I think vibe coding would greatly struggle with large open source projects unless your planning was exceptional and your comments on optimal coding style was exceptional, however...... For those small open source tools that many of us use daily and find invaluable, I actually think vibe coding is ideal for that. It can make a functional version quickly and you can iterate and improve it, and feel no loss for making it free to use.
I was very sceptical but I will admit I think vibe coding has a place in society, just what it is yet is still to be determined. It can't help most for sure but it can help some in some situations.
No problem! Just give the agents the ability to autonomously report issues, submit patches, and engage with library authors. Surely nothing can go wrong.
perhaps also more forking (not only absolute but also relative)
contribution dynamics are also changing
I'm fairly optimistic that generative ai is good for open source and the commons
what I'm also seeing is open source projects that had not so great ergonomics or user interfaces in general are now getting better thanks to generative ai
this might be the most directly noticeable change for users of niche open source
"Vibe coding raises productivity by lowering the cost of using and building on existing code, but it also weakens the user engagement through which many maintainers earn returns."
I think the title is clickbait.
The conclusion is:
"Vibe coding represents a fundamental shift in how software is produced and consumed. The productivity gains are real and large. But so is the threat to the open source ecosystem that underpins modern software infrastructure. The model shows that these gains and threats are not independent: the same technology that lowers costs also erodes the engagement that sustains voluntary contribution."
The dangers I see rather in projects drowning in LLM slop PR's, instead of less engagement.
And the benefits of LLMs to open source in lowering the cost to revive and maintain (abandoned) projects.
I asked Claude several times to resolve this ambiguity and it suggested various prioritisation strategies etc. however the resulting changes broke other functionality in my library.
In the end I am redesigning my library from scratch with minimal AI input. Why? because I started the project without the help of AI a few years back, I designed it to solve a problem but that problem and nuanced programming decisions seem to not be respected by LLMs (LLMs dont care about the story, they just care about the current state of the code)
I believe we will see a new huge wave of useful open source software. However don't expect the development model to stay the same. I was finally able to resurrect a few projects of mine, and many more will come. One incredible thing was the ability to easily merge what was worth merging from forks, for instance. The new OSS will be driven not much by the amount of code you can produce, but from the idea of software you have, how the software should look like, behave, what it should do to be useful. Today design is more important than coding.
I am a huge proponent of using AI tools for software development. But until I see a vibe coded replacement for the Linux kernel, PostgreSQL, gcc, git or Chromium, I am just going to disagree with this premise. If I am on a system without Python installed, I don't see Claude saying, oh, you don't need to download it, I'll write the Python interpreter for you.
I don't really read papers and haven't read this one either but that summary.
> In vibe coding, an AI agent builds software by selecting and assembling open-source software (OSS),
Are they talking about indirectly due to prior training of the model? No agent I use is selecting and assembling open source software. That's more of an integration type of job not software development. Are they talking about packages and libraries? If yes, that's exactly how most people use those too.
I mean like this:
> often without users directly reading documentation, reporting bugs, or otherwise engaging with maintainers.
and then,
> Vibe coding raises productivity by lowering the cost of using and building on existing code, but it also weakens the user engagement through which many maintainers earn returns.
Maintainers who earn "returns" must be such a small niche as to be insignificant. Or do they mean things like github stars?
> When OSS is monetized only through direct user engagement, greater adoption of vibe coding lowers entry and sharing, reduces the availability and quality of OSS, and reduces welfare despite higher productivity.
Now the hypothesis is exactly the opposite. Do agents not "select and assamble" OSS anymore? And what does this have to do with how OSS is "monetized"?
> Sustaining OSS at its current scale under widespread vibe coding requires major changes in how maintainers are paid.
Sustaining OSS insofar as maintainers do it for a living requries major changes. Period. I don't see how vibe coding which makes all of this easier and cheaper is changing that equation. Quality is a different matter altogether and can still be achieved.
I am seeing a bunch of disjointed claims taken as truth that I frankly do not agree with in the first place.
What would the result of such a study even explain?
Small bespoke personalized on the spot apps are the future with LLMs.
The future will absolutely not be "How things are today + LLMs"
The paradigm now for software is "build a tool shed/garage/barn/warehouse full of as much capability for as many uses possible" but when LLMs can build you a custom(!) hammer or saw in a few minutes, why go to the shed?
Bespoke personalized apps have been around forever in the form of shell scripts. LLMs will expand that capability to a lot more people but won’t change the basic dynamics. Just like shell scripts, LLMs will need lower level building blocks to work with. The main difference will be that the LLM ecosystem will include some libraries as well. Though, I believe, most of those libraries haven’t been written yet.
Getting poor software made by Indians is a cycle that takes months.
With AI you get the first poor version within hours. Then you can iterate. In a week of hourly iteration you'll most likely have something usable.
The Indians are still having meetings about when to have meetings and you can see from the logs they haven't accessed any of the documentation you gave them for the issue. Billing is on point though.
LLMs have been a godsend for throwaway software. I wouldn't want to count the number of hours I've spent hand rolling an application that exists for one purpose for about two weeks and then to the digital graveyard.
Now that LLMs can do it while I make coffee, I can do more useful things with my time like writing business code and posting on HN.
Finally a world where everyone uses different hammers and no one can work on anything unless you can ask the hammer-oracle to reverse engineer the other people's hammers and nails
This is also the general threat to books and encyclopedias. Grok for all its point of view flaws still generated a full encyclopedia in two months that Wikipedia by humans took 25 years and Britannica over three centuries. The only remaining issue is when we can finally trust AI to be correct more than expert humans and be aligned with human moral consensus.
It is effective but once cost of creating something is down, then you have less reason to collaborate and depend on each other vs asking your own LLM to build your own bubble.
When paired with new-found cognitive laziness and lack of motivation when you then use no AI it's not sure of the second order effects.
Something I've noticed is that AI code generation makes it easier/faster to generate code while shifting more work of the work of keeping code correct and maintainable to the code review stage. That can be highly problematic for open source projects that are typically already bottlenecked by maintainer review bandwidth.
It can be mitigated by PR submitters doing a review and edit pass prior to submitting a PR. But a lot of submitters don't currently do this, and in my experience the average quality of PRs generated by AI is definitely significantly lower than those not generated by AI.
interesting as an econ thought experiment .. but it assumes OSS revenue comes from direct developr engagement .. In practice .. most successful OSS is funded by enterprises .. cloud vendors .. or consulting engagements .. where broader adoption, including AI mediated usage, often increases demand of said OSS project
On the contrary, I hope vibe coding revives Linux desktop into a truly usable platform.
e.g. Vibe coding defeats GNOME developers' main argument for endlessly deleting features and degrading user experience - that features are ostensibly "hard to maintain".
Well, LLMs are rapidly reducing development costs to 0.
The bottleneck for UI development is now testing, and here desktop Linux has advantage - Linux users have been trained like Pavlov's dogs to test and write detailed upstream bug reports, something Windows and macOS users just don't do.
For me spending time on my open source projects doesn't make sense anymore.
People (the community and employers) previously were impressed because of the amount of work required. Now that respect is gone as people can't automatically tell on the surface if this is a low effort vibe code or something else.
Community engagement has dropped. Stars aren't being given out as freely. People aren't actively reading your code like they use to.
For projects done before llms you can still link effort and signal but for anything started now.. everyone assumes it's llm created. No one want to read that code and not in the same way you would read other humans. Fewer will download the project.
Many of the reasons why I wrote open source is gone. And knowing the biggest/only engagement will come from llms copying your work giving you no credit.. what's the point?
> When OSS is monetized only through direct user engagement, greater adoption of vibe coding lowers entry and sharing, reduces the availability and quality of OSS, and reduces welfare despite higher productivity. Sustaining OSS at its current scale under widespread vibe coding requires major changes in how maintainers are paid.
I can't think of even a single example of OSS being monetized through direct user engagement. The bulk of it just isn't monetized at all, and what is monetized (beyond like a tip jar situation where you get some coffee money every once in a while) is primarily sponsored by enterprise users, support license sales, or through grants, or something like that. A few projects like Krita sell binaries on the steam store.
I have written so many small scripts and apps that do exactly what I want. The general purpose OSS projects are always a compromise. I believe if LLMs mature some more years we will see a decline in these general purpose projects and will see a rise in personal apps. I dont think its something to worry about.
There's a balance between coding by hand and vibe coding that is important. The less you understand the code, the more boring maintaining the software becomes. It's OK for throw away code, but not for serious open source projects. Use it as a powerful tool rather than your replacement.
I wonder how many OSS projects are using AI to actively squash bugs so their projects are more rock-solid than before.
Also, seems to me if your project underwent a full AI standardized code-quality check (using 2 or 3 AI models), it would be considered the "standard" from which other projects could use. For example, if you needed a particular piece of code for your own project, the AI tooling could suggest leveraging an existing gold-standard project.
61 comments
[ 3.6 ms ] story [ 69.9 ms ] threadI was very sceptical but I will admit I think vibe coding has a place in society, just what it is yet is still to be determined. It can't help most for sure but it can help some in some situations.
more open source, better open source
perhaps also more forking (not only absolute but also relative)
contribution dynamics are also changing
I'm fairly optimistic that generative ai is good for open source and the commons
what I'm also seeing is open source projects that had not so great ergonomics or user interfaces in general are now getting better thanks to generative ai
this might be the most directly noticeable change for users of niche open source
I think the title is clickbait.
The conclusion is:
"Vibe coding represents a fundamental shift in how software is produced and consumed. The productivity gains are real and large. But so is the threat to the open source ecosystem that underpins modern software infrastructure. The model shows that these gains and threats are not independent: the same technology that lowers costs also erodes the engagement that sustains voluntary contribution."
The dangers I see rather in projects drowning in LLM slop PR's, instead of less engagement.
And the benefits of LLMs to open source in lowering the cost to revive and maintain (abandoned) projects.
However trying to get it to do anything other than optimise code or fix small issues it struggles. It struggles with high level abstract issues.
For example I currently have an issue with ambiguity collisions e.g.
Input: "California"
Output: "California, Missouri"
California is a state but also city in Missouri - https://github.com/tomaytotomato/location4j/issues/44
I asked Claude several times to resolve this ambiguity and it suggested various prioritisation strategies etc. however the resulting changes broke other functionality in my library.
In the end I am redesigning my library from scratch with minimal AI input. Why? because I started the project without the help of AI a few years back, I designed it to solve a problem but that problem and nuanced programming decisions seem to not be respected by LLMs (LLMs dont care about the story, they just care about the current state of the code)
> In vibe coding, an AI agent builds software by selecting and assembling open-source software (OSS),
Are they talking about indirectly due to prior training of the model? No agent I use is selecting and assembling open source software. That's more of an integration type of job not software development. Are they talking about packages and libraries? If yes, that's exactly how most people use those too.
I mean like this:
> often without users directly reading documentation, reporting bugs, or otherwise engaging with maintainers.
and then,
> Vibe coding raises productivity by lowering the cost of using and building on existing code, but it also weakens the user engagement through which many maintainers earn returns.
Maintainers who earn "returns" must be such a small niche as to be insignificant. Or do they mean things like github stars?
> When OSS is monetized only through direct user engagement, greater adoption of vibe coding lowers entry and sharing, reduces the availability and quality of OSS, and reduces welfare despite higher productivity.
Now the hypothesis is exactly the opposite. Do agents not "select and assamble" OSS anymore? And what does this have to do with how OSS is "monetized"?
> Sustaining OSS at its current scale under widespread vibe coding requires major changes in how maintainers are paid.
Sustaining OSS insofar as maintainers do it for a living requries major changes. Period. I don't see how vibe coding which makes all of this easier and cheaper is changing that equation. Quality is a different matter altogether and can still be achieved.
I am seeing a bunch of disjointed claims taken as truth that I frankly do not agree with in the first place.
What would the result of such a study even explain?
The future will absolutely not be "How things are today + LLMs"
The paradigm now for software is "build a tool shed/garage/barn/warehouse full of as much capability for as many uses possible" but when LLMs can build you a custom(!) hammer or saw in a few minutes, why go to the shed?
You go to the shed because you know the hammer has undergone extensive testing and won’t immediately be hacked, costing you hundreds of millions.
Getting poor software made by Indians is a cycle that takes months.
With AI you get the first poor version within hours. Then you can iterate. In a week of hourly iteration you'll most likely have something usable.
The Indians are still having meetings about when to have meetings and you can see from the logs they haven't accessed any of the documentation you gave them for the issue. Billing is on point though.
Now that LLMs can do it while I make coffee, I can do more useful things with my time like writing business code and posting on HN.
Beyond that, meh.
It can be mitigated by PR submitters doing a review and edit pass prior to submitting a PR. But a lot of submitters don't currently do this, and in my experience the average quality of PRs generated by AI is definitely significantly lower than those not generated by AI.
e.g. Vibe coding defeats GNOME developers' main argument for endlessly deleting features and degrading user experience - that features are ostensibly "hard to maintain".
Well, LLMs are rapidly reducing development costs to 0.
The bottleneck for UI development is now testing, and here desktop Linux has advantage - Linux users have been trained like Pavlov's dogs to test and write detailed upstream bug reports, something Windows and macOS users just don't do.
People (the community and employers) previously were impressed because of the amount of work required. Now that respect is gone as people can't automatically tell on the surface if this is a low effort vibe code or something else.
Community engagement has dropped. Stars aren't being given out as freely. People aren't actively reading your code like they use to.
For projects done before llms you can still link effort and signal but for anything started now.. everyone assumes it's llm created. No one want to read that code and not in the same way you would read other humans. Fewer will download the project.
Many of the reasons why I wrote open source is gone. And knowing the biggest/only engagement will come from llms copying your work giving you no credit.. what's the point?
I can't think of even a single example of OSS being monetized through direct user engagement. The bulk of it just isn't monetized at all, and what is monetized (beyond like a tip jar situation where you get some coffee money every once in a while) is primarily sponsored by enterprise users, support license sales, or through grants, or something like that. A few projects like Krita sell binaries on the steam store.