I don't understand this article. ChatGPT made Google search and SO irrelevant so vibe coding is killing open source? ... That's a stretch.
> The LLM will not interact with the developers of a library or tool, nor submit usable bug reports, or be aware of any potential issues no matter how well-documented.
Arn't these interactions responsible for the claimed burn-out suffered by open-source maintainers? If you want interaction then, I don't know, go to a conference? Again, I don't get the issue. Seems like a good thing! Users are able to find answers and solutions to their quesitons more efficiently--all the while, still using the open-source library. The usage chart is still seeing tremendous growth! Developers are still using the library to solve their problems. It seems like exactly what open-source was intended for.
The issue to me is that, the incentives for investing in open-source have changed for some maintainers in such a way that they're no longer in alignment with their return on their investment. Maybe there are fewere people interacting with them and so fewer people to discover how "great" they are. Maybe fewer eye balls on their resume. The point is, open-source was a means to an end. And, so, frankly, I don't give a shit.
LLMs are making open-source technology accessible to more people and that's a good thing.
> with only 0.1% of users still choosing GPT‑4o each day.
LOL WHAT?! I'm 0.1% of users? I'm certain part of the issue is it takes 3-clicks to switch to GPT-4o and it has to be done each time the page is loaded.
> that they preferred GPT‑4o’s conversational style and warmth.
Uh.. yeah maybe. But more importantly, GPT-4o gave better answers.
Zero acknowledgement about how terrible GPT-5 was when it was first released. It has since improved but it's not clear to me it's on-par with GPT-4o. Thinking mode is just too slow to be useful and so GPT-4o still seems better and faster.
I agree - I use 4o via the API, simply because it answers so quickly. Its answers are usually pretty good on programming topics. I don't engage in chit-chat with AI models, so it's not really about the personality (which seems to be the main framing people are talking about), just the speed.
I decided to learn IPv6 recently and I'm pleasantly surprised how simple and elegant it is. Truly a joy. Highly recommend, if you've never worked with IPv6 to try it. It's like discovering a bidet.
You could describe Richard Stallman as someone who refuses to use proprietary software because he sees using it as becoming complicit--however indirectly--in a technology ecosystem that violates the values he’s committed to.
"Just don't use X" is in fact a very engaged and principled response. Try again.
The important thing is to relate to other humans and to be sure of what kind of human you're interacting with: the creators of the game or other players.
The staleness that actually "shipping once"[0] gives is precisely the space where human player creativity grows and thrives in.
----
[0] I understand you can get the similar results and better base games if you patch things occasionally, but constant patches[1] hides the jank and repetitiveness with novelty.
[1] And dynamically creating "content" with LLMs is like a constant stream of patches.
A new MUD needs a way to build several thousand rooms, mobs, items, etc. LLMs can help with that process, though I wouldn’t trust them alone with things like balance.
Similarly, existing MUDs adding new areas need hundreds of rooms, mobs, items, etc. In my experience MUDs tend to stagnate when there’s no new content for long time players.
Some of the coolest MUDs I played in had effectively only two useful rooms, and no mobs or items to really speak of. They were barely more than a couple of IRC chat rooms, but with the ANSI colors support and complex script languages a MUD Engine directly over telnet could provide to a good MUD client.
There were far more genres of MUDs than just the Diku-style ("EverQuest-like", to use as analogy the graphic MMO that took a lot from the Diku-style of MUD) that needed to be "endless" content farms of mobs and items and new areas full of more mobs and items.
But also many of the fan favorite Diku-style MUDs were procedurally generated and no one was actually building all those thousands of rooms/mobs/items by hand even then. In theory you could use an LLM as a part of procedural generation process, but that's not the kind of content I would have wanted from a good MUD at the time I was heaviest playing MUDs. (But then I also didn't play many Diku-style/Diku-inspired MUDs, either. I was more on the Socializer side of things at the time.)
I’ll admit YMMV and my comment should’ve been better scoped — but it sounds like you’re not disagreeing that for those, LLMs are useful in the way I suggested.
I took a moment to participate in an ARIN (https://www.arin.net) call. They drive policy for IP addresses that run the Internet. Anyway, one thing that stood out to me was during a portion of the conference the organization was holding elections for open seats. The seats are open to anyone, I'm pretty sure.
There were a bunch of people, from across the country, MANY in retirement, trying their best to sell themselves that they are the right candidate.
Sergey can just make a phone call and he gets to build Gemini and run a billion dollar organization and have meaning in old-age. This is what wealth buys you. The rest of us, I guess, we will be arm-wresteling for the few open oppertunities to make an impact.
> The LLM will not interact with the developers of a library or tool, nor submit usable bug reports, or be aware of any potential issues no matter how well-documented.
Arn't these interactions responsible for the claimed burn-out suffered by open-source maintainers? If you want interaction then, I don't know, go to a conference? Again, I don't get the issue. Seems like a good thing! Users are able to find answers and solutions to their quesitons more efficiently--all the while, still using the open-source library. The usage chart is still seeing tremendous growth! Developers are still using the library to solve their problems. It seems like exactly what open-source was intended for.
The issue to me is that, the incentives for investing in open-source have changed for some maintainers in such a way that they're no longer in alignment with their return on their investment. Maybe there are fewere people interacting with them and so fewer people to discover how "great" they are. Maybe fewer eye balls on their resume. The point is, open-source was a means to an end. And, so, frankly, I don't give a shit.
LLMs are making open-source technology accessible to more people and that's a good thing.
reply