior and reckless optimism."
There are definitely people abusing AI and lying about what it can actually do. However, Crypto and NFTs are pretty much useless. Many people (including me) have already increased productivity using LLMs.
This technology just isn't going away.
The house is poorly put together cause the carpenter used a cheap nail gun and a crappy saw.
LLMs are confidently wrong and make bad engineers think they are good ones. See: https://en.wikipedia.org/wiki/Dunning–Kruger_effect
If you're a skilled dev, in an "common" domain, an LLM can be an amazing tool when you integrate it into your work flow and play "code tennis" with it. It can change the calculus on "one offs", "minor tools and utils" and "small automations" that in the past you could never justify writing.
Im not a Lawyer, or a Doctor. I would never take legal advice or medical advice from an LLM. Im happy to work with the tool on code because I know that domain, because I can work with it, and take over when it goes off the rails.
Frankly, I don't like this kinds of takes. Yes, people are seeing more spam in their pull requests, but that's just what it is - spam that you need to learn how to filter. For regular engineers who can use AI, it's a blessing.
I'm a long time linux user - now I have more time to debug issues, submit them, and even do pull requests that I considered too time consuming in the past. I want and I can now spend more time on debugging Firefox issues that I see, instead of just dropping it.
I'm still learning to use AI well - and I don't want to submit unverified slop. It's my responsibility to provide a good PR. I'm creating my own projects to get the hang of my setup and very soon I can start contributing to existing projects. Maintainers on the other hand need to figure out how to pick good contributors on scale.
[dead]
[dead]
[dead]
AI training is information theft. AI slop is information pollution.
This is a deeply pessimistic take, and I think it's totally incorrect. While I believe that the traditional open source model is going to change, it's probably going to get better than ever.
AI agents mean that dollars can be directly translated into open-source code contributions, and dollars are much less scarce than capable OSS programmer hours. I think we're going to see the world move toward a model by which open source projects gain large numbers of dollar contributions, that the maintainers then responsibly turn into AI-generated code contributions. I think this model is going to work really, really well.
For more detail, I have written my thoughts on my blog just the other day: https://essays.johnloeber.com/p/31-open-source-software-in-t...
It didn’t take AI to destroy Open Source, we were already doing it to ourselves. LLMs just magnified the existing structural issues and made them even easier to exploit. But the trajectory was already clear.