Just for my understanding of what actually happened, because I think 3dEyes is not the only one who’s confused here, you’re not talking about this 24-hours suspension, right?
After checking Atomozero’s post and comments activity, am I right to assume some of his posts were deleted afterwards? Because I cannot see the one he got 30 days ban for, I just see a 2-months gap in his activity on the forum. Also, I could swear I’ve seen a post from him a few days ago where he apologized for his past behavior and mentioned that his HaikuPorts recipes are not being merged. But after a few hours that post was gone. If I understand correctly, that’s the post you mentioned as a first closed version of “On LLM policy of HaikuPorts”.
So my bigger question is, are the closed posts being actively deleted by the moderators, or were the posts deleted by the user himself? If it’s the former, could the mods please stop doing it? I believe it would help regular users as I, who doesn’t have access to the deleted posts, to track the events timeline and to better understand further moderators’ decisions and actions, avoiding misunderstanding and heated discussions afterwards.
@nephele I’m not going to apologise for holding a different view, and I won’t accept being singled out as the reason someone left the forum. People make their own choices.
Let me also be clear about my actual position, since it keeps getting lost. I’ve never argued that everyone should use LLMs. My point has only ever been that an LLM is a tool, and using one is not in itself wrong or shameful. That’s a position several long-time contributors share. I also don’t believe a contributor should have to “declare” which tool they wrote something with, any more than they’d declare whether they used vim, Pe, Genio or an IDE; what matters is whether the code works. Yet ever since I said this openly, I’ve gone from being a contributor with a different opinion to being treated as a problem to be removed. I hold nothing against the people who disagree with me, and I’d ask for the same in return.
I want to say plainly that @Begasus and everyone who has carried HaikuPorts over the years have done outstanding work. My respect for that is genuine, and my disagreement is not with them.
While I’m at it, one small thing: some people read the name “BubiCam” as a deliberate provocation. It wasn’t. Where I’m from, “Bubi” is just an affectionate word for something small and cute. There was no hidden meaning, and reading one into it says more about the reader than about me.
My honest concern is with how disagreement itself is handled here. The current climate has reached a point where contributors feel they have to “admit” or “apologise” for having used an LLM to build something that simply didn’t exist before, as if producing working software were something to be ashamed of. And people who were actively working on writing new programs for Haiku have stopped, because the verdict was already decided before anyone had even seen the program. When the judgement comes before the work, you don’t just lose one contributor; you discourage everyone watching from trying at all. That, to me, is the real cost, and it deserves to be faced honestly instead of being reduced to the conduct of one person.
So you can ban me, close my account, do as you see fit. None of it changes anything for me. I’ll keep using Haiku, because I love this system, and I’ll keep writing the small programs I need and that the OS is missing, on my own, without asking anyone for permission. I’ll leave it there.
Moderating is one thing, trying to put foot on the scale is another… you can disagree with Andrea’s ideas but you have to respect the fact that he is a true enthusiast and a member of the BeOS-Haiku community for decades… accusing him of being a provocateur, is not acceptable in my personal opinion, it is just a twist in a discussion that instead of remaining on the tracks of objectivity has become ideological.
Funny you should say that, because in my practical experience as a moderator elsewhere, there’s a very high correlation between “concerned citizens” who “only worry about free speech” and people who start foaming at the mouth about “wokeness” the moment you push back a little bit, then move on to far-right talking points faster than a mod can click “Lock Topic”.
Why do we need an “official policy” towards any tool? Isn’t the official policy high quality code? I have a calculator but I like doing math in my head. Some don’t. So what? In the end all that matters is that the math is correct.
I’m not a fan of AI, I’m a fan of Haiku. I want Haiku to be the best operating system in the world.
All this time arguing could have been used reviewing code contributions by enthusiastic community members who know and love Haiku and are trying to contribute tools we have dreamed of having for decades. Thanks guys.
“I don’t really care about being right, I just care about success.”
The arguments against AI have been made many times over in the other topic, the one that was closed. To ask yet again is starting to look disingenuous.
I 100% agree, we should have banned him indefinitely, so that we can have a serious discussion instead of trolling. If we let him continue he will be the best ally of people trying to discredit LLM users
I recently stubled upon this thread about toxic people in online communities and I think it applies pretty good to the situation we have here currently:
To sum it up: Sometimes people really hate a community,but they still stay and ruin the place for everyone and make both themselves and everyone else feel bad.
I think that’s pretty good advice to the moderators and also certain people (if you’re only here for trolling,just leave).
The negative attitude towards moderators and admins makes this a quite toxic place and drives people away from the community.
This is a fair criticism (and also actually on-topic for a thread that’s supposedly about moderation, so yay :-).
I’ll raise that issue and see if anyone else is up for reconsidering it.
This, however, is where I think the real issue that needs dealt with is.
(It’s not @Andrea himself; that reply I’m quoting from was fine.)
The issue is that:
not everyone here agrees that LLMs are actually tools
of those who do, not everyone agrees that using them is desirable, or even acceptible
of those who do, some insist (as @Andrea does in that reply) that everyone must accede on that point and treat LLMs as just another tool
That insistence is the issue that’s getting in the way of addressing the underlying disagreement, just as anyone insisting that no one use LLMs would.
No one, on either side, is being convinced to change their mind about the fundamental issue.
This forum, and the Haiku space in general, will not be an LLM-free space; the users, moderators and developers don’t agree among themselves on that.
The OS itself will not have LLM-generated code in it, that’s been decided, but haikuports and the forums aren’t going to take that strong a position. I don’t like it, but I’ll get over it.
But equally the forums (and maybe haikuports as well) are not going to be a totally welcoming space for LLMs either; the users, moderators and developers don’t agree among themselves on that.
It’s long past time for arguing over whether LLMs and their output are OK; no one is changing their minds at this point. I would sincerely appreciate it if we could all just put that part aside and consider instead what we do about it.
@us3r1d That’s a fair point, and you’re right. I don’t actually need anyone to agree that an LLM is “just a tool”; people can think what they like about it. All I’d ask is that the output be judged on its merits, not pre-judged by how it was made. And you’re right that we should move from arguing about LLMs to deciding what to do in practice.
On the practical side, here’s one concrete thing. Back in March, on the HaikuPorts list, Adrien (@PulkoMandy) himself pointed out that the “only committers can vote” rule belongs to the Haiku project and doesn’t apply to HaikuPorts, which has no written governance rules, and that we should “have a proper discussion with everyone” and build a consensus. I think he was right then. A decision that affects the whole community works best when it’s discussed somewhere the whole community can actually take part; otherwise the people most affected end up excluded, and the result feels imposed rather than agreed.
So my suggestion is simply this: keep the technical work where it belongs, on the mailing list and GitHub, but community-wide policy decisions should be made, and visibly recorded, here on the forum, where every user and developer can read along and weigh in.
That’s it from me. Thanks for keeping this on-topic.
But that would be treating the LLM as another tool.
You’re going to have to accept that LLM-generated content will be judged based on how it was made, because a significant chunk of this community disagree with you on whether making things with LLMs is acceptable.
Just like I have to accept that this forum (and haikuports) will have LLM-generated stuff around, because a significant chunk of this community disagree with me on whether making things with LLMs is acceptable.
If we can start from there, then we should be able to come up with a way for both perspectives to share a forum.
I don’t actually disagree with this, since I haven’t used email lists to any real degree for a decade or more now. Email for me is largely a wasteland of automated notifications and spam. Ick.
But the forum isn’t the place for that discussion at the moment, because this forum isn’t where that decision will be made.
It’s up to the haikuports mailing list to make that decision.
In many of your posts, I found often the word “hate” and you clearly demonstrated your bitterness towards … – I will not name it to avoid your triggering - you seem to enjoy the segregation of those who think differently from you, applying labels that allow you to ignore them - The sum of all this does not seem to lean towards demonstrating love in the community
