You actually probe my point (which you seem to have missed).
My point was that it was not âthe communityâ that flagged the post, as the system message claims, but just two people (or something?) that disliked the post or the OP for whatever reason. Thatâs deceiving, at the very least.
This is just your personal opinion, which I can respect. But this is certainly not ground for reporting/hiding the post. It was just a joke after all, that many people do get.
Also, as indicated by the number of likes (7) the initial post received before it was hidden, and the responses from a member who actually used the app introduced in the post, it is clear that some of us found the post to be useful. This is another indication that a good number of people donât share your view that the post was distasteful. It would be nice if the moderators take this into account before taking such drastic measures are hiding posts or comments because it is not of their personal liking.
I suppose you are referring to the link issue (correct me if I am wrong). That could have been easily solved by editing the link, and perhaps a warning/reminder to OP.
What worries the most of all is that there is no recourse for appeal, so it always comes down to what a few people (not the community at large) feel is right.
A more community-driven (and less punishing) system could avoid these uncomfortable situations. Perhaps an upvote/down vote based system, where heavily down voted posts/comments are hidden but their title still visible so that those users who choose to read them can do so by clicking on the title. This would actually allow the community to better self regulate the forum, and in the process lower the burden of the moderators too.
I say all of the above in good faith and with total respect for other peopleâs opinions.
This post and its comments are not showing anymore in the âLatestâ posts page of the forum top page. I donât know if this is deliberate or by accident, but it is effectively silencing the voices of the community. So sad for an open source projectâŚ
Where and how would you copy a link from a search result, in which the link text was a URL and the link target was a different URL? I have never seen a search engine generate a link like that. So how could this have happened?
The fact is obvious that the post was not proofread and the links were not checked before posting, since the link does not actually go to where it says it will go to. That by itself makes it very suspicious. Besides as other moderators have pointed out, even if the post was somehow not LLM-generated, it would still have needed to be edited to be allowed to stay up.
I did nothing of the sort. No suspensions were issued or even threatened. I was not even the one to flag the post, and if it hadnât already been reviewed and marked by another moderator, I probably wouldâve just ignored it. He was not banned, and even now there is nothing preventing him from creating a new account if he wanted to.
This is regarding the use of LLMs for translating posts written in another language, not for generating posts. And furthermore, the results indicate there is a majority support for âencourating non-LLM methodsâ (if we combine policy 1 voters with policy 3 voters, as logically makes sense.)
To be honest, I have no idea exactly how or where Andrea copied that link, and it doesnât matter. The fact is that search engines like Google use indirect redirect URLs for tracking and search statistics. When a human copies a link directly from a search results page (especially on mobile devices or via certain browsers), they very often copy this messy tracking URL instead of the clean destination.
This is purely a human error of copying and pasting (although I may be mistaken and it was done intentionally), but nitpicking such trivialities is going too far.
You donât need to hit the âBanâ button to force someone to leave. Publicly accusing a user of lying, deleting their software announcement based on a flawed âhunch,â and conducting a constant witch-hunt creates an environment so hostile that any self-respecting person would choose to invoke GDPR and leave. That is social exclusion, and the responsibility for it lies with those who wield the administrative power.
Ha-ha-ha. No, it does not make sense. Policy 3 is fundamentally about allowing the use of LLM translators (albeit with warnings). Combining a policy that allows choice with a policy that bans it completely, just to claim a majority for your personal stance, is intellectually dishonest.
@3dEyes, a number of your posts were flagged and auto-hidden as a result (not by moderators.) I restored most of them already and will look at restoring the others.
Why would people who voted for Policy 1 be in favor of Policy 2, if they had to choose between that and Policy 3? If thatâs whatâs in dispute, perhaps we should have a ârunoff pollâ, or just ask the people who voted for Policy 1 if they agree?
People who repeatedly and shamelessly break the rules, and brag about breaking the rules gradually lose their rights to have their actions given that benefit.
Honestly, threads like this one make me feel like I never want to go on this forum again. Where does all this hate and distrust come from? Yes, the whole AI topic is important and people have strong feelings about it, thatâs OK. But itâs not an urgent life or death situation, so we can take a little time and all calm down and use our collective intelligence to figure this out. Even if there are conflicting views maybe we can find a way to handle the AI stuff that most of us (or ideally all of us) can live with.
No need for a runoff. The trend is completely clear. Here is a summary of the survey results.
Which policy would you prefer?
Policy 2 - Allow LLMs for translation 42%
Policy 3 - Allow LLMs for translation, but encourage non-LLM methods 42%
Policy 1 - Ban LLMs for translation 16%
Which policy do you like the least?
Policy 1 - Ban LLMs for translation 71%
Policy 2 - Allow LLMs for translation 22%
Policy 3 - Allow LLMs for translation, but encourage non-LLM methods 7%
How much do you like/dislike policy 1, which bans LLMs for translation?
Strongly for 10%
For 6%
Weakly for 3%
Neutral 6%
Strongly against 36%
Against 34%
Weakly against 5%
How much do you like/dislike policy 2, which allows LLMs for translation?
Strongly for 21%
For 31%
Weakly for 10%
Neutral 16%
Strongly against 13%
Weakly against 7%
How much do you like/dislike policy 3, which allows LLMs for translation, but encourages non-LLM methods?
Strongly for 16%
For 23%
Weakly for 7%
Neutral 31%
Strongly against 6%
Against 7%
Weakly against 10%
By any measure, this poll shows that the majority is against a ban, and either are OK with using LLMs or are tolerant to the use of it. How you can spin this as a majority support for âencourating non-LLM methodsâ is a real mystery.
We know you donât like AI, but please be honest and accept that there is a gap between your position and that of the community at large.
Itâs the position with the least opposition, is it not? And itâs a pretty natural compromise position between the âtotal banâ and âtotal allowanceâ. Some portion of the community would really prefer the total ban, while indeed a large majority is opposed to that. So the compromise is the most natural fit, even if by itself it does not have majority support.
There is a vocal minority that are pro LLM as far as I can tell - donât mistake this for the âcommunityâ. Some of us agree wholeheartedly with the position of the moderators and the stance against LLM written posts and software.
From all the question results, I am sure we can agree that the âtotal allowanceâ position is significantly larger than the âtotal banâ one. That is the main trend you can see in this survey.
Does the compromise position sway that trend? Considering that 31% of the people are neutral relative to Policy 3 (the compromise), it does not seem to have enough weight to support a âmajority prefers encourating non-LLM methodsâ, does it?
Well, we may disagree on the AI/LLM issues, and thatâs OK. I am sure, though, that we all agree that we want the community to stick together and not polarize. Right?
As a prominent leader of the project, you have a lot of influence to make sure this is a welcoming place in spite of the differences that may exist among members. After all, Haiku is a very diverse project with people from all over the world, so cultural differences are bound to appear (nobody is the owner of the absolute truth).
So, perhaps you can be the bigger person here, and accept that the community can self-regulate itself without the extreme measures that we have seeing recently. Tolerance goes a long way.
In fact, itâs the complete opposite; it is the anti LLM crowd that is a minority, not the other way around. Please look at the summary of the survey that I posted here.
Nowhere did I talk about software. I am only referring to the use of LLMs for forum posts, which is the topic of this discussion, as is the survey I was referring to. So, no conflation as you claim.
I have read back through your replies, and OK I accept not the LLM written software. But I think this whole discussion was reignited by an LLM generated post, which is not the same thing as a translated post (with the original language underneath). So thatâs a conflation.
It seems very clear that the overwhelming majority of developers,admins,moderators and contributors is against LLMs.
And on the other hand,itâs mostly users who have never contributed anything to the project (with a few exceptions) who demand that rules are changed in their favor,that they are not âcensoredâ (itâs not censorship by the way,this place belongs to the Haiku project and the admins can set the rules however they want) and that theyâre allowed to spam everything with their shiny new toy,the great bullshit machine.
Well,no,there is no law that forbids moderating a forum,also none that demands a moderator or admin has to waste days explaining every single moderation action theyâve taken.
These flamewars which start again and again and again only show that it seems impossible to make compromises with LLM fanboys because nothing seems good enough,and when they cross the line and moderators have to take action,the next flamewar is guaranteed to start.
My suggestion: Itâs really time for a hard ban on anything LLM-related.
Most developers and forum staff seem to be against LLMs,and theyâre those who made the project great and should not be annoyed everytime they look on a project-owned website.
LLM fanboys can have own unofficial communities where they donât annoy anyone who doesnât want to be there.
Make a hard cut,make a few people unhappy,but finally end the drama.
Please!
You are giving a bad name to the community with your ideas that actively suggest repressing people, both in this forum and in the IRC channel. Please consider not sharing your harmful ideas in public community spaces.
A forum comes with moderation. When people are repeatedly ignoring the moderation rules, they get banned. I think you know that very well.
It is not the first time it happens. It will, sadly, not be the last.
Thanks to this, the forum can keep a civil tone and we can have useful discussions, instead of people insulting and trolling each other.
And when several persons flag a post as inappropriate, it gets hidden automatically (this allows the reaction to be very fast). Moderators can then intervene to restore the post if the flagging itself was abusive. Forunately, this happens only very rarely (or at least thatâs how it was when I was a moderator, who knows what has happened in the last week!)
