The current AI/Low effort blocking system doesn't work.

[godlikeion]

I have my tolerances of AI/Low effort reports set to one. I am still getting multiple AI developed/assisted mods per page.

I understand its a community effort that takes time to report low effort mods, But when a mod with 100+ downloads and a description that boils down to "I had Claude write this?" doesn't have a single report by anyone who used it? Please just change the AI content/use disclaimer to a tag rather than a disclaimer. It would be far more accurate and immediate. The tags could also be expanded to follow individual preference for AI use. EX: "AI-Art." "AI-Assisted." "AI-Created"

And also, I know mods are downloaded automatically when you join a server with mods. Please allow the automatic download to forbid AI/Low effort mods too.

[LadyWYT]

Welcome to the game and forums.

10 hours ago, godlikeion said:

Please just change the AI content/use disclaimer to a tag rather than a disclaimer. It would be far more accurate and immediate. The tags could also be expanded to follow individual preference for AI use. EX: "AI-Art." "AI-Assisted." "AI-Created"

A tag alone won't prevent the mod from showing up on the user's database feed though, and slapping a tag on a mod if it gets enough reports also opens up the door to the mod author disputing the tag via tickets(because false positives happen).

A hidden tag isn't foolproof either, but it allows users to filter out stuff they don't want to see.

10 hours ago, godlikeion said:

I have my tolerances of AI/Low effort reports set to one. I am still getting multiple AI developed/assisted mods per page.

Perhaps report the mods that you think are low-effort/AI? You might also consider reporting the issue via the Discord or perhaps a support ticket, if you think the system might have a problem.

Edit: Incidentally, I gave the filters another test, cranking it all the way down to the strictest tolerance of one report before a mod gets hidden. And the filters still work.

 

10 hours ago, godlikeion said:

And also, I know mods are downloaded automatically when you join a server with mods. Please allow the automatic download to forbid AI/Low effort mods too.

This isn't really necessary, in my opinion, and will more likely lead to users getting confused/frustrated when they can't play on the server they want to play on because the mod download fails, in which case they're probably going to blame the game/Anego Studios for the problem instead of checking their own settings/picking a different server to play on.

Edited September 26 by LadyWYT

[godlikeion]

Sorry, I don't believe that I explained the point I was making the best, Allow me to explain again.

On 9/26/2026 at 9:33 AM, LadyWYT said:

A tag alone won't prevent the mod from showing up on the user's database feed though, and slapping a tag on a mod if it gets enough reports also opens up the door to the mod author disputing the tag via tickets(because false positives happen).

A hidden tag isn't foolproof either, but it allows users to filter out stuff they don't want to see.

Mod authors are required to disclose if and how they use AI to make their mods. Simply making them tag their mod instead of a written disclosure would be far more informative than having to sift through the mod page for the disclosure. Which is sometimes a single sentence in the middle of a paragraph. It would also allow people with different tolerances for AI use to decide what they are ok with rather than any AI at all. Someone might be fine with AI assisted/written code but not AI art and vice-versa. Someone might be fine with AI assisting code writing but might have issues with purely AI written code.

To be fair I don't know how the tagging system works, But I assume the mod author tags their own work.

On 9/26/2026 at 9:33 AM, LadyWYT said:

Perhaps report the mods that you think are low-effort/AI? You might also consider reporting the issue via the Discord or perhaps a support ticket, if you think the system might have a problem.

Edit: Incidentally, I gave the filters another test, cranking it all the way down to the strictest tolerance of one report before a mod gets hidden. And the filters still work.

Once again bad communication skills on my part. I wasn't saying the system was broken, I'm saying the system is quite flawed. Yes it blocks a large amount of mods that use AI. However there are still mods with a large amount of downloads that use AI that slip through the gaps. And knowing what counts as an AI mod is based on the person. I report mods that use any amount of AI as AI/Low effort. Including AI art, But others clearly disagree as some popular mods that use AI art are not reported as AI/Low effort content.

This causes an issues that gets back to my tag argument. Popular mods that use AI are going to get more reports by the people who bother too. But some people might not like a fully AI mod, But are fine with an author that uses AI to write code and then reviews it. The reporting system does not address this. Its either blocking AI mods, Or not at all. Plus there's an upper limit for blocking. Once again, Popular mods that use AI may get more than one-hundred reports. The limit for the tolerance. These mods may never be seen by anyone with a tolerance set. And there's no way to address personal preference.

On 9/26/2026 at 9:33 AM, LadyWYT said:

This isn't really necessary, in my opinion, and will more likely lead to users getting confused/frustrated when they can't play on the server they want to play on because the mod download fails, in which case they're probably going to blame the game/Anego Studios for the problem instead of checking their own settings/picking a different server to play on.

Sorry that was one dumb idea on my part, But its a simple fix. When downloading mods rather than blocking it it gives the user a popup showing the mods and the AI content tags. People can allow download anyway if you just want to play. Or ignore it entirely.



And now to properly explain my issues with the current system.

First, Everyone has their own preference like I stated above. The current system has no control on preference. Its based on reports, Not the actual way AI was used.

Second, The tolerances don't work if people use the system properly, Or don't use it. Or are responsible and report mods. If mods aren't reported the issues is simple, People who have tolerances have to have it at one to five reports to actually block any mods. Or mods are reported too much, And the tolerance scale may as well become a toggle.

Third, People may be hesitant reporting mods. To most people reporting a mod makes it feel like the mod may be taken down, Not that your just letting the system know AI was used. Some people may not think the authors use of AI counts as AI/Low effort. Once again. AI art, Is AI art something a mod should be reported for? Is AI art fine? Personal preference may cause mods that to some count as AI use to be unreported. And most people don't report. Meaning that for the most part the only people who report mods are the ones with active tolerances.



Id like to apologize for the poor quality of my original post. I didn't explain my points well. I was tired and I was simply frustrated that I had to go through and report mods, Questioning if it counted as AI/Low effort. Seeing a mod with one-hundred downloads admit to being fully vibe coded yet not having a single report was the last straw for me. In a huff I made my first statement and didn't really think it through. I hope what I've written above clarifies my opinion on the matter.

 

[LadyWYT]

1 hour ago, godlikeion said:

Mod authors are required to disclose if and how they use AI to make their mods.

The mod database doesn't require such disclosure. https://mods.vintagestory.at/terms

Most users, however, are polite enough to include AI disclosure in their mod description though.

 

1 hour ago, godlikeion said:

It would also allow people with different tolerances for AI use to decide what they are ok with rather than any AI at all.

There's already a filter feature for this in your database profile settings, and it can be adjusted in regards to how many reports a mod has to have received before it gets filtered out. There's also a report button to report mods for using AI or otherwise being low effort. If it's not reported, the filters won't catch it.

 

1 hour ago, godlikeion said:

Third, People may be hesitant reporting mods. To most people reporting a mod makes it feel like the mod may be taken down, Not that your just letting the system know AI was used. Some people may not think the authors use of AI counts as AI/Low effort. Once again. AI art, Is AI art something a mod should be reported for? Is AI art fine? Personal preference may cause mods that to some count as AI use to be unreported. And most people don't report. Meaning that for the most part the only people who report mods are the ones with active tolerances.

I mean, I was messing around with the filters the other day, and as far as I could tell, setting the tolerance to the strictest possible setting caught most AI content. Some stuff still slips through the cracks, because not everyone is gonna look at every single mod, so it can take some time for something to actually get reported. However...that's also why the report button exists. Some people don't really care about making reports, this is true, however the fact remains that if you see content that you think qualifies as AI or low-effort...you should be reporting it if it really bothers you that much, when you encounter such mods. Otherwise, the mod isn't going to get filtered out.

 

1 hour ago, godlikeion said:

Simply making them tag their mod instead of a written disclosure would be far more informative than having to sift through the mod page for the disclosure.

I will note that a tag system requires someone to actually enforce the tagging rule. Which means someone has to be looking through the mod's code to see whether or not AI was actually used. Reports alone aren't enough, because false reports happen for various reasons, and mod authors are also gonna challenge the tags if they feel their mod was tagged unfairly. Needless to say, manually checking mods requires more time and effort, which means time and effort that's not being put towards actually developing the game.

Edited September 28 by LadyWYT
Wrong link

[Teh Pizza Lady]

Who appointed you the judge of what counts as acceptable? Low effort has nothing to do with how the code was written. Plenty of hand-coded mods are half-baked, and plenty of AI-assisted mods went through many guided development passes and are better than most of what's out there. The report button exists for any low-effort mod regardless of how it was made, and the moderators have already said false reports carry severe penalties, so what actually gets reported is the low-effort stuff. That is the system working as intended.

Your tag idea also has a fatal flaw: nobody can verify it. Code is code. It works or it doesn't. Once any telltale comments are stripped, and an AI tool can strip them itself, there is no way to tell whether a mod was AI-assisted, hand-written, or something in between. A tag system would only measure who is willing to volunteer that they used AI, which punishes honest authors and rewards the ones who say nothing. That is the opposite of what you're asking for.

And the bigger issue is the attitude behind it. You already have a filter you can set to whatever tolerance you like, so your own experience is fully in your hands. Asking for the game to block or gate what other people can download and play in their own sandbox is a different thing entirely. You can decide what you install. You don't get to decide what everyone else is allowed to.

Plus, if you decline mods from a server that were generated using AI and one of those mods is core to the function of the server, then you will not be able to join the server entirely... but I'm sure you'll be okay with that, especially if it's one of those super popular servers that EVERYONE is playing on... but you.  

[Rainbow Fresh]

Wow, this went into questionable territory real fast despite being the most objectively sane discussion including the word "AI" I have seen so far. OP has a point and somehow you didn't address it at all as LadyWYT just went repeating what OP already stated as if it answers a question never asked and Teh Pizza Lady just dogpiled as if OP insulted their personal taste.

What is stated and the concept behind are quite simple and in my opinion pretty logical and straight forward: Tags and filters instead of some washy report threshold system. And I wholeheartedly agree with it. Not just for AI/Low Effort related stuff. The ModDB currently has a very barebones tag system and search functionality. You can look for some stuff, but you cannot look for not X stuff. Which may be due to the fact that tags in general are somewhat useless as they currently are and there wouldn't be anything to filter out in the first place and aside from one or another known vibe coder people could have a gripe with (and a certain individual center of a certain recent controversy) no reason to exclude author/contributor from the search with all other filter options being mutually exclusive categories to search in in the first place. But the lack in such option could and should be addressed and, AI related or not, OP has made a good example for why that is and why the simplistic report threshold system isn't cutting it at best and making it worse at worst.

3 hours ago, LadyWYT said:

There's already a filter feature for this in your database profile settings, and it can be adjusted in regards to how many reports a mod has to have received before it gets filtered out. There's also a report button to report mods for using AI or otherwise being low effort. If it's not reported, the filters won't catch it.

OP already stated how the system works, yes. The point tried to be made here is that this system is fundamentally flawed/insufficient. You can report. And, minus whatever may or may not happen behind the scenes for such a report to count, this contributes to a user-set threshold. If a mod has more reports than you "allow to see", the mod is hidden. That part functionally works right now. The problem brought up here is that this system does not allow for fine-tuned personal preference. The more reports a mod gets, the fewer people will see it until, with enough popularity and "honesty" of the reporting users, a mod becomes invisible to anyone filtering *anything* due to numbers alone, numbers based on other people's decision of what you do and don't want to see in AI/Low effort content. Additionally, reporting is a weird action for a filter system. Reporting innately feels like "This person did something wrong and needs to be punished for it". You report scams and phishing attempts on Discord. Stolen identities on Steam. Bad actions by bad actors you expect to get nuked by Moderation. So "reporting" a mod for AI/Low Effort content would carry the same undertone of "This author did bad and their mod needs to be removed/restricted", not "I want to let my fellow ModDB users know this is AI/Low Effort content so they can filter it out". A tagging system with a rule-given "enforcement" to properly tag your stuff would have the exact same result (with more detail) than the current system of reporting and "Don't make false reports or else", just with a more positive spin on it. Which brings us to

3 hours ago, LadyWYT said:

I will note that a tag system requires someone to actually enforce the tagging rule. Which means someone has to be looking through the mod's code to see whether or not AI was actually used. Reports alone aren't enough, because false reports happen for various reasons, and mod authors are also gonna challenge the tags if they feel their mod was tagged unfairly. Needless to say, manually checking mods requires more time and effort, which means time and effort that's not being put towards actually developing the game.

Correct, but how does this differ from the current system? You report something (which by design of a "report" function would just notify moderation to check it anyway) and "False reports get punished", meaning that someone HAS to check it, verify it, and take action if it is clearly wrong. The end result doesn't change if we have a detailed tagging system and a rule that you need to tag your stuff properly (even outside of the typical AI discussion). The workload would however change as now there isn't a report to verify per person reporting per AI/Low Effort mod. But instead just reports, if any, on someone notably so mis-tagging/ not tagging their mod. And if people are already kind enough to put AI disclaimers in their text, they can also be kind enough to just put a tag instead. Which, as OP secondarily mentioned, opens up the option to more granular control than "AI yay or nay". A win-win in my eyes.

 

Now for the unnecessarily hostile response to a valid idea by a person that already admitted to and apologized for phrasing things badly out of personal emotion.

3 hours ago, Teh Pizza Lady said:

Who appointed you the judge of what counts as acceptable? Low effort has nothing to do with how the code was written. Plenty of hand-coded mods are half-baked, and plenty of AI-assisted mods went through many guided development passes and are better than most of what's out there. The report button exists for any low-effort mod regardless of how it was made, and the moderators have already said false reports carry severe penalties, so what actually gets reported is the low-effort stuff. That is the system working as intended.

Who is taking the role as judge for anything here? The very idea is to NOT have people dictate stuff for other people by NOT having a mere number of people reporting X thing remove it for everyone else with vaguely Y setting. But instead just having tags, that everyone, themselves, on their own volition, can decide to hide for whatever reason from their own search results. Be it a blacklist kinda "I hate AI" thing or simply just "I want to check if there is X category mods that aren't also Y". You know, for people looking to expand their modlist in a very specific direction. Do you use or have used Nexusmods at some point? Imagine looking for a mod that has some name you search for, but filter out translations cause you want to find the mod, not the Brazilian, Russian, Chinese or French translation of said mod, which would show up first as they were made/updated more recently than the mod itself. It's utility for personalization and the very opposite of someone taking on the role of judge for others. It does also, natively, apply to many more things than AI and/or Low Effort discussions. Granted, noone would ever voluntarily put a "Low Effort" tag on their mod and we shouldn't go around judging people's skill at coding or mod making by reporting things so moderation puts the dunce hat on them, that is a whole 'nother topic - and as you correctly stated, any human can make lower effort slop than AI puts out if prompted with general AI code being extremely hard, if at all, provable - but a simple personal mod/author blacklist could do for such things.

3 hours ago, Teh Pizza Lady said:

A tag system would only measure who is willing to volunteer that they used AI, which punishes honest authors and rewards the ones who say nothing. That is the opposite of what you're asking for.

I don't think it punishes honest authors. Because if you would need/want to hide certain information, knowing that it would piss certain people off, just so that these certain people still download your mod in ignorance you quite likely have very questionable goals/morals. We don't earn money making mods so download numbers don't matter. If people like your stuff they use it. If you did ask Claude a single question and that rubs someone the very wrong way and they hate your guts and want nothing to do with you, their problem. Using a tagging system combined with a personal blacklisting feature as was already suggested in other threads before would make this whole ordeal much easier and much nicer for everyone involved - moderation doesn't have to confirm hundreds of reports for false positives, the variety of tags allows fine-tuned filtering per personal preference or use case (as OP stated the example of separation AI assistance from full Vibe Coding, and AI code from AI art - the latter of which obvious and not able to be hidden) and the honest authors stay un-harrassed by the masses of "Eww you used AI go kys" comments. Or if not, atleast reporting *those* will be much easier and straightforward because with the tools given to them they have virtually no reason to be here and be that toxic, and people actively seeking to be bad actors by purposefully mis-tagging or not tagging their mod can still be dealt with the same way the current system requires us to: report and wait.

3 hours ago, Teh Pizza Lady said:

And the bigger issue is the attitude behind it. You already have a filter you can set to whatever tolerance you like, so your own experience is fully in your hands. Asking for the game to block or gate what other people can download and play in their own sandbox is a different thing entirely. You can decide what you install. You don't get to decide what everyone else is allowed to.

Plus, if you decline mods from a server that were generated using AI and one of those mods is core to the function of the server, then you will not be able to join the server entirely... but I'm sure you'll be okay with that, especially if it's one of those super popular servers that EVERYONE is playing on... but you.  

Granted, as OP already retconned after LadyWYT's response, this whole blocking downloads thing is a bit unnecessarily far. But there is still no reason to be so offended about it - it would still be a per-user system based on personal preferences. "But I'm sure you'll be okay with that, especially if it's one of those super popular servers that EVERYONE is playing on... but you". Eh... yeah? If someone, personally, voluntarily goes so far as to enable an optional setting in *their own* game to block server mod downloads from mods tagged AI XYZ then they surely value not having anything touched by AI on their PC, ever, over playing the game on a server. It solely affects themselves and their ability to have fun with others and, as you yourself started your reply with, who makes you the judge about how these people want to play (or not play) their game? As per the updated idea from the OP, making it rather a "Hey, the server has mod XYZ that has your blocked tag ABC, do you want to proceed?" is better than just silently failing and being like "Oop, can't play on this server sorry bucko" but the idea is no worse and no more offensive than the tagging system in general.

 

So Tl;dr, a better tagging and accompanying filter system would be a very nice, very welcome addition to the ModDB. I am one of the honest people that would *love* to tag my mods, but everytime I uploaded one and tried I found the list of available tags so lacking that none would apply other than the category-redudant "It's a content mod". Having vastly more tags available or, even better, the ability to set custom tags would open up a very convenient system for people. And imho, any filter system needs a blacklist mode aswell. We can search for tags, why can't we also search for "Not tag X"? Outside of personal AI crusades, the ability to narrow down specific searches is a great tool. Nexusmods has such a comprehensive tag-and-filter system and it's one of the best features for me as mod user there. It would fully address the imo valid complaints OP has about the current AI reporting approach we have and I do not see how the supplied counter-complaints would actually apply here.

[LadyWYT]

7 hours ago, Rainbow Fresh said:

Outside of personal AI crusades, the ability to narrow down specific searches is a great tool.

I don't mind more tags, but I'm skeptical that a tagging system alone really fixes anything, chiefly because of personal AI crusades. If that's why I come across as harsh in some responses, that's why. There are several users that don't really seem to want to use tags or filters, as much as they want to ban anything remotely related to AI and demonize others who used such tools. 

 

7 hours ago, Rainbow Fresh said:

As per the updated idea from the OP, making it rather a "Hey, the server has mod XYZ that has your blocked tag ABC, do you want to proceed?" is better than just silently failing and being like "Oop, can't play on this server sorry bucko" but the idea is no worse and no more offensive than the tagging system in general.

As I was trying to say before, the reason I don't think this is really better, is that something like this will likely lead to several players complaining about not being able to play on the server they want to play on, because there's mod(s) with the "AI content" tag. To me it seems that if a player is incredibly concerned about what mods are running on a server, they should check the modlist themselves ahead of time in addition to other rules and conditions that may apply to that server as well. That way they know exactly what they're signing up for.

I suppose it's a similar case for mods themselves. Most mod authors will provide an AI disclaimer in their mod description, so the user generally still has prior warning if something slips through the filter. And it seems to me that it should be common sense to read the mod description before you download the mod, so that you have a better idea of what you're getting into.

In any case, apologies to the OP if I was overly harsh. A specific "AI Content" tag itself would not be a bad thing to have, but I don't believe that it will prevent a user from seeing any AI content at all, and I'm under the impression that's not what most users(not the OP themselves necessarily) are actually after when they're complaining about AI content(aside from wanting it purged from the database entirely). You're still going to have things slip through the cracks, and as @Teh Pizza Lady noted, it may not be possible to manually verify if something used AI content or not since it's possible to cover the AI's tracks.

[Teh Pizza Lady]

9 hours ago, Rainbow Fresh said:

Now for the unnecessarily hostile response to a valid idea by a person that already admitted to and apologized for phrasing things badly out of personal emotion.

Who is taking the role as judge for anything here? The very idea is to NOT have people dictate stuff for other people by NOT having a mere number of people reporting X thing remove it for everyone else with vaguely Y setting. But instead just having tags, that everyone, themselves, on their own volition, can decide to hide for whatever reason from their own search results. Be it a blacklist kinda "I hate AI" thing or simply just "I want to check if there is X category mods that aren't also Y". You know, for people looking to expand their modlist in a very specific direction. Do you use or have used Nexusmods at some point? Imagine looking for a mod that has some name you search for, but filter out translations cause you want to find the mod, not the Brazilian, Russian, Chinese or French translation of said mod, which would show up first as they were made/updated more recently than the mod itself. It's utility for personalization and the very opposite of someone taking on the role of judge for others. It does also, natively, apply to many more things than AI and/or Low Effort discussions. Granted, noone would ever voluntarily put a "Low Effort" tag on their mod and we shouldn't go around judging people's skill at coding or mod making by reporting things so moderation puts the dunce hat on them, that is a whole 'nother topic - and as you correctly stated, any human can make lower effort slop than AI puts out if prompted with general AI code being extremely hard, if at all, provable - but a simple personal mod/author blacklist could do for such things.

I don't think it punishes honest authors. Because if you would need/want to hide certain information, knowing that it would piss certain people off, just so that these certain people still download your mod in ignorance you quite likely have very questionable goals/morals. We don't earn money making mods so download numbers don't matter. If people like your stuff they use it. If you did ask Claude a single question and that rubs someone the very wrong way and they hate your guts and want nothing to do with you, their problem. Using a tagging system combined with a personal blacklisting feature as was already suggested in other threads before would make this whole ordeal much easier and much nicer for everyone involved - moderation doesn't have to confirm hundreds of reports for false positives, the variety of tags allows fine-tuned filtering per personal preference or use case (as OP stated the example of separation AI assistance from full Vibe Coding, and AI code from AI art - the latter of which obvious and not able to be hidden) and the honest authors stay un-harrassed by the masses of "Eww you used AI go kys" comments. Or if not, atleast reporting *those* will be much easier and straightforward because with the tools given to them they have virtually no reason to be here and be that toxic, and people actively seeking to be bad actors by purposefully mis-tagging or not tagging their mod can still be dealt with the same way the current system requires us to: report and wait.

Granted, as OP already retconned after LadyWYT's response, this whole blocking downloads thing is a bit unnecessarily far. But there is still no reason to be so offended about it - it would still be a per-user system based on personal preferences. "But I'm sure you'll be okay with that, especially if it's one of those super popular servers that EVERYONE is playing on... but you". Eh... yeah? If someone, personally, voluntarily goes so far as to enable an optional setting in *their own* game to block server mod downloads from mods tagged AI XYZ then they surely value not having anything touched by AI on their PC, ever, over playing the game on a server. It solely affects themselves and their ability to have fun with others and, as you yourself started your reply with, who makes you the judge about how these people want to play (or not play) their game? As per the updated idea from the OP, making it rather a "Hey, the server has mod XYZ that has your blocked tag ABC, do you want to proceed?" is better than just silently failing and being like "Oop, can't play on this server sorry bucko" but the idea is no worse and no more offensive than the tagging system in general.

To clear up what I meant by "who appointed you judge": it wasn't about tags. It was about the original post, where the complaint was that certain mods with 100+ downloads hadn't been reported because they were AI-assisted. The report form lists "Very low effort / AI-generated content" as a single reason, which is part of the problem, since it treats how a mod was made and how good it is as the same question. They're not. A mod can go through plenty of guided AI-work sessions and still be good, and a hand-coded one can be half-baked. As I understand it, the moderators have said the report button is for low effort mods, not for anyone who used AI as a tool along the way. It's not the OP's place, or mine, to decide that for everyone else based on how a mod was made.

I do see the appeal of tags as a personal filter, and I'm not against more tags or a "not X" search for people who want to narrow things down, the way Nexusmods does. My worry is narrower: those filters only work if authors have a reason to tag honestly. It's true that modders don't earn money, but audience still matters to the people making the mods, and I'd expect a visible AI tag to become the thing people aim their frustration at, not a way to reduce it. The moderators have also said they don't plan to tag mods for AI, and if a tag shrinks your audience, the practical move is to stop disclosing, and honestly I'd probably do the same. That's my main worry: a tag would end up measuring who is willing to say they used AI, not who actually used it, so the honest authors get filtered and everyone else stays invisible.

For me, code is code. It either works or it doesn't, and where a given piece came from doesn't change what it does. In my case I consulted an AI tool while going through the game's source to work out where my patches needed to go, and I don't think that makes the mod any less deserving of a fair platform. Judging work by where it came from instead of how well it performs is a mistake history has made before, and I'd hate to see it repeated here.

I also want to be clear that I'm not directing this at anyone in this thread. I know there are people elsewhere in the community who would like AI-assisted mods segregated or banned, and that's what concerns me. If someone doesn't want to use AI products, that's completely fine, and the filter they already have lets them do that. It just seems like the person with the stricter preference should be the one doing the extra work. You want something different from everyone else, put in the work to make it that way. That's how mods get created, after all. Mods should be judged by what they do, and the low effort filter already covers what most people care about.

TL;DR: Use of AI doesn't automatically mean that a mod is low-effort. The quality of a mod is not determined by the number of downloads it has, nor by whether the author used AI to create it, nor by how extensively the AI was used. The quality is determined by the content of the mod and whether it works, and that is how mods should be judged. A voluntary AI tag also gives authors a reason to stop disclosing, so it would end up measuring who is willing to say they used AI, not who actually did. Judging mods by how they were made instead of what they do is the part I disagree with.

Edited September 28 by Teh Pizza Lady

[Dilan Rona]

For the time being I'm making extensive use of AI. And compared to ChatGPT, Copilot, and Workik; the one I am using is far superior. My first atempt was using signs (that needed a lot of harmony patching to actually detect the signs, and the writing on them, and that was with copilot assisting chatgpt). Eventually I gave up, tried another one, and it had a proper mod going on the second recompile.

Busy with a second mod for visuals (thats at a point where I can take over myself, and use AI for assistance if I get stuck), with a third mod that i'm thinking of making. Content mods is easy enough to start maintaining myself now, now that I have a proper understanding of how they work. Anything C# related will take a bit longer for me to get to grasp with properly, and I would just as soon have mods thats of a high enough quality (even if its AI generated) that there is little to no issues, bugs, or crashes present, and as I get better I can start doing it more myself, and rely less on the AI.

I know how everyone here stands regarding AI (for and against, and I myself am absolutely against having AI shoved into my face in everything and on every app and device. If I want to use it, I'll look for one to use that can do the job.), and all I can say is to each his own. Love it or hate it. Its here. From my point, it mustn't get to the point where we stop thinking for ourself how to do anything, especially coding. And especially with children, who absolutely have to be able to think and reason for themselves, and not rely on AI tools.

End of the day, thats what it is. a tool. And it can be used for good, or bad. Just like any other tool.

[LadyWYT]

Incidentally, a better option would be to just add a feature that lets users ignore mods based on mod author, similar to how you can ignore users and their content on the forums. Because I mean, chances are if the user is that strict about AI content, they aren't going to want to see anything else from a mod author that's okay with using AI. It's also the kind of feature that's useful for more than just AI content, since there are other various reasons that a user might want to filter out content from specific mod authors, and it can't be bypassed as easily as a tag can, since in order to upload a mod your name has to be attached to it.

[Dilan Rona]

but as stated before by others, there are servers that could be relying on such mods. And the game will download the mods regardless. It will be up to the user to decide if they want to continue to join, or to stay on said server.

And I for one have a minimum standard regarding quality that I will not upload unless it matches exactly how I need the mod to work, and will not cause problems for the user. Or conflicting with other mods for that matter.

[Zaph42nd]

13 minutes ago, LadyWYT said:

Incidentally, a better option would be to just add a feature that lets users ignore mods based on mod author, similar to how you can ignore users and their content on the forums. Because I mean, chances are if the user is that strict about AI content, they aren't going to want to see anything else from a mod author that's okay with using AI. It's also the kind of feature that's useful for more than just AI content, since there are other various reasons that a user might want to filter out content from specific mod authors, and it can't be bypassed as easily as a tag can, since in order to upload a mod your name has to be attached to it.

For sure in favor of this, even regardless of the AI discussion extra filtering options make it easier to find specific mod author published mods (or contributed mods) without needing to necessarily always visit an author's page.... Maybe be interesting to see if in the future we can follow/block authors on moddb too 

[Teh Pizza Lady]

7 minutes ago, Dilan Rona said:

but as stated before by others, there are servers that could be relying on such mods. And the game will download the mods regardless. It will be up to the user to decide if they want to continue to join, or to stay on said server.

And I for one have a minimum standard regarding quality that I will not upload unless it matches exactly how I need the mod to work, and will not cause problems for the user. Or conflicting with other mods for that matter.

yeah obviously if the mod is full of bugs that the author can't or doesn't know how to fix then it should be marked as low-quality. I'm okay with that as long as we're not docking points on quality just because the mod author got help from a digitized assistant.

[LadyWYT]

4 minutes ago, Dilan Rona said:

but as stated before by others, there are servers that could be relying on such mods. And the game will download the mods regardless. It will be up to the user to decide if they want to continue to join, or to stay on said server.

Right, but when it comes to servers, it's on the user to do their own research on what mods the server uses and what rules are enforced before they join. So if a user has stricter preferences, they're the ones that need to do the extra legwork when it comes to figuring out if a server supports their playstyle. To my knowledge, games in general don't offer players the option to filter servers by specific mods; they just allow players to sort servers via game mode and 18+. Anything more feels like an unnecessary hurdle and likely to generate confusion and complaints.

When it comes to mod dependencies though, I think it's acceptable to filter out mods that have an ignored mod author's mods listed as dependencies. It saves the player some hassle when it comes to building their personal modlist. Players are still gonna complain with this system, of course, since inevitably they're going to be excluding some popular mods by setting filters that restrictive. But I mean, at that point that's entirely on them.

[Dilan Rona]

4 minutes ago, Teh Pizza Lady said:

I'm okay with that as long as we're not docking points on quality just because the mod author got help from a digitized assistant.

the problem is there are people that will do that just because of that fact, or if they think (with no evidence) that it was. It will only hurt the modding community in the long run. Only reason I resorted to AI end of the day is the fact I couldnt get anyone that was willing to do the coding for me (or in the one case where I did, wanted to charge what amounts to over a quarter of my monthly salary for a mod, where he himself was planning to use an ai to make the mod). So in a way, the people that wasnt willing to assist is to thank for me resorting to relying on an AI to get the job done properly.

[Teh Pizza Lady]

6 minutes ago, Dilan Rona said:

the problem is there are people that will do that just because of that fact, or if they think (with no evidence) that it was. It will only hurt the modding community in the long run. Only reason I resorted to AI end of the day is the fact I couldnt get anyone that was willing to do the coding for me (or in the one case where I did, wanted to charge what amounts to over a quarter of my monthly salary for a mod, where he himself was planning to use an ai to make the mod). So in a way, the people that wasnt willing to assist is to thank for me resorting to relying on an AI to get the job done properly.

And see, based on that statement, I'm curious what the mod is, what it does, and why. But some people in the community would want it sequestered away from the public eye just because you chose the route you did, and that's where I have a problem. It reminds me too much of all the scientific contributions from the 50s and 60s that either got ignored or improperly attributed just because the ideas came from a woman or a black American.

[Dilan Rona]

people cant complain over the use of ai when the same people refused to help others when they needed modding work done for their ideas. eventually it gets to the point where they realise its better to do it themselves. Took me a while to get fed up enough to get to that point though. and in my eyes for the best. i got a mod that does EXACTLY what I need it to do. Got a second one thats well underway (and i can finally take over myself), with another 3-4 being planned.

[Feyd-Rautha]

I'd like to add a mod author's perspective, because I think godlikeion and I want the same thing from opposite sides of the issue.

I use AI tools in my mod development, and I've been open about it from the start. My first releases carried an AI disclaimer in the description. When an AI tag became available, I switched to the tag because it was clearer and easier for players to find. A few weeks ago I added links to the source code for every one of my mods, so anyone can see exactly what they're installing. I've also never deleted a player's comment on any of my ModDB pages, including the critical ones.

The AI tag has since been removed. The closest remaining option is "Entity Logic (AI)," which describes creature behavior, not how a mod was made. It confuses players and doesn't tell them what they want to know. For now I'm going to put my written disclaimer back into my descriptions, but as godlikeion pointed out, a sentence buried in a description is easy to miss.

On LadyWYT's point that tags would need someone to enforce them: I think there's a middle ground that needs very little enforcement. Give authors voluntary, self-applied tags, such as AI-assisted code, AI-generated art, and AI-generated content. Authors who are already disclosing would use them, and players could filter on them immediately, without waiting for reports to build up. The report system would still be there for anyone who doesn't disclose. Right now the only way an honest disclosure reaches the filter is through "AI/Low effort" reports, which treats "this author told you they used AI" the same as "this mod is low effort." Those aren't the same thing, and a self-applied tag separates them.

I'd also ask for two additions:

1. Filter by AI tag. Let players hide every mod tagged as AI-assisted, or only the categories they care about. Someone who's fine with AI-assisted code but not AI art should be able to choose that.
2. Block individual authors. If a player doesn't want to see my work, they should be able to hide it entirely with one click.

This would help mod authors as much as players. Right now, every update I post brings a new round of anti-AI replies in the comments. Players end up arguing with each other, and the people who actually want help installing or troubleshooting the mod get buried. The comment section should be for the people who want to play the mod. If players who don't want AI-assisted content could simply never see it, everyone would get a better experience, including the players who dislike AI.

I'm not complaining about the criticism itself. People are entitled to their views on AI. I just think clear labels and real filtering would take the heat out of these arguments, and I'd welcome any system that lets players decide for themselves with accurate information.

[Tj Pepler - Critcher]

First I would like to know which mod programmer (code mods) you know that doesn't have any AI influenced tools in their workflow?  Can you name a single mod? No.
Then I would like very much for an AI protester to show me a workflow and tools I or anyone could use to create these mods for you for free as we do?  What can I use as my OS?  I don't even know of a AI free OS, but after you solve that, what IDE do I use that is not AI???  Can I go to coding sites and copy code?  What if that was made by AI and I don't know? I will need a new language and compiler because .NET is of course heavily developed by AI systems.  Gee that makes me wonder how this very game is compiled...  Imagine how dumb someone would sound trying to prevent AI from being used in a mod for a game in which AI was used in development...  that would be something eh?

So let's make a rule then, if you want to complain about AI use, then you provide the author with the tools and workflow to produce it to your standards.  (hint : no such thing exists anymore)
 

[Dilan Rona]

Excellent point there, @Tj Pepler - Critcher. Couldnt have said it better myself. And its shoved in literally everything now. Where its not shoved into something, it was used to develop something.

[Rainbow Fresh]

15 hours ago, LadyWYT said:

[...] the reason I don't think this is really better, is that something like this will likely lead to several players complaining about not being able to play on the server they want to play on, because there's mod(s) with the "AI content" tag. To me it seems that if a player is incredibly concerned about what mods are running on a server, they should check the modlist themselves ahead of time in addition to other rules and conditions that may apply to that server as well. That way they know exactly what they're signing up for.

I suppose it's a similar case for mods themselves. Most mod authors will provide an AI disclaimer in their mod description, so the user generally still has prior warning if something slips through the filter. And it seems to me that it should be common sense to read the mod description before you download the mod, so that you have a better idea of what you're getting into.

You are certainly correct that the onus should always be on the users to inform themselves. However a system like this could still be another friendly step of holding out a hand, especially since all of the information is already there - the server already sends the mod list, the client already loads it from the ModDB itself where it could also get the information about the mod's tags, and the user is already logged in (was the game account the same as the ModDB account or the forum account?) to fetch the personal blacklist.

What if there is a comparatively smaller server that doesn't have a descriptive presence informing about the specific modlist prior to joining? Much like not everyone writes an informative essay in their mod description. What if the person just wants to try a new server out "real quick", dipping in to see what's up before committing to playing long term on what turns out to be a rather small world? If such a warning was automatically given (specifically warning, not prevention) that could save them some trouble. And there is always a chance that an automated warning telling them there is a mod they won't like on this server could prevent them from getting pissy and angry after the fact at the server owners. Of course, some people will always complain - if not about one thing, about another - but no system other than moderation telling them to stay civil will help against that.

14 hours ago, Teh Pizza Lady said:

It was about the original post, where the complaint was that certain mods with 100+ downloads hadn't been reported because they were AI-assisted. The report form lists "Very low effort / AI-generated content" as a single reason, which is part of the problem, since it treats how a mod was made and how good it is as the same question. They're not. A mod can go through plenty of guided AI-work sessions and still be good, and a hand-coded one can be half-baked. As I understand it, the moderators have said the report button is for low effort mods, not for anyone who used AI as a tool along the way. It's not the OP's place, or mine, to decide that for everyone else based on how a mod was made.

Fair. It does showcase how, if not a more advanced tagging system being "better", the current system is lacking though. It's not only overly simplified it is also subjectively debatable if most people are already using it "wrong" just because the word "AI" shows up in it. I am aware this was mostly just added to quench all the loudly voiced demands to ban AI in some capacity but that doesn't mean it can't be done better still.

14 hours ago, Teh Pizza Lady said:

those filters only work if authors have a reason to tag honestly. It's true that modders don't earn money, but audience still matters to the people making the mods, and I'd expect a visible AI tag to become the thing people aim their frustration at, not a way to reduce it. The moderators have also said they don't plan to tag mods for AI, and if a tag shrinks your audience, the practical move is to stop disclosing, and honestly I'd probably do the same. That's my main worry: a tag would end up measuring who is willing to say they used AI, not who actually used it, so the honest authors get filtered and everyone else stays invisible.

Personally, I see it the opposite way. As long as there is some form of backing security net - which in its simplest form just being active moderation against people being assholes in the comments - a tagging system that is being used in honest detail with the accompanying filter options for users should foster a healthier atmosphere. If people correctly tag their work for what it is, people who don't like can just filter out that tag and everyone is happier. Who deliberately seeks out "AI XYZ" tags just to harass the author in the comments - gets timed out or banned, simple as that; generic netiquette rules should already be enforcing this, filter system or not. A filter system would just further nail that down because "Bro, just blacklist it if it bothers you" becomes a valid response. You said "audience still matters" which I see this being the best result for that; it's not the quantity of users but the quality - which I don't mean derogatory against anyone but more in the veins of "Better have fewer people actually caring about the mod than a bunch of trolls just for the numbers". Again, there is no money involved so download numbers don't matter outside of rather niche leaderboard comparisons. And while I am certainly not pure myself in that regard, one shouldn't determine their hobby based on numbers. If you like making mods, make mods regardless of if 1 or 1,000,000 people download it.

So outside of the concerns of "losing audience to honesty", everyone else profits from proper tagging. People can more easily find the mods they specifically look for. People can exclude the things they don't want to see. Authors benefit from having a "better" audience by offering the necessary tools for people to do this filtering themselves. Moderation has less work as they don't need to sort through dozens if not hundreds of reports for people wanting AI mods "tagged" for them. What remains is normal levels of moderation work against trolls that no system could satisfy, and the still open question of what form of moderation rules are necessary to enable better use of such a tagging system - as enforcing "tag correctly" is about as feasible as banning any and all AI use ever.

P.S. On reading further, Feyd-Rautha already mentioned many of the points I made here, but alas.

[LadyWYT]

7 hours ago, Rainbow Fresh said:

What if there is a comparatively smaller server that doesn't have a descriptive presence informing about the specific modlist prior to joining? Much like not everyone writes an informative essay in their mod description. What if the person just wants to try a new server out "real quick", dipping in to see what's up before committing to playing long term on what turns out to be a rather small world? If such a warning was automatically given (specifically warning, not prevention) that could save them some trouble. And there is always a chance that an automated warning telling them there is a mod they won't like on this server could prevent them from getting pissy and angry after the fact at the server owners. Of course, some people will always complain - if not about one thing, about another - but no system other than moderation telling them to stay civil will help against that.

Yeah, no, sorry, but I think the responsibility should still fall on the player to do their own research before joining a server if AI is something they're terribly worried about. If the modlist isn't posted, they can always message the server owner/admins, but otherwise I think it's fair to point out that if the server isn't advertising the modlist they're using, there's probably other lurking issues and it's not a particularly good choice of server.

It also very much feels like turning responsibility over to Anego Studios for protecting players from servers, or at least opening the door for such, rather than placing the responsibility of decision-making on the player. If there has to be any kind of warning, I think it's more appropriate to post a blanket warning for all servers stating that the multiplayer experience isn't rated, and that people may say things you don't find agreeable or the server may use mods that you find problematic(basically, multiplayer is "use at your own risk"). It's still not the most ideal thing, but a warning box you only have to deal with maybe once is better than multiplayer itself getting bogged down by a bunch of a different rules and checkpoints that caters specifically to players with incredibly strict preferences. Because quite honestly, I don't expect those players to do their own server research when the system will just do all the work for them.

[Teh Pizza Lady]

11 hours ago, Rainbow Fresh said:

Personally, I see it the opposite way. As long as there is some form of backing security net - which in its simplest form just being active moderation against people being assholes in the comments - a tagging system that is being used in honest detail with the accompanying filter options for users should foster a healthier atmosphere. If people correctly tag their work for what it is, people who don't like can just filter out that tag and everyone is happier. Who deliberately seeks out "AI XYZ" tags just to harass the author in the comments - gets timed out or banned, simple as that; generic netiquette rules should already be enforcing this, filter system or not. A filter system would just further nail that down because "Bro, just blacklist it if it bothers you" becomes a valid response. You said "audience still matters" which I see this being the best result for that; it's not the quantity of users but the quality - which I don't mean derogatory against anyone but more in the veins of "Better have fewer people actually caring about the mod than a bunch of trolls just for the numbers". Again, there is no money involved so download numbers don't matter outside of rather niche leaderboard comparisons. And while I am certainly not pure myself in that regard, one shouldn't determine their hobby based on numbers. If you like making mods, make mods regardless of if 1 or 1,000,000 people download it.

I can see your point, but whether or not it's moot depends solely on whether the person downloading the mod can tell it was made with AI or not. If it is not disclosed, then how would they feasibly know?

[Rainbow Fresh]

15 hours ago, LadyWYT said:

It also very much feels like turning responsibility over to Anego Studios for protecting players from servers, or at least opening the door for such, rather than placing the responsibility of decision-making on the player.

I mean, going on a tangent here but once VS reaches Minecraft levels of popularity and people start making servers that are just money-sucking internet slot machines predatorily targeting children for their parents credit cards with useless shiny pixels, I would certainly expect Anegos to step in and protect the players from such servers by taking them down - and hopefully more actively than Mojang does/did. Considering how many videos this one C-Tier "I ruined this P2W server by duping" YouTuber has and how little actually changes afterwards.

That said,

16 hours ago, LadyWYT said:

[...] but a warning box you only have to deal with maybe once is better than multiplayer itself getting bogged down by a bunch of a different rules and checkpoints that caters specifically to players with incredibly strict preferences. Because quite honestly, I don't expect those players to do their own server research when the system will just do all the work for them.

It's not about warning boxes and guardrails bogging down Multiplayer in general. This would be a simple convenience option based on data already present, already technically available at this point in time (pretty sure) simply presented for the players affected anyway by their own choosing. You block a certain mod/author/tag for yourself in your ModDB profile, you join a server for the first time/after a modlist update, the game gives you a kind warning that the modlist this server uses includes stuff you chose to blacklist. Nothing more. Doesn't even need to be more specific than that.

Would this be a level of convenience that amplifies the laziness of certain such individuals resulting in them not doing their own background research about the server they try to play on? Most certainly - however, honest opinion; would they do it without this convenience in the first place? Probably not, then complain after the fact about how they wasted time and energy playing on a server just to find out the people running it are *insert whatever insults and slurs necessary* for using *whatever this person dislikes THAT badly*. If they get "sent away" by an automated warning based on their own preferences, atleast there is a reasonable chance these people (or rather some of them) never get far enough to become a nuisance reducing the workload of everyone involved a fair bit (less mental stress having to deal with idiots in the comments/DMs, less pending reports to verify) for a rather small one-time investment of coding effort. And as with most other things, slap an opt-in setting into the game itself to turn this warning on or off; who wants it that badly can get it who thinks it's a bother/hindrance doesn't.

10 hours ago, Teh Pizza Lady said:

but whether or not it's moot depends solely on whether the person downloading the mod can tell it was made with AI or not. If it is not disclosed, then how would they feasibly know?

Indeed. If it can't be identified as AI and isn't labeled as AI, nobody would know it's AI. Sadly we are long past the point where people look at what is and rather look at what could be. If a mod has one too many fancy filler words in a bullet point list, the description is probably AI. And if the description is AI, clearly the code is too. And if the code is AI, then it's slop. And if it's AI slop and you didn't disclose that, then wow does that make you an asshole. Probably all your other mods are also AI. There is no cure for collective social discrimination and idiocy but an honesty system could potentially help establish a sense of order. If things get tagged properly - and as mentioned before the incentive for people to do so is less bullshit by trolls and AI inquisitors (but not limited to AI as a topic) - everyone has it a bit easier. Less trolls in the comments because people genuinely just against AI content for themselves can filter out results with the appropriate tags, and the rest can be warned/timed out/banned for their shitty behavior. Less moderation work because people don't need to report every single mod they think could contain a trace of AI, but only reports of things reasonably believable to have been mis-tagged. (If it doesn't look like AI and isn't tagged as AI, why would you think and report it as AI?) People honestly tagging their stuff would get credibility so that if they say a specific mod isn't AI, it likely isn't. (And honestly most people's gripe is based on the mods that just spam post ChatGPT summaries and AI artwork thumbnails - hiding those is hard to impossible.)

But I don't want to make this topic solely about AI if only for fear of escalation and topic shutdown so please don't count how often those two letters appeared in the previous paragraph alone and remember that a general tagging (and or blacklisting) feature with the right options, used properly and supported by filters in the search works for everything. Want a mod that adds weapons (like Swordz) but doesn't add new tiers of metal (like Swordz)? Something that adds NPCs while filtering out the common settlement building sim mods. And that is without adding personal gripe against X topic/person.

[godlikeion]

Its 3 Am and I'm tired. but holy hell the same point has been stated multiple times. Please just take the time to understand what is actually being said rather than just saying "Well this." I just want the modding forum to be the best it can be. Tagging AI mods wont cause a problem. And allowing more types of tags for how AI was used and allowing separate blocking helps everyone without adding some insane workload or causing crazy problems. I don't want to be a dick but honestly. The point being made is so simple. So I'm going to sum it it as simply as I can.

Authors can tag mods and specify how AI was used in the tags.

The report system will stay because dishonest actors exist and there needs to be a system to address that.

Users can block certain tags but allow others. EX: Block AI art allow all AI code. Block vibe coding allow AI assisted code.

There's a toggle in game that when you join a server if you have tags blocked it warns you what mods are using those tags and whether you want to proceed.



And all I can gather from the counter argument is:



What if people don't use the tags for AI mods.

Bad actors can use the system for harassment.

It punishes honest actors.

Its the players job.



And every one of those arguments has been countered.



People wont. And they might get away with it. But most mod authors are honest. And they can still be reported.

They can get banned or timed out for harassment. Simple as.

No. Transparency is almost always apricated. Even if you disagree with something you don't bully the person who did what you disagree with for being honest, You appreciate that their honest and leave. And for most others it makes the person look trustworthy

That doesn't mean the system should be better, Ill say it. The modding page is bare bones. It has almost no functionality. But this is a step in the right direction because some players don't have the time (like me.) or patience to look. This system would streamline things and make them less frustrating.

 

 

Now im going to go to bed. But please, Make an actual argument. Im tired so i may have missed some points but they have probably been countered pretty dang well. Doing something is more effective than doing nothing. It will make everyone happier. Your arguments against it can be reasoned out after five minutes.