👀 … https://sfconservancy.org/blog/2026/apr/15/eternal-november-generative-ai-llm/ …my colleague Denver Gingerich writes: newcomers' extensive reliance on LLM-backed generative AI is comparable to the Eternal September onslaught to USENET in 1993.
-
I just noticed the version posted didn't incorporate various final edits. I've been defending *that* version of the post (which almost no one saw) *not* the one you all read.
@ossguy confirmed some final changes may have been lost (possibly moving from Etherpad to website).
@ossguy & I are working to fix that now.
The disconnect this evening hopefully makes sense now. I'll reply to this post when we've updated the public URL.@bkuhn @linux_mclinuxface @josh @wwahammy @cwebber @burnoutqueen @ossguy ah ha! thank you! It did feel off. -
@josh @wwahammy I definitely agree with discouraging developers who should know better from making LLM-generated commits that aren't very good. But this is a separate issue from communicating with the people who are just getting excited about buildings software, so we can encourage them to do so in FOSS-friendly ways. (2/2)
For what it's worth, if your blog post had come across saying what you are *currently* saying on Fedi, I would be much more enthusiastic and appreciative of it, and I suspect others would be too. -
One of *many* arguments against: codebases substantially contributed to by LLMs will develop a tolerance for complexity that is not conducive to being maintained by anything *other* than an LLM.
@josh @silverwizard @ossguy @bkuhn @karen @kees @wwahammy You can prevent it by asking LLM tho add comments and check those comments I'm pretty sure you can make a very good PR with a LLM.
That said without bounds this will definitely not be the default and yes what you said will happen.
Although with the current rate things are going, a LLM will probably be able to rewrite a complete program source-code and re-format it in anything that is currently possible...Which is way worse for FOSS.
-
"Words that aren't there" like this?
> Historically, software freedom has has typically necessitated interacting with others
Suggesting that this is merely "historically"?
> more easily with LLM-backed generative AI coding tools (and the ease with which changes can be made generally) there is less of a natural tendency for people to work with existing FOSS communities. And we should be ok with that!
We should be okay with that? We should not treat it as an *existential threat* and respond accordingly? Those are the words that aren't there?To be clear, I am genuinely trying to understand your position because it seems distinct from the (traditional) LLM criticisms (many of which I share). But what is the existential threat? I would understand that in this context to mean a threat to the existence of FOSS. How do you see people improving their software with LLMs as a threat?
My simplified model of the situation is: a person who was previously unable to change their software now can. Then they can either:
A) never contribute it upstream
B) contribute it upstream
(BTW these are also the same 2 outcomes for people who can change their software without LLMs.)I don't see how "A" poses a threat. There is no interaction with the FOSS upstream.
I don't see how "B" poses a threat. Upstream can either ignore it (no change to FOSS) or engage with it (FOSS improved).
What threat to FOSS do you see?
-
To be clear, I am genuinely trying to understand your position because it seems distinct from the (traditional) LLM criticisms (many of which I share). But what is the existential threat? I would understand that in this context to mean a threat to the existence of FOSS. How do you see people improving their software with LLMs as a threat?
My simplified model of the situation is: a person who was previously unable to change their software now can. Then they can either:
A) never contribute it upstream
B) contribute it upstream
(BTW these are also the same 2 outcomes for people who can change their software without LLMs.)I don't see how "A" poses a threat. There is no interaction with the FOSS upstream.
I don't see how "B" poses a threat. Upstream can either ignore it (no change to FOSS) or engage with it (FOSS improved).
What threat to FOSS do you see?
Leaving aside for a moment the issue that (B) can leave maintainers drowning in slop...
There is a massive game-theoretic problem here. Employers are forcing some developers to deal with LLMs. Some people of their own volition are excited about LLMs. Some people want nothing to do with LLMs. People who heavily use and rely on LLMs have different standards for acceptable complexity and maintainability. LLMs encourage people to work more in silos without collaboration and use LLMs instead of collaborators, and that serves LLM purveyors. It's much easier to collaborate with "You're absolutely right!". Codebases and ecosystems and communities diverge. -
@kees @karen @josh @silverwizard @wwahammy @ossguy @bkuhn
This is an aside, but
I am surprised to see anyone say there's nothing novel to object to about LLMs. I think though that I might post about that tomorrow as it's late now where I am. But I definitely would love to know more about why you think that because a major concern with LLMs I have is what Sean calls epistomological collapse which is it not talked about how it's destroying trustwortiness of info pervasively? Anyway, I should collect up my sources and do a complete argument for that on my personal instance if anyone cares what I think on it (which, feel free to not)@firefly_lightning @karen @josh @silverwizard @wwahammy @bkuhn @ossguy
I have been trying to keep the scope of my replies as narrow as possible because I think there are unique benefits of LLM use in software development. To your specific point, I think software is more resilient to epistomological collapse in the sense that is has provable characteristics (e.g. it has to compile). Perhaps I am being naive!
The larger scopes around LLMs in prose, art, etc are IMO substantially different and much more alarming.
-
@firefly_lightning @karen @josh @silverwizard @wwahammy @bkuhn @ossguy
I have been trying to keep the scope of my replies as narrow as possible because I think there are unique benefits of LLM use in software development. To your specific point, I think software is more resilient to epistomological collapse in the sense that is has provable characteristics (e.g. it has to compile). Perhaps I am being naive!
The larger scopes around LLMs in prose, art, etc are IMO substantially different and much more alarming.
I think software is not at all immune, in the sense that just as LLMs can produce grammatically correct sentences that make no sense and have no factual basis, they can produce code that *compiles* but is utterly alien to what any sensible human with taste would write. -
Leaving aside for a moment the issue that (B) can leave maintainers drowning in slop...
There is a massive game-theoretic problem here. Employers are forcing some developers to deal with LLMs. Some people of their own volition are excited about LLMs. Some people want nothing to do with LLMs. People who heavily use and rely on LLMs have different standards for acceptable complexity and maintainability. LLMs encourage people to work more in silos without collaboration and use LLMs instead of collaborators, and that serves LLM purveyors. It's much easier to collaborate with "You're absolutely right!". Codebases and ecosystems and communities diverge.Also, the "drowning in slop" problems have real-world social consequences too! Some projects are having to go closer to "we don't take patches from people we don't know", and that's damaging the ability to do drive-by or one-off contributions, or to onboard new contributors. That feels like the prologue of ecosystem collapse. -
Also, the "drowning in slop" problems have real-world social consequences too! Some projects are having to go closer to "we don't take patches from people we don't know", and that's damaging the ability to do drive-by or one-off contributions, or to onboard new contributors. That feels like the prologue of ecosystem collapse.
I think the "attention competition" will find a viable solution. It has been solved many times before when we've all fought spam in its many forms. Slop is the byproduct of LLM usage the way spam is a byproduct of email usage, as a grossly simplified comparison. (It's not *good* to have spam of any kind, of course, but for example I can't avoid email spam unless I stop using email entirely, and I'm not about to do that nor stop writing software.)
I see where LLMs are making things genuinely easier for humans (review, debugging, etc), though, so I don't share the same sense of impending ecosystem collapse.
-
I just noticed the version posted didn't incorporate various final edits. I've been defending *that* version of the post (which almost no one saw) *not* the one you all read.
@ossguy confirmed some final changes may have been lost (possibly moving from Etherpad to website).
@ossguy & I are working to fix that now.
The disconnect this evening hopefully makes sense now. I'll reply to this post when we've updated the public URL.https://sfconservancy.org/blog/2026/apr/15/eternal-november-generative-ai-llm/ now reflects what I thought was posted hours ago. Sorry for the confusion.
You all got an insight into how much you have to draft & redraft to consider difficult policy questions. Anyone who works in policy drafted a dozen things that were not quite right before getting it right.
Anyway, if you still think it's terrible, I refer you to all my other posts from this evening.
@ossguy @josh @wwahammy @linux_mclinuxface @burnoutqueen @cwebber @silverwizard @mjw @mmu_man
-
Leaving aside for a moment the issue that (B) can leave maintainers drowning in slop...
There is a massive game-theoretic problem here. Employers are forcing some developers to deal with LLMs. Some people of their own volition are excited about LLMs. Some people want nothing to do with LLMs. People who heavily use and rely on LLMs have different standards for acceptable complexity and maintainability. LLMs encourage people to work more in silos without collaboration and use LLMs instead of collaborators, and that serves LLM purveyors. It's much easier to collaborate with "You're absolutely right!". Codebases and ecosystems and communities diverge.Most of your reply didn't seem to be describing threats to FOSS. (Using/not using LLMs, etc.) The only statements I could see maybe being a threat to FOSS was this:
> LLMs encourage people to work more in silos without collaboration and use LLMs instead of collaborators
Are you suggesting existing contributors will exit FOSS because of their LLM use? I don't understand how these two things are related. And getting back to @ossguy 's post, it looks like quite the opposite: there are people *entering* FOSS due to LLMs.
> Codebases and ecosystems and communities diverge.
Through what mechanism?
-
@josh @silverwizard @ossguy @bkuhn @karen @kees @wwahammy You can prevent it by asking LLM tho add comments and check those comments I'm pretty sure you can make a very good PR with a LLM.
That said without bounds this will definitely not be the default and yes what you said will happen.
Although with the current rate things are going, a LLM will probably be able to rewrite a complete program source-code and re-format it in anything that is currently possible...Which is way worse for FOSS.
> You can prevent it by asking LLM tho add comments and check those comments
You really can't; it is not anywhere close to that simple. The problem isn't just line-level, it's (among many other things) systemic design complexity, tolerance for technical debt, unbounded (except by token budget) capacity to duplicate or reinvent rather than reuse, none of the programmer's virtue of "laziness", and a substantial multiplier on the hubris.
-
Most of your reply didn't seem to be describing threats to FOSS. (Using/not using LLMs, etc.) The only statements I could see maybe being a threat to FOSS was this:
> LLMs encourage people to work more in silos without collaboration and use LLMs instead of collaborators
Are you suggesting existing contributors will exit FOSS because of their LLM use? I don't understand how these two things are related. And getting back to @ossguy 's post, it looks like quite the opposite: there are people *entering* FOSS due to LLMs.
> Codebases and ecosystems and communities diverge.
Through what mechanism?
I'm suggesting, as the article we're replying to points out, that it's now easier for people to go "eh, I don't need FOSS collaborators, I have LLMs and look how many lines of code I produce per day!". And conversely, projects developed heavily by LLM will not be welcoming environments to people who don't want to work with LLMs. This creates silos. -
I'm suggesting, as the article we're replying to points out, that it's now easier for people to go "eh, I don't need FOSS collaborators, I have LLMs and look how many lines of code I produce per day!". And conversely, projects developed heavily by LLM will not be welcoming environments to people who don't want to work with LLMs. This creates silos.And the problem isn't just *new projects* that are LLM-written, it's the LLM-cordyceps taking over the bodies of existing projects and driving out developers who want to work with humans and don't have the complexity-and-debt-and-NIH tolerance of LLMs. (And solving that isn't as simple as forking, because it's possible one or both groups don't have the critical mass that they would have had together.)
-
I think software is not at all immune, in the sense that just as LLMs can produce grammatically correct sentences that make no sense and have no factual basis, they can produce code that *compiles* but is utterly alien to what any sensible human with taste would write.
@josh @firefly_lightning @silverwizard @ossguy @bkuhn @karen @wwahammy
> but is utterly alien to what any sensible human with taste would write.
This implies no humans are doing code review. If it's crap code then it goes nowhere and collapse is avoided.
And yes, I'm aware of some projects that are utterly YOLOing everything into their codebases, and I think the results will speak for themselves, in either outcome! Either they flame out with no damage to larger FOSS, or the LLMs become so good that we get beautiful FOSS code and proprietary software becomes a thing of the past. Limping along in between seems unlikely to me.
-
@josh @firefly_lightning @silverwizard @ossguy @bkuhn @karen @wwahammy
> but is utterly alien to what any sensible human with taste would write.
This implies no humans are doing code review. If it's crap code then it goes nowhere and collapse is avoided.
And yes, I'm aware of some projects that are utterly YOLOing everything into their codebases, and I think the results will speak for themselves, in either outcome! Either they flame out with no damage to larger FOSS, or the LLMs become so good that we get beautiful FOSS code and proprietary software becomes a thing of the past. Limping along in between seems unlikely to me.
> This implies no humans are doing code review. If it's crap code then it goes nowhere and collapse is avoided.
No, it implies no humans *without the aid of LLMs* are reviewing *how easy it would be to maintain without LLMs*. And that's an easy state to get into.
I think the "in between" outcome seems much more likely to me than it does to you: projects can limp along for a long time, and be popular enough to discourage competition or hold onto users for a while.
Diseases that are contagious before people are symptomatic are especially hazardous. LLM-written technical debt takes time to become symptomatic. The epidemic is time-delayed from the initial outbreak, and exponentials are hard to see from the middle. -
> You can prevent it by asking LLM tho add comments and check those comments
You really can't; it is not anywhere close to that simple. The problem isn't just line-level, it's (among many other things) systemic design complexity, tolerance for technical debt, unbounded (except by token budget) capacity to duplicate or reinvent rather than reuse, none of the programmer's virtue of "laziness", and a substantial multiplier on the hubris.@josh @silverwizard @ossguy @bkuhn @karen @kees @wwahammy ok so basically your point is that someone hires a bunch of machines to do a lot of work, but then when the machines leaves you are stuck with all the stuff the machines made which you cannot maintain alone, because it's too much work.
And that's so true.
It's the same with math and a calculator, but a calculator isn't subscription based. Which in my opinion is the real issue. -
> This implies no humans are doing code review. If it's crap code then it goes nowhere and collapse is avoided.
No, it implies no humans *without the aid of LLMs* are reviewing *how easy it would be to maintain without LLMs*. And that's an easy state to get into.
I think the "in between" outcome seems much more likely to me than it does to you: projects can limp along for a long time, and be popular enough to discourage competition or hold onto users for a while.
Diseases that are contagious before people are symptomatic are especially hazardous. LLM-written technical debt takes time to become symptomatic. The epidemic is time-delayed from the initial outbreak, and exponentials are hard to see from the middle.@josh @firefly_lightning @silverwizard @ossguy @bkuhn @karen @wwahammy
> The epidemic is time-delayed from the initial outbreak, and exponentials are hard to see from the middle.
I agree with this, and I expect to see some evidence of slop-code in real software (especially proprietary) in the coming years. Where I differ, though, is that I see *benefits* being time delayed too. I just don't think any of this is going to be all bad or all good.
If the cordyceps made some people zombies and made other people able to fly. And we could shift the ratio through education and experience.
And getting cordyceps in the first place required boiling all our oceans.
-
@josh @silverwizard @ossguy @bkuhn @karen @kees @wwahammy ok so basically your point is that someone hires a bunch of machines to do a lot of work, but then when the machines leaves you are stuck with all the stuff the machines made which you cannot maintain alone, because it's too much work.
And that's so true.
It's the same with math and a calculator, but a calculator isn't subscription based. Which in my opinion is the real issue.@MisterMaker @josh @silverwizard @ossguy @bkuhn @karen @wwahammy
I am reminded of Kernighan’s Law: because debugging is twice as hard as writing code, writing code as cleverly as possible makes you, by definition, not smart enough to debug it.
So I really don't want the LLM writing clever code.
But yes, now we have to rent "thinking".
All the more reason to have FOSS LLM models to resist rentier capitalism. -
@bkuhn @ossguy I have to admit that I am pretty surprised by this post. Not in terms of being welcoming to newcomers, which is something I have advocated for and made the center of all of my FOSS work.
However, the post says the following:
> I encourage all of us in the FOSS community to welcome the new software developers who've adopted these tools, investigate their motivations, and seriously consider cautiously and carefully incorporating their workflows with ours.
While the sentence which follows acknowledges that "seasoned software developers understand the benefits and limitations of LLM-assisted coding tools", there are two big things I expected at least acknowledged:
- Many maintainers are facing *burnout* over the situation. However, I agree that addressing this in terms of norms is something we can consider
- The biggest thing I am surprised to not see addressed at all is the licensing and copyright implications(cotd)
welcome the new software developers who’ve adopted these tools
how about No
incorporating their workflows with ours
yea No
get a backbone maybe?
