Samsung Electronics has integrated Anthropic’s Claude into its semiconductor design processes. Consequently, support for Haiku on all Samsung devices(including its memory products and smartphones) should likewise be banned.
AI impact in hardware, firmware or even softwares supported on Haiku from third parties are no concern and not the responsability of Haiku devs.
While code present or merged in Haiku code base is, and therefore should be maintained (and this, understood) by these devs.
I agree with that. So, I don’t think anyone needs to get mad about using AI agents to build tools or third-party apps for Haiku. It blows my mind how the dev team jumps out of their skin at the mere sight of the letter ‘L’ in LLM.
I agree with you, let’s delete posts just because we do not like contracditory opinions on a debate thread. Or maybe simply we can agree to not agree like normal adults would do.
Lets start from beginning. Some user started topic about Haiku 3rd-party software for using AI. Then a lot of people (including moderators) come and started to blame how bad AI tools are, that it is used for personal task scheduling, that it is immoral, force children to work in mines, destroy planet etc.. And such content become >95% of topic. It is definitely off-topic and should be deleted by proper moderation. Nobody tells that AI tools pros and cons can’t be discussed, but it should be done in proper topic, not everywhere where words “AI”, “LLM” are mentioned.
Yes, topic author violated forum rules for making LLM generated summary in topic head post, but it is not an excuse to start tons of flood. Post should be hidden and asked by topic author to rewrite, thats all.
Ah yes, rewriting history as usual.
Here is a newsflash for you. We will kick out or suspend users when they break the rules, we won’t delete posts because you, personally, do not agree with them.
Also consider this an explicit warning to not harass forum members for daring to write their opinion on a topic (in the forum section for that topic..) in the future.
So, to summarize: comments that agree with what the usual four say will never be hidden or blocked (even if they are off-topic and aggressive), while all the others can be hidden or removed just for having two L’s in their last name 
Not sure what you are talking… Following the forum rules is not that difficult
Interesting rules, but they aren’t applied to everyone 
A referee isn’t allowed to act as a player and kick the soccer ball into the goal. You are the moderator, yet you participate in the discussion as a player, fighting with others, pushing the rules link, and threatening bans. This is highly inappropriate for the development of an open-source community project. Stop your sarcastic attitude.
Forum rules serve a purpose for “healthy” discussion and debate. The forum rules are not a mirror of Orwell’s “Animal Farm”… c’mon!
Moderators are supposed to delete posts users disagree with, and are not allowed to write in the forum, did I get all that…?
I don’t understand why the expectation to follow forum rules is offensive to you. on the handfull of times people were banned on this forum it was after repeated rule breaking. Not because they posted something other users disagreed with. Do you prefer moderators moderate less (which is what writing in the forum acomplishes aswell) and instead keep a tally list to more quickly kick everyone out?
Also a discussion is not a soccer, it’s an exchange of ideas. If you go into it as a “fight” you have already lost, and completely missed why there are discussion in the first place.
Oh. Finally, my post got hidden. Then shouldn’t this post be hidden as well? The ongoing genAI debate, vol. 1 - #101 by nephele
Since when is building an entire OS more or less from scratch with a tiny fraction of the resources that Linux has, let alone Windows or macOS, “realistic”?
I don’t intend to use “AI” ever, and not because it “doesn’t work well” or “uses too much resources”. I think those are problems, but even if they were solved, I still would not want to have a machine do a human’s job, and I wouldn’t want to be part of a project that did that, even if we unfortunately have to deal with upstream dependencies that sometimes do.
The AGENTS.md says no AI-generated contributions will be accepted and makes no exceptions here. So I think we do have such a policy?
What forums are you a part of where the moderators aren’t members of the community who participate in discussions? I don’t know of any myself; this is pretty standard for how forums operate.
@nephele explained he meant “general you” and not “specific you”. This can be confusing, I admit, and so it’s generally better to use “one” instead of “you”. I edited the post to use that language instead to make things clearer.
This is true; the “translate this” button on Mastodon is pretty nifty, though I certainly don’t trust it’s results to be particularly accurate now that many translation sources are re-tooling to use LLMs.
But linguistic inclusion is a lower priority for me, personally, than all the reasons to not use LLMs so again I’m ok with a no-LLM-translators policy. Again, though, that is not this forum’s policy.
The policy is that you can use whatever translation services are available to you.
Do you find a referee in a team of a developers in any company?!
A collaborative software project is not a match between two oponents, it’s a single shared project.
A referee is needed to rule between two players / teams / competitives “projects” for the victory, according rules defined before the players even started to play.
Here, the project is started since very long, and nobody is outside the teams / sides fighting, everybody are mostly involved in the same, the quite undefined “Haiku project community”.
Calling for a referee in that context will be like calling a tennis referee to decide on an open source project: to be a neutral outsider, he will need to have no interest himself at all on the open source project, and yet will be accepted by the project community as being the one to decide.
Would you accept that a lambda guy come in and decide instead of yourself in your own project?
Nitpicking over words isn’t a debate; we call that sophistry.You talk about collaboration, but haiku is open source. You’re not just a company team working on a project; you’re part of an open-source initiative that should always welcome newcomers. Sadly, this community lacks the ability to actually accommodate newbies. Happy “project”.
What I meant is we don’t and can’t have a neutral role to arbitrate debates and/or decisions here because by definition everyone willing to spend time and energy to do that will be also involved in the project community and therefore subjective and participating to it already.
You can’t hope from someone to act as a moderator on a forum dedicated to Haiku ecosystem without him/her the be also interested to that topic.
And therefore willing to express his/her views on it.
Do you ever volunteer to moderate a forum of a community you don’t care one bit about and you’re not part of!? And how any decisions of such moderator could get any legitimacy from that community?!
you’re part of an open-source initiative that should always welcome newcomers
About that part, I’m not sure open source implies to always welcome newcomers.
Open source software, AFAIK, is about allow everyone to use, study, modify and redistribute it.
Nowhere it implies that the project itself should welcome everyone to contribute under theirs own conditions instead of following the current project conditions for new contributors.
It seems interesting to hide my post and then display it again to argue back. With this kind of community operation, no one except the haiku developers and administrators will feel safe.