Welcome to Inkbunny...
Allowed ratings
To view member-only content, create an account. ( Hide )
Kadm

Inkbunny Staff Recruitment 2024

by
Edit: This journal is now locked and applications are closed. See a summary of the initiative so far at the bottom of this journal. Commenting is now locked.

In 2020 Inkbunny performed its first additions to the staff in nearly a decade, adding three community moderators to the staff through a process that ran from September 2020 to May of 2021 before final selections were made. The results of these additions have been overwhelmingly positive, driving down average ticket response time significantly and cutting ticket backlog to a fraction of its original high. We've been happy with the results, but mindful of the idea that we would need to do additional recruitment sooner than later.

The original process was performed on approximately twenty candidates that had over the years offered assistance to Inkbunny in some form or another. We solicited information on their skillsets and backgrounds and compared with their history on the site itself to narrow down to a much smaller group to move on at each stage, finally selecting three user to advance into shadowing and eventual placement as Community Moderators.

Today we're looking to solicit applications more broadly than the previous run. We know from feedback following the original run that there are a number of people out there interested in assisting with Inkbunny who had not expressed it before we closed out our process and ran through it. Additionally, as a result of policy changes over the last several years we have workload that necessitates more assistance than we accounted for when we made those changes.

Inkbunny is entirely volunteer run. There is no compensation involved for its operators. As a result, we make no explicit demands of your time as staff. We ask only that people give what time they're able, when they are able, and contribute to discussion when possible. There is a possibility that roles on the site or access to information may be restricted if someone is particularly inactive over a long period of time, but there is no penalty associated.

Likewise there is an expectation that staff members be mature and at least somewhat professional in their interaction. A majority of Inkbunny's staff is nearing or at least forty years old, and Inkbunny is a site designed by adults, for adults. Especially in staff to staff interaction which can at times become contentious when discussing policy, it is important to remain level headed and treat others with respect.

We do not have an explicit language requirement, but all internal communication is performed in English. English is a second language for several staff members. Proficiency in other languages is definitely a plus.

A large amount of our current workload centers around the moderation of AI generated works. A technical background and at least cursory knowledge of AI work will be helpful in moving forward. More in-depth expertise may be weighted in the process.

While we are soliciting applications broadly for this run of the process, we still expect to select a very limited number of candidates from this run. We expect that we will re-run this process in the future as the need arises, or as people feel the need to step back from their work.

If you're interested in applying to moderate Inkbunny, please send me (Kadm) a private message titled
"Inkbunny Staff Recruitment 2024" indicating as such and we will document it. This initial application does not need to contain significant information. We will leave the applications open for approximately two weeks (ending March 30th, 2024 at 8PM CST), after which we will close out the list of applicants and begin selecting people to move forward in the process. We will solicit more information from those selected in the second stage of the process. There will be limited or no responses to messages sent until we move to the second stage. All users that apply will be notified whether they are selected to move forward in the process or not.

Edit: Thank you everyone who took the time to participate, both in the discussion and by applying to help moderate Inkbunny. As of closing we have received ninety-one applications, exceeding the number we originally expected to receive (approximately eighty).

Now that applications have closed, we will take some time to evaluate the pool of applicants based on their activity on Inkbunny (including how often they use the site, what they use it for, and how they interact with others), along with previous contributions in terms of Support Tickets. Once we have completed this, we will contact all applicants to inform them whether or not they'll be progressing to phase 2 of the recruitment initiative. We expect to progress twenty applicants to phase two, and eventual selection following phase 3 will be no more than five applicants.

Please be understanding that while we intend to grow, we intend to do it slowly and in such a manner that we ensure there are no growing pains with bringing on new staff members. This process is designed to be thoughtful and thorough. If you are not selected in this process, please do not take it personally. There are a large number of motivated individuals interested in the position, and this will not be the last time that we solicit for new staff. We will re-run this process in the future and invite anyone not selected at this time to re-apply at that time.

If you have questions or comments feel free to reach out via Support Ticket.
Viewed: 4,398 times
Added: 1 month, 1 week ago
Site News Item: yes
Commenting Locked
 
IceAgeChippies
1 month, 1 week ago
Enemies of Chippy will surely apply! D:

*prays for a happy, healthy InkBunny*
Bloodhawk
1 month, 1 week ago
ALL HAIL CHIPPY!
IceAgeChippies
1 month, 1 week ago
<3
Mousington
1 month, 1 week ago
Who is Chippy?
IceAgeChippies
1 month, 1 week ago
I am the Chippy, and Chippy means ME! *thunder*
KevinSnowpaw
1 month, 1 week ago
Gandolf Reference? I respect that.
hanage1
1 month, 1 week ago
Protect Chippy! <3
Bloodhawk
1 month, 1 week ago
do you live under a rock or something? It's chippy!
Yiffox
4 weeks ago
chippy is watching you under rock
nelson88
1 month, 1 week ago
Chippy the King!XD
CosmicMutt
1 month, 1 week ago
Best of luck with this round kitty.
Kz
Kz
1 month, 1 week ago
all the best <3

--Kz
MurryCurie
1 month, 1 week ago
how would the CMs do Ai content moderation manually? theres like 100 Ai image posted here every hour (more or less :P)
Kadm
1 month, 1 week ago
Part of the problem is the volume of users that we're not able to keep up with right now. If we reach a point where we're able to respond in a timely manner to individual issues, I think a majority of the issues will clean up over time. We'll weed out the users who are unwilling to comply with the rules, and that will slow down the amount we're having to moderate overall.

Not all of the alleged 100 AI images per hour are violations, and we're only concerned about the ones that violate policy or are missing required tags.
Kavukamari
1 month, 1 week ago
what are the kinds of actions that need to be taken on AI works? just making sure they follow the guidelines?
Kadm
1 month, 1 week ago
Yes. Moderator actions require a  non-trivial amount of time for each instance. We document the problem both in the support ticket (usually) and on the users account, and if it's related to a submission specifically and it's not being deleted, on the submission itself. We prepare a notice for the user that hopefully explains what they did wrong, what they can do to avoid it in the future, and in some cases what they can do to fix it now. Then we need to actually do whatever moderating (edit, delete, lock), and we probably need to follow-up in some of those cases.

Most days the flow of tickets isn't too insane, but no one wants to spend hours of their day on that stuff. So the goal should be to have enough people that we can reliably handle the workload without burning anyone out. And currently we're not at that stage. I'm not sure that this run will solve that problem entirely, but we'll expand slowly and combine that with my work to improve the tools to make us more efficient, and eventually we'll reach a point where it feels good to run this site.
Kavukamari
1 month, 1 week ago
I feel like i have good potential to be a moderator, but at the moment my job and hobbies are taking all of my time, unfortunately, so i don't know if I would be able to dedicate the time without sacrificing something else that's important to me
Kadm
1 month, 1 week ago
And that's okay. This isn't for everyone. The hope is that eventually we have both better tools, and more people, and it's easier for all involved. But it definitely takes some time right now. We don't make demands, but we do expect people to participate, so if you don't have time let us know in advance!
AutoSnep
1 month, 1 week ago
" A large amount of our current workload centers around the moderation of AI generated works. A technical background and at least cursory knowledge of AI work will be helpful in moving forward. More in-depth expertise may be weighted in the process.

My concern is that there're essentially 2 groups of people proficient in the AI-generated art: AI directors and AI haters.

AI image generation is at the bleeding edge of machine learning, so only people passionate about it will devote a lot of time to learning hundreds of tricks while climbing the steep learning curve and be ready to relearn everything every three months. Naturally, this includes AI directors who are interested in generating content, because learning improves what they can produce. And slightly counterintuitively, AI haters also have to obtain the necessary skills, be it for applying anti-AI watermarks or catching users faking prompts on InkBunny.

Experience with generative models doesn't happen randomly, as I can't imagine someone dedicating so much time to learning tools which will become obsolete in a year.

Both aforementioned groups seem like a poor choice for moderators, and I'm not sure people proficient in generative models outside those two groups exist, not to mention are willing to dedicate time to moderation.
Kadm
1 month, 1 week ago
I don't see this as a concern. We're not unleashing people with unlimited permissions at a whim. We're going to gather information and then interview candidates, and by the time we finally install them as staff on the site we'll have significantly more tools for auditing what staff members are doing. The Community Moderator role is already limited compared to the other roles on the site, and whether someone is enthusiastically for or against AI is unimportant. What is important is that they enforce policy, and we're aiming for sufficient diversity to ensure that people can double-check everything.
AutoSnep
1 month, 1 week ago
The problem with people passionate about AI being on the team is that it'll cause lots of emotional reactions, be it hatred towards administration or living in constant fear. But okay, let's assume emotions don't exist.

The rules, as they're written, are so vague and hard to follow that it's totally possible to interpret them in a way that'll get 95% of AI directors banned. Some people use simple prompts, a few people build their workflows around InkBunny rules, but a lot of people put their first or last prompt into the description and call it a day — not because they're malicious but because rules are impossible to follow when using their workflows, existence of which the rules ignore.

Someone using Photoshop or Krita based tooling — what's the solution? Someone going through 10 image-2-image full conversions by relying on vague shapes (yes, I know someone who does exactly that) — what's the solution? In the eyes of AI haters, the solution would be to ban the user because the image isn't reproducible from the prompt.

As to AI directors on the team, the obvious practical issue is counteracting moderation becoming easier.
VarraTheVap
1 month, 1 week ago
That is true.. Gotta agree that the ruling does not suit hybrid workflows well and makers of those are the closest to a true artist you can be with an AI-based workflow. And think they are also the most successful users of that tech here on IB.
We are beyond the point where just throwing a prompt at a generator gains you significant attention.
AutoSnep
1 month, 1 week ago
I've seen complex hybrid workflows being used by both digital artists (with manual drawing experience) and by AI directors drawing stick figures.

" We are beyond the point where just throwing a prompt at a generator gains you significant attention.

Different AI directors are moving in the opposite directions.

Just throwing a prompt into a tool gets you better and better results as the models get better. On DA, some users seem to post 10 images per day by using dynamic prompt generators and just picking images which look the best. And trust me, it totally gets you watchers. 😆

And of course, as more complex tooling becomes available and AI directors dedicating a lot of time to the generation learn to draw at least a bit, the hybrid workflows become more popular too.
VarraTheVap
1 month, 1 week ago
Hmm, on DA I quickly ended up unwatching those who post multiple a day since that just floods the feed with similar works q_q And they just look too "ai-like" anyways. But taste surely differs as usual xD

What I'm thinking though is that as an art community, IB should rather encourage the hybrid workflow more than the "prompt'n filter"-approach and to a degree the ruling causes the opposite. At least if the rule were to be enforced really strictly.
Balmung
1 month, 1 week ago
I agree such workflows are a concern and would love to work out a solution to that. More complex workflows give better results, but are much harder to make reproducible.

I am not sure what the solution is there, it's tricky. But I think something should be possible to work out.
AutoSnep
1 month, 1 week ago
A1111 supports API logging. I assume ComfyUI supports it too. So someone using Photoshop or Krita can configure their tools to dumps full API logs.

It'll be more data, but it doesn't get us anywhere close to reproducibility because these logs won't contain input image data, as well as tons of other parameters. With the way drawing and generating interact, I don't think there's a sensible way to fully verify a workflow outside of maybe a video recording of the whole process. Demanding a video for every image upload is beyond reasonable though.
VarraTheVap
1 month, 1 week ago
This gets to the point where it's equivalent to requiring a traditional artist to write down information how to draw the image accurately a second time.

Maybe staff should become more clear regarding what exactly the rule wants to have for an effect?
What should it encourage or discourage?
Is the prevention of copying artist's styles via AI, the only goal?
In that case having to list all models and all keywords used may be a more sensible approach since the exact step-by-step process doesn't matter.
AutoSnep
1 month, 1 week ago
The problem is users saying, "I totally didn't use artist names in the prompt!" and having Chunie signature in every image. 😆 Without some sort of reproducibility, it's impossible to verify that the user isn't lying. And if it's impossible to verify, then the rule can't be enforced, thus the rule doesn't exist.

One may think that disallowing AI is one of the possible solutions to resolving the dilemma. However, in the long term, it's even worse:
(1) A year from now, it'll be impossible to tell whether an image was generated with AI
(2) Everybody will lie about not using AI
(3) You lose the last bit of openness, as nobody will be sharing workflows

DA chose the "fuck it, we're doomed anyway" path, FA chose the "it's future me's problem" path, and IB is trying to find a sensible compromise in between. Maybe solving 80% of the problem is the solution, who knows.
VarraTheVap
1 month, 1 week ago
Solid arguments, but nobody will try to reproduce anyways, would they?
With that premise, it becomes more like a rule just for the sake of a principle already...

But yeah, it's a really bad dilemma...
AutoSnep
1 month, 1 week ago
" Solid arguments, but nobody will try to reproduce anyways, would they?

Ask KammyKay, a notable example of an AI-hater who is getting better and better at generating AI art. 😂 And I believe there're more.
VarraTheVap
1 month, 1 week ago
I heard that name.. Interesting, lol xD
Xennos
1 month, 1 week ago
And that's exactly why I created a journal back then stating that AI should have its own space as an additional measure. Not because I want it out of my face as a non-AI artist, but because I want to give it an opportunity to be adopted without the fear of backlash from both sides of the purists. Like you said right now, IB has so far chosen an adequate path to handle this matter. But why stop at 80% and not make it better by trying to fix people's issues too, right?
AutoSnep
1 month, 1 week ago
What is your proposal on separating pre-AI and AI art? How do you want to see it implemented on InkBunny? I just don't see a simple solution. 😆
Xennos
1 month, 1 week ago
Nothing is simple, I know, but at least it won't give another reason for the two groups to step on each other's toes and create more hatred. Maybe in the future, when people are more understanding of the tools or the nature of AI and realize it isn't as bad as we think, maybe then we can roll back and merge the two. But that will never happen if there is still a reason to hate each others.
AutoSnep
1 month, 1 week ago
People will get over it either way. I don't think it'll take long. Hating AI will stop being mainstream the day Disney, Dreamworks and James Cameron release their AI-assisted films which are received positively by the general public.

And this will happen. Disney has bought half of AI startups, is creating digital copies of all actors like crazy and owns huge datacenters. People just don't realize the scale of what big corporations can do.

I bet the whole AI debate will probably be over in two years, four years max, unless something wild happens.
Xennos
1 month, 1 week ago
Weren't you just trolled by a cub hater the other day? (Welcome to the club, by the way. XD) Never underestimate the ability of a few loud individuals to bring everything down for everyone else. I'm just suggesting a way to reduce the percentage of those kinds of people to spawn in our site. That's all.
Balmung
1 month, 1 week ago
What about AI assisted art? Which category does a sketch processed with AI go in? And pictures with AI backgrounds?
Xennos
1 month, 1 week ago
If it's just assisted, then I suggest it should be in the regular gallery with just a tag as "AI-assisted." That way, it doesn't need the hurdle of adding a long list of prompts to incentivize regular artists to try and adopt it more.
Zagroseckt
1 month, 1 week ago
Considering how much time and work goes into a curated ai genoration wether or not it's started with a drawing is mut.
Trying to get the pose the two or more different subjects what there doing fixing glitches in the matric with inpainting and img to img after genorating like 100+ first runs refining and tweeking the prompt to almost get what you want. then another 50 to 100 ish regenorations in img to img and inpaintig and twiddeling reprompting and sub promting inversions loras stacked loras custom trains of public models to tweek out unwanted junk or add enough refferinces to get things working properly when prompted....

Silly thing.  Getting the models i use to give me a spotted skunk morph. or a spotted skunk period....
Has been a nightmare.. allways with the hognose single or duel striped verent not even going into tail poofeness getting mixed up with limbs.

Well the other day while goofing off i added "My Little Pony" spotted skunk colored to my primpt run thinking i'd make a funny responce to someone in the discord i was in..
first run it actuily gave me a brony with the right Friggen furpattern....
nice simple plop laugh....
Poke at prompt a bit. added morph.

Got render after render after render of Perfict spotted skunk morphs NON ponyized even with the my little pony prompt.
take out the pony part and poof junk output. put it in perfict

able to  minipulate the primpt for pose mixed morph groups were hit and miss alot of them seemed to enheret the pony part so i had to move things around but i got good results.

ive spent over 100 hrs coaxing this ai to give me this and suddenly pow.

now i'm not one for posting much i may drop an ai post in when i finaly get MY fursona to come out properly with all atachments intact (and im getting dam close now) but it's not easy and i wish folks would stop saying it's not real art or its easy to do.

or sure for a quick soda like what you can get away with a short prompt but it's not gona give you what you realy had in your head .. unless ya get lucky :p
Xennos
1 month, 1 week ago
And that's why I don't use AI, not because I dislike it. It's because it can't do precisely as I want it to or perform as well as I can.

Hell. I'll go down the rabbit hole too the moment it's good enough to take some of the workload off my shoulders.
weblurejoltik
1 month, 1 week ago
Typically, when it comes to AI, these are the bare minimum points that reasonable people care about:

1.) Only post the final product of a piece you've been working and had an actual goal in mind for. People generating random images and posting a raw output of something they thought looked cool will lead AI-generated images flooding out actual drawn artwork. If your website has a place for scraps, this kind of 'spam' should be contained to that area, or perhaps even a special area made specifically for unpolished AI images.

2.) AI artwork should be labeled, so that people may choose to blacklist it. This one's more difficult, since there are many different ways for AI to generate or assist in the creation of artwork. For example, it'd be unfair to force someone to put an "AI generated" label on their artwork over minor AI involvement, such as an AI-generated background (plenty of artists use royalty free photos with a filter slapped over it for their backgrounds) or simply using an AI-generated piece as a reference for an image they completely redrew from scratch. There are also artists that supply their own artwork to the AI as an img2img input, to 'enhance' what they drew. Then there are artist that take an AI-generate piece and spend hours touching it up by hand, repainting much of the image. AI is a very broad tool that can be used in many different ways, so we either need to come up with a broad yet concise set of labels for it, or only focus on the few use cases that the majority actually cares about.

3.) The AI-generate image should not be plagiarism. This is the most important, but also the most difficult point to enforce. Not to be confused with the blind argument of "all AI art is theft!!", most of these involve a bad actor intentionally going out of their way to target specific artists in order to reproduce their hard work.
There are three main ways to get an AI to plagiarise, or otherwise copy, another artist. The first is to directly feed an existing image that someone else drew into the AI; this is the most blatant and obvious form of AI plagiarism.
The second is to create a dataset focused around a particular artist, in order to greatly influence the AI into copying that artist's artstyle; this one can be harder to spot and even harder to form a legal argument against, as you cannot copyright an art style, and because the AI is still technically generating a unique piece of artwork without directly copying an existing piece (though depending on how small and overtrained the dataset is, this could be debatable).
The third is to create a broad dataset consisting of many artists, then to add artist tags to each image in the set, thus allowing users to type in an artist's name in order to nudge the AI into adhering to their style; this last one is the hardest to spot, and since the AI will still be pulling from a much larger and diverse dataset, it might not even be worthwhile to fight against.
Xennos
1 month, 1 week ago
That was a spot-on assessment of what should be a requirement to make AI content acceptable anywhere. But that also leads to another problem that is directly related to this journal. In order to enforce these requirements for all users here, I would say I'm not envious of the amount of work the mods here will soon have to undertake. XD
Kadm
1 month, 1 week ago
The selection and interview process exists to pre-empt any 'emotional' issues. I don't care if someone is passionately for or against, as long as they're willing to enforce the rules. If an issue arises, we'd deal with the issue, and if it persisted, we're perfectly capable of removing someone if they're not willing to be a productive part of the team. It's happened in the past.

You're already aware of the fact that we're working to clarify certain points of our rules, but overall I think that what you describe is a much smaller portion of uploads than content that simply violates basic rules. Certainly we'll have to address it at some point, but it's not as much a concern as all of the people ignoring the rest of our rules. I feel like people failing to tag works or using disallowed tools probably aren't utilizing complex workloads.
AutoSnep
1 month, 1 week ago
If the primary focus is obvious rule violations (artist names, no metadata, closed tools etc.), then I guess I'm totally okay with whoever wants to deal with that — it's mostly mechanical and alternative interpretations are unlikely. It's full reproducibility becoming the focus that worries me the most.
SomeAIArtist
1 month, 1 week ago
👍
bullubullu
1 month, 1 week ago
You make some good points in your posts. Passion can be an issue, but it certainly would be better to have people that are very engaged with the AI tech to be on board as moderators, both anti's and directors as a balancing factor.
Finding someone who is "neutral" on it while also be skilled at looking through the prompts is not likely, whereas the two "camps" have in-depth knowledge of the field.

Also, on the AI director side of things, I think that those who follow the rules and respect them for their purpose (basically allowing AI in the first place instead of outright banning the stuff) that these creators of those images have a vested interest in those rules being followed by other creators, including those who are new to the tech and unaware of the rules regarding it.
Cirn0
1 month, 1 week ago
" The rules, as they're written, are so vague and hard to follow that it's totally possible to interpret them in a way that'll get 95% of AI directors banned.


That's precisely why I save all the intermediate generations and all Photoshop files with all the layers intact in case I'll need to show them to the mods. I don't save inpaint masks though because it's way too tedious even for me, and they can be easily inferred from the pictures themselves
AutoSnep
1 month, 1 week ago
Same. I've written a script which dumps all metadata from all selected files. I also use a consistent naming scheme, so I can decipher a workflow from a filename like "...!!!ld-eh-2x-cmb-e-bsr-ldm-e-2x-resr-i-e-2xe-2xe-2x1!-yn-i2-e-zoom" and the metadata index I generate. 😆 With Regional Prompter, I chose to manually draw a mask using a palette, instead of drawing with mouse in UI, which is probably for good. 😁 Couldn't find a sensible way to save inpaint masks either.
Nightwing57
1 month, 1 week ago
Oh ok
VarraTheVap
1 month, 1 week ago
Certainly an interesting preposition...

Regarding AI, wonder, have you considered semi-automating the checking moderation does?
Am I right to assume that currently, moderators need to manually look at every submission tagged as AI to check whether prompts are provided (and don't contain artists etc.) and the used model/LoRa is allowed?
While it would be wishful to truly verify that the provided information is truthful (beyond plausibility), that is sadly not feasible no matter what.
However the check itself could be automated in my opinion (not talking about using AI algorithms for this here, just some string searches). That way moderation can focus again on what people report.
Kadm
1 month, 1 week ago
We don't check every submission on the site as it's submitted. Moderation happens in one of two ways.

1. As things are reported as violations by users in Support Tickets.

2. Things we see in our regular use of the site.

We don't have anywhere near the manpower to proactively check every submission as it comes in, and it's not even a goal.

I don't really see short-term submission changes around AI works to be viable. There are other things we can improve to make support more responsive and timely.
VarraTheVap
1 month, 1 week ago
Oh, I had a misconception then.
Thank you for clarification!
Zagroseckt
1 month, 1 week ago
Just be carfull not to ban someone over old art that sudenly is getting atintion.

i hate it when i see someone get band or restricted for stuff theyve had up for weeks or months and becos they had multiple of it two different moderators see X was disiplened X was disiplend so i'm gona ban and through up the paw to the face when the artest goes wtf thats been there for months..

Just shove the bad art into a scaps bin with a note to get it back on the main you need to do XXX or XYX
Kadm
1 month, 1 week ago
We're generally extremely lenient when it comes to banning. We would rather coach people to post acceptable works than remove their ability to do so. There are circumstances where we feel like we need to go that far, but they're not the norm.
IBp
IBp
1 month, 1 week ago
:3
thegolem
1 month, 1 week ago
:3 :)
BelovedBnnuyBaby
1 month, 1 week ago
It's baffling you haven't banned AI art yet or dedicated a whole new gallery to it.
Posting art and seeing it disappearing after 1 minute because people post "fun" characters with 17 fingers in one hand is really good for the site. /s
Kadm
1 month, 1 week ago
We mandate tags to allow people to block AI works if they don't want to see them. You're not taking advantage of that to avoid what you don't want to see. If you want help with how to do that, feel free to let any of us know. If there are things getting past a block, submit a Support Ticket with details and we'll deal with it as staff are available.
BelovedBnnuyBaby
1 month, 1 week ago
It doesn't fix the fact that people that don't use the blacklist are still gonna have trouble seeing real art getting AI art cock blocked...
Balmung
1 month, 1 week ago
I don't know about other people, but I don't believe I'm missing anything.

I regularly search the site for everything I'm interested in, so it doesn't need to be in Latest or Popular for me to see it.
PamperedDiaperFox
1 month, 1 week ago
Same here. I usually just glance at popular rarely even before the AI art. Searching and watching allows me to find my niches much easier.
CBear624
1 month, 1 week ago
Finally, someone gets it

And yet another reason why every art site must have something that only allows AI art included, and seperate that from the others so that it is not so overcrowded and not overshadow real art, making it difficult to gain traction
PamperedDiaperFox
1 month, 1 week ago
It might be best to work on getting rid of the whole entitled self-centered whiny attitude early.

This is a popular website. Your artwork would be quickly buried even if AI art wasn't allowed. What you can do is properly Title, tag, and describe your work so that people who are interested in it can find it and people who don't want to see it can avoid it. Other than that, there's really nothing else to be done.

If you want to actually get seen, you'll need to either wait for it to happen gradually or put yourself out there on other websites. If you're basing your whole career off of a single website and staying quiet, (Eggs in one basket and not being noticed.) AI art isn't the only issue you'll have later on.
AutoSnep
1 month, 1 week ago
I see several pretty good images in
CBear624
CBear624
's gallery actually. I'm surprised none of them reached the front page's Popular section. Busty women are popular in general, aren't they? 😁

I suspect it's cel-shading that affects the popularity. Thumbnails need to be recognizable to get people to click on them. Using AI for coloring could massively increase the perceived quality. I could even help with that. 😁 I think I'm currently at 5 digital artists having a collab with me (their line art or sketch, my AI coloring).
PamperedDiaperFox
1 month, 1 week ago
It might be best to work on getting rid of the whole entitled self-centered whiny attitude early.

This is a popular website. Your artwork would be quickly buried even if AI art wasn't allowed. What you can do is properly Title, tag, and describe your work so that people who are interested in it can find it and people who don't want to see it can avoid it. Other than that, there's really nothing else to be done.
CBear624
1 month, 1 week ago
Im fine with what i have now, its not about me that im referring to about getting traction

i refer to the artists that cares more about getting views and followers, believe me, i know a few artists that are obsessed with getting attention, and i dont envy them, that kinda attitude is dangerous and they need to sort things out about that


Then i'll put this way - its not so much is whats happening "now", its what to come thats the worrying part

We've seen how quickly AI art have showered the art world, every art site wants to exploit the hell out of it, it is insane how quickly people are jumping on it


I could care less about what the outcome of my art would bring, im only doing it for fun - the only concern i have with AI is how fast it has thrown itself into our lives, and i can only imagine what the next year would bring, or the next 5 years if it keeps going at this pace


Im fine with AI art as long as its used for fun or just to go crazy, but if its used for profit or to disciminate others, thats where i draw the line, because that is not the encouragment we should be giving to people who're not artists, to generate ai art for those purposes, it feels... wrong


Hell i've heard of some small cases where some artist that does art, has not only tried AI but has stopped making it themselves and just resorted to AI only - as an artist myself, that hurts to watch


So im sorry if i came off on the wrong foot but seeing AI art and the potentially dangerous things it can bring, is worrying and we need to figure something out that can help both sides
AutoSnep
1 month, 1 week ago
" Hell i've heard of some small cases where some artist that does art, has not only tried AI but has stopped making it themselves and just resorted to AI only - as an artist myself, that hurts to watch

I'm not a huge fan of this happening, but it depends on motivation, I guess. At the core, some people enjoy the process, some people enjoy the result, some people enjoy the attention, and some people enjoy the money. 😆 For someone suffering from art-block or burn-out, AI can totally become some sort of solution.

The best scenario, of course, would be to incorporate AI into the workflow to enhance a manually drawn image, rather than replacing drawing with generating completely. This is how you can have human touch and vision, plus top-tier quality, plus speed up of the process. 😎

In the end, if it's the choice between creating nothing and creating something, it's hard for me to justify complaining. I know an artist who hasn't been active for years, but when he got his hands on AI tools, he got interested in creating art again. Maybe I'll get him to learn the technology better, so that he could rework his unfinished sketches into art rather than generate everything from scratch, to give his AI art more personality. 😁

" i can only imagine what the next year would bring, or the next 5 years if it keeps going at this pace

The future is unknown. That's what makes it fun. 😁 And gives more reason to prepare for it. 😁
CBear624
1 month, 1 week ago
Eh, those are valid arguments

i dont have art block as often so im in no position to comment about what others are going through with that, but my point still stands, AI art is good when it helps artists with references on things they cant get right, motivation from art block, etc, but we shouldnt encourage people to use it for the sole purpose on making a career path with it in a field where creativity from the person themselves, made with their own two hands, is what truly shines

its just how its always been, and its how it always should be
AutoSnep
1 month, 1 week ago
" we shouldnt encourage people to use it for the sole purpose on making a career path with it in a field where creativity from the person themselves, made with their own two hands, is what truly shines

Nice in theory, almost never works in practice. 😆 Vast majority of artists give in to the corporate overlords and produce whatever the chain of managers above them choose.

Artists who create only what they want and somehow don't die in poverty are extremely rare exceptions. Telling someone to aim for the top 0.01% of all creators is setting up unrealistic expectations and will only result in suffering and depression. 😐 99.9% of freelance artists live in poverty or change their job (source: trust me bro 😆).

Artists who don't live in poverty, adapt and use the tools which allow them to produce art at competitive rates and competitive quality. This is what gets them hired and pays their bills. And in the nearest future, that means using AI tooling to at least some degree, judging by what Disney and big corporations are doing.

I too would probably love to live in a perfect society where art is pure and disconnected from the vanity and the greed, but sadly, we have to consider the reality. 😆
Rakuen
1 month, 1 week ago
What exactly is the problem? Do you think people don't know there is a possibility to block tags? If they know and don't block the tags, then presumably there is no trouble because they are seeing what they are happy to see. Just because it's not what you want them to see doesn't mean it's a problem.

Also, I haven't blocked "AI" art but it's not overwhelming everything. (That's not not the case on DeviantArt for example.) My watchlist is free of AI art and while it comes up on the popular tab, it's not the majority and it's usually fairly good pieces.
trider
1 month, 1 week ago
I'd say that's on them for not using blacklists. I'm not ok with AI images, but the blacklist does most of the work for me.
Kavukamari
1 month, 1 week ago
i don't understand this, is there a specific reason people don't use the blacklist? it's there to make your life easier
BelovedBnnuyBaby
1 month ago
Coming 1 week later to say I added AI art to the block list and I see them still because people refuse to put the AI tag :))
AutoSnep
1 month, 1 week ago
" Posting art and seeing it disappearing after 1 minute because people post "fun" characters with 17 fingers in one hand is really good for the site. /s

Don't MS Paint-level artists produce even more content? 😁 In general, I think that relying on the Recent Submissions section on the front page is totally non-viable, especially as the website grows. A measly 10% growth in daily submissions due to AI won't change that in the long term.

If you're interested in growing your popularity, the strategy is well-known:
(1) prefer NSFW over SFW
(2) draw popular characters
(3) tag your images well
(4) images must be enjoyable

This is what gets you watchers, no matter what content you produce. AI artists can certainly outpace digital and traditional artists in the short term, but in the long term, the upper bounds are largely similar, I think.
DanteAffinityXD
1 month, 1 week ago
Just a gentle reminder that the only thing being popular gets you is hate mail and pictures of strangers feet. :P If you draw for fun instead, you're more likely to enjoy the path to said strangers feet. Stop worrying about what other people do and just have fun with your creations :3.

I'm looking forward to our new moderators and I know the staff will do a great job making IB better each day. This place is awesome!
MagyarMilo
1 month, 1 week ago
(0) Character-viewer interaction - gets you comments

Example: If Tails asks if he should raise and spread his tails - you know the rest. Profit! But be more creative than that. :3
AutoSnep
1 month, 1 week ago
Comments don't get you watchers or views though... 🤔
MagyarMilo
1 month, 1 week ago
Raises chances to land in Popular

Inspires fun comments.
Those inspire fun conversations.

People are more likely to return to a fun place, than an empty one.

They may become watchers.

Some have lots of watchers, but no comments, and not even 10% views of the amout of watchers they have.

(Also, of course, exceptional case: drama is no fun. Empty views are just traffic, with zero interaction.)
AutoSnep
1 month, 1 week ago
No idea how you manage to get watchers through comments. You must be pro at it. I don't think I've ever succeeded at this. 😁

It's hard to measure how active the watchers are because we have very few numbers: watchers, views and comments. Artists who have been active on InkBunny for many years accumulate tons of no longer active users as watchers. It'd be unfair to calculate ratios when most of watchers are from 8 years ago and it's a second submission after a year of inactivity.

If, in addition to the simple number of watchers, we had stats on how active these watchers are in general (visited/commented/faved/+watched 1/10/50 times over day/week/month) — then we'd be able to draw conclusions. 🤓

I inspire conversations by writing "CREEPY COMMENTS WELCOME" under my every submission. Sometimes leads to amusing results. 😂

Surprisingly, "CRITICISM WELCOME" succeeds way less often. Somehow people are very hesitant to criticize something even when explicitly asked to do it. 😆
MagyarMilo
1 month, 1 week ago
I used to have a gallery.

But what I suggest is something else. People indeed like to fantasize, and a talking character is inviting.

People first see a thumbnail. Small text may arouse curiosity. The next thing they check, is the picture. Then they may fave and just move on. Unless there's some reason to stay and get involved. If they stay, then they may check out the title and description and drop a comment.

Now, whether they like you enough to want more from you, is another thing.
Some people may just post art and say nothing. Chances are, even if they get popular to some degree, they probably miss out on a lot of potential loyal fans or even friends.

But this is just an opinion. When I drew, I myself forgot adding a bit of interacting with the viewer inside the artwork. xD
But that's why every post needs a plan, so we don't forget, nor neglect steps. - Yours are pretty good points 1-4, btw!
Zagroseckt
1 month, 1 week ago
theres a blacklist for a reason.

Try it :)

or get rid of sonic art and stuff that isnt adult orented and rough sketches that are barly visible doodles that have barly any shape or form. oh ban non cub related art how about we just shut the site down becos what you just sead triggerd me.

bla bla bla and so on.

Were an edge site trying to ride that razor line of being legit and not dip into criminal lighten up :)
Zagroseckt
1 month, 1 week ago
not sure but i think this got ploped under the wrong main sub thread
eeveefan
1 month, 1 week ago
its hard to use a blacklist effectively when many many artists dont tag their arts properly. sure they have the community to add tags but often times it doesnt happen and it slips through. you can add the tags yourself, assuming the artist hasnt blocked that off, or you can send in a ticket to have a mod force the tag in not giving the artist a choice.

all in all its not really an option since alot of artists are lazy and dont tag correctly.
AutoSnep
1 month, 1 week ago
" get rid of sonic art and stuff that isnt adult orented and rough sketches that are barly visible doodles that have barly any shape or form

Man, I tried getting rid of all images of MS Paint tier quality by hiding submissions of individual users. When I gave up, my submission-blocklist grew to 300 users or something 🤣
MagyarMilo
1 month, 1 week ago
When it comes to maintaining a community, I would recommend a type of person with the ability to handle a given situation emotionally independently. The person should be mature and experienced. Most people fully develop physically and mentally (to become a true adult) around the age of 25; while their personality develops around the age of 35.

A potential moderator should have rational and logical thinking. The person should be able to focus on finding the solution to a given problem, without getting personal.

The situation should be taken seriously by both parties (moderator and community member), perhaps formal language should be used, respect should be practiced, and a certain etiquette should be worked out and required to be followed.

Personally, I would team up moderators: one that focuses on facts, and one that considers circumstances. (They don't have to be a Sherlock and Watson, of course. A more "cold" and a "warm" hearted person have different perspectives.)

Don't just consider everyday personality and activity, but also the person's behavior during work. The two can be completely different.

And yes, criminal background (even if not caught!), and psychological issues should be considered, when making a careful decision, regarding whom to hire.

I hope some of this can be inspiring.

I also would recommend that moderators "refresh" their training now and then. In a way like it's a good idea to re-read ToS, and Rules every now and then.
MagyarMilo
1 month, 1 week ago
Bonus: avoid people that are hungry for power, and influencers. So if somebody has a huge following, and uses it non-stop for so-called "brainwashing", then you will likely regret hiring a person like that.
MagyarMilo
1 month, 1 week ago
And a golden bonus: I think, moderator badges aren't necessary. Moderators should have a separate, anonymous moderator account.

All too often I see people becoming chummies with mods, then try to soften their hearts, and manipulate them. This way, that could be avoided. That, and other issues.
Beedy
1 month, 1 week ago
neat
SanisPifo
1 month, 1 week ago
Maybe others have already asked this, but I wanted to hear from someone who's been through it. Why isn't there a separate category for AI-generated art? I don't want AI art to be stifled; that feels like it's against freedom. But having them lumped together in the same section seems like an odd choice.

Maybe tagging AI artists with an official badge that requires them to always include three mandatory and automatic AI tags could encourage good behavior and reduce problems with stubborn AI artists who don't tag their stuff at least? Would that be possible? Or would it be a programming impossibility or just too irritating to set up?
Kadm
1 month, 1 week ago
Because any solution like that would require developer time, something that we're often sorely lacking. And the person that works on it would need to want to work on it. We don't pay for developer time (or any other time), so what gets done is at the discretion of the people working on it. I'm not saying more customization is impossible in the future, but I'm prioritizing (as I personally work on the site) features that improve our responsiveness and ability to deal with problems on the site, and small features that make sense in the areas I'm working in.
SanisPifo
1 month, 1 week ago
Fair enough, it makes sense. It's a shame, but I understand the reason for the lack of updates on the Inkbunny website as well. I guess there really isn't much to be done at the moment, just small, subtle steps forward.
KammyKay
1 month, 1 week ago
Is there any interest in expanding on the number of developers? If so, what skillsets would you need?
Kadm
1 month, 1 week ago
Inkbunny is primarily PHP and Postgres. Little bit of Javascript. General web design is helpful.

There are two big deterrents to bringing on random people. First is, we don't mandate that people work on specific things. They'd work on whatever they want. But that may not be things that we want pushed to the site, or it may be implemented in a way that we're not comfortable with. But it's hard to complain, because it's all volunteer.

The second thing is that the site is generally pretty easy to read, but is in no way designed for a lot of simultaneous contributors. Some of that is down to process on our end, but also because it's a lot of interworking methods that should be cleaned up and separated better.
KammyKay
1 month, 1 week ago
This sounds like something that could be fixed by setting up a basic git strategy. What is your CI/CD like? I've worked a ton on this type of stuff.
Kadm
1 month, 1 week ago
We have Git(hub). Inkbunny's got it's own (private) project. And we're pushing towards better change process, but for so many years, one or two people alternately were working on the sites, making changes and pushing them live without much review.

We've got test servers now, and I'm really trying to instill a culture of branches and pull requests into it, but there are some habits that are difficult to break, and fighting the urge to hotfix on live and then forgetting to commit it back to Git is a real problem.
KammyKay
1 month, 1 week ago
Oh okay interesting. I've wondered for the longest time where Inkbunny's source code lived. I'm more familiar with Gitlab, but I know Github has branch protection rules like requiring pull request reviews before merging, which can prevent devs from pushing directly to main/master. Not sure if you're already doing that.
Kadm
1 month, 1 week ago
Traditionally we haven't. But if I haven't been clear, I am not a web developer by trade (I am a storage software engineer), and so I've been insistent that Salmy and GreenReaper review my pull requests before merging them in. But fixed processes are a new thing that I'm pushing more and more, and I want to get to a point where we're branching and merging for regular releases in the future, committing hotfixes if they need to happen and can't be done on test servers in advance. But in general I think we're more aware than we have been in the past. A lot of my actual job is DevOps style work, and I'm trying to inject that into our running of the project.
AutoSnep
1 month, 1 week ago
I have a strong itch to move all the 'style="width: 960px"' and 'color: #9c9c9c' from view templates to CSS and then add proper support for the dark theme and the mobile layout, instead of relying on crazy hacks in UserCSS. But I'm not sure I can devote enough resources to get every single inline style fixed. 😆
Kadm
1 month, 1 week ago
I'll be honest, I don't know that much about the actual web design stuff. I can make features and put elements on the page, but I don't really do visual things. I design back-end features and improve existing features. If you've got feedback with specifics, submit it as a support ticket and I'll try it on a test server.

I'm also just not a huge mobile fan in general. I don't browse things on a tablet or phone. I don't like mobile websites, or apps. But I accept that a lot of people do.
AutoSnep
1 month, 1 week ago
I mean I'm considering asking to become a dev, like KammyKay above.

I've implemented a dark theme for InkBunny as a userstyle, but it has its limitations and I'm sure a lot of users who would prefer a dark theme don't even know what a userstyle is, not to mention are aware of my userstyle.

I've tried doing a mobile layout style and had some success, but only Firefox supports userstyles on mobile which limits potential userbase to about 2 people on the planet, which doesn't even include me. 😆 And with the way layouts are implemented on InkBunny (with lots of inline styling outside of CSS, as well as floats and other mess), these userstyles are full of hacks and often miss some obscure features (I still find some random miscolored buttons in my userstyle).

Ideally, something like this would be implemented within InkBunny's codebase itself. That way, users would be able to switch to a dark theme and have a proper mobile layout without jumping through the hoops, and these features will work properly on every single page, without any hacks and random glitches.

This can't really be explained within a Support Ticket. It's about moving 250 inline styles in view templates into CSS, fixing layouts to use modern layout system (flexbox or grid) instead of floats and fixed widths — and after that preparation work is done, styling features can be properly implemented:
1. mobile layout with fully flexible width
2. dark theme (selected in settings or based on OS settings)
3. vector graphics for high-density monitors
4. configurable story design (ideally something like FiMFiction)
5. etc.

I'm not sure what value the dark theme brings (I myself was very hesitant to move to the dark themes, but eventually got used to them), but I believe majority of people browse websites on mobile devices nowadays, and some don't even have a PC, so the mobile layout thing will likely be useful to a lot of users.
AutoSnep
1 month, 1 week ago
Currently AI art is 10% of daily submissions. I believe it just isn't seen as much of a problem, in comparison.

How often do you see untagged AI art? Established AI directors almost never forget tags, so in 99% of cases, it's newbies, in my experience. How exactly you deal with newbies (post a comment, send a private message, add a badge) is largely irrelevant as you have to deal with every one of them individually anyway. There's no programmatic way of dealing with it.

(Well, one possible solution is detecting AI metadata in submitted images, but it's a complex task and will take a long time to implement. And it'll help only with some of the submissions, not all of them.)

In the end, it's the same as with untagged cub/death/scat porn — if a tag is missing, add it and possibly comment. It's inevitable.
SanisPifo
1 month, 1 week ago
" (post a comment, send a private message, add a badge)


What if it were a mandatory option before making your first post? Something that pops up on your screen and prevents you from proceeding until you confirm that you're an AI artist, thus earning you the label?
" Currently AI art is 10% of daily submissions. I believe it just isn't seen as much of a problem, in comparison.

But:
" A large amount of our current workload centers around the moderation of AI generated works.

If the moderator mentioned that, then I'll assume there's still some issue surrounding AI artists, right?
AutoSnep
1 month, 1 week ago
" What if it were a mandatory option before making your first post? Something that pops up on your screen and prevents you from proceeding until you confirm that you're an AI artist, thus earning you the label?

You ignore the existence of users who upload both pre-AI and AI art. And if AI art wasn't as hated by some members of the community, probably 1/4 of AI directors would have posted images on their main accounts instead of creating alts specifically for AI art.

" If the moderator mentioned that, then I'll assume there's still some issue surrounding AI artists, right?

The solutions you propose (popups, badges etc.) don't solve the actual problems the moderators have (support tickets backlog, technical complexity of prompt verification, model availability etc.).
threeracoons
1 month, 1 week ago
I hope that even if you do find developer time for this you would very much consider the cons of segregating AI art. We're already treated like "suspect" citizens on this site, we don't need to live under apartheid too.

It takes me between 4 and 12 hours to "create/direct/steal" (choose whatever word that fits one's particular bias) each single AI piece I submit to this brilliant community (and I'm not talking about processing time I'm talking about human effort hours). Theoretically the spam/flood you mention would hurt me just the same, and yet I seem to be doing okay reaching the people interested in my stuff, either via tagging the correct kinks and showing up in search, or because my thumbnails in New and Popular look interesting.

Sending me down some sort of garbage chute to a place everyone agrees is "only for AI spam" would feel like an extraordinary betrayal.
Xennos
1 month, 1 week ago
I completely understand that pain and sincerely wish there were a better solution to it. But you must understand that not all AI pieces are art, and not all of them took as much effort to create as yours did. I understand there's a significant gap in quality between AI pieces that took a few minutes to generate and those that took hours and care. But in thumbnails, those spam-grade ones are, unfortunately, as eye-catching as any others due to AI's proficiency in lighting. While any traditional art that is remotely good enough to gain attention on the popular tab requires days or hours of work at least.
threeracoons
1 month, 1 week ago
Collective punishment is a breach of the Geneva convention. Is the annoyance of "Recent" being flooded by lazy AI spammers really worth committing war crimes? ( ͡° ͜ʖ ͡°)ノ⌐■-■

More seriously though, I too want creative and competent artists to be more visible, and not forgotten amongst a bunch of disappointing AI solos with good looking thumbnails. But this is not the way to do it: you would in one stroke cripple all the high effort AI directors. Why even try to work hard on something when we would feel so unwanted.

Besides, I have a hard time seeing this new reality that supposedly permeates IB "Recent" and "Popular" submissions. If I look, right now, at the last 60 submissions, I find less than 10 AI pieces. If I look at a random sample of 60 "Popular" submissions, I find maybe 6?

Xennos
1 month, 1 week ago
That came at the cost of heavy moderation and regulations. Isn't that exactly what makes anyone feel like a second-class citizen more than anything else? Having to be thoroughly inspected for every prompt used, every model picked, and every dataset trained? I was wishing for a way that doesn't need to be the case and that IB can be more user-friendly for new AI users.
AutoSnep
1 month, 1 week ago
I do feel like a second-class citizen, having to paranoically check every tag and metadata dump. 😐 But I think we just have to wait. The attitude towards AI among the general public will inevitably change, and with that lax restrictions will become the norm, eventually. It won't happen tomorrow, but we'll get there.
Xennos
1 month, 1 week ago
I wish I could be as optimistic as you, really. But seriously, I hope you are right for the sake of everyone on both sides.
Zagroseckt
1 month, 1 week ago
Same goes for non ai art. there are alot of shitty regular posts as well.
Xennos
1 month, 1 week ago
Those almost never got a chance in any spotlight. You can't pass off a stick figure as a masterpiece just because the thumbnail is small, you know, right?
TribalDragon
1 month, 1 week ago
There's people on here who actually posts one single upload per image they generate of a similar batch, that's a pretty big issue.

Even myself on my alt I thought a lot about this which is why I make a single post containing over 10 images and I only do that within the 4th week of the month before I return to my normal schedule in here.
Those who really spam one image a minute of their daily batch are generally users who are hungry for appearing in the popular section, like seriously a lot of those accounts I've noted on the popular tab have very little time to have joined inkbunny and are harvesting a lot of watchers, it's why I also agree to multiple comments where the AI art should be split to have their own section, but of course, developing that would take time.
Zagroseckt
1 month, 1 week ago
I would say and agree that limiting posts per day should be a thing. to save on server space at the least.
But not just AI users. everyone.
Zagroseckt
1 month, 1 week ago
why not a catagory for every kink subkink style set post level and.. oh wait thats what the tags are for.
sillylittlemaxwell
1 month, 1 week ago
hoping for a happier and healthier inkbunny!!!!

-mxwl, eclipse system host
TribalDragon
1 month, 1 week ago
Just hope you don't hire people that are registered sex offenders, I remember seeing a very infamous lolcow here requesting more than once to be a mod in here, don't tarnish the site reputation!
Flooftura
1 month, 1 week ago
Yup, a basic background check should be important.
Kadm
1 month, 1 week ago
We're members of this community too. We're well aware of people that are publicly unacceptable for some kind of leadership role.
MagyarMilo
1 month ago
A certain hawk that "didn't relies" comes to mind... xD
R3DRUNNER
1 month, 1 week ago
i wish i was more optimistic about this. unfortunately i think the site is less in need of more moderators and is actually in need of existing moderators paying more attention.

i had a longer convo with several established inkbunny artists on discord the other day and the general consensus has, over time, shifted to "we're all diversifying past inkbunny, because the site isn't paying enough attention."

the community wasn't asked if we wanted AI art or not, it was just a journal that got put up. it took far, FAR too long to ban all the MAP slop that gave this site such a bad reputation for literally a decade -- that was a rule change that could have been written in an hour.

sometimes i don't update my inkbunny for days because it doesn't have a mobile site and it's a pain in the ass to use on mobile. telegram, discord, twitter, bsky, e6, everything else has a mobile site. fucking FA has a mobile site. if FA can make a mobile site, anyone can.

i see these new artists and i tell them to get on inkbunny. half the time i can't convince them; their telegram channel 'works fine' or 'everyone finds them off discord.' before anyone says it, i am completely in agreement that telegram/discord/twitter ARE NOT good platforms for posting art. telegram is probably the best out of the three, but even then, it's not designed for it.

but does it matter? people adapt. when i made my telegram account i never thought it'd get all the features it did. i can run a YCH off the same app i message all of my friends on. hell, koko is selling commissions faster than she can draw them. she's never posted anywhere but telegram, as far as i know.

why use inkbunny if you're sick of hearing about AI/AI art models/false flagging AI art? i have two tickets on AI art being mistagged that i never got a response on. know where i don't have to worry about it? me and my artists friends' telegram channels, where it's just a rule. again, not saying it's perfect, or even that it's better, but it's definitely easier AND faster.

i still use inkbunny, but i get older every day. and, for the record, i don't hate the site. i like the site. i just hate watching it slog around in the mud and make slow decisions as the rest of the fandom, even shit sites like FA, seem to move quicker. i hope the site doesn't get confused and think that something functional means something people, especially younger, new NSFW artists, want to use.

so, i wish you all the best in finding community mods, and i'm sure they'll be a net positive to the site, and a lot of help. but having more people on the deck of the titanic wouldn't have stopped it from sinking faster, would it?
Kadm
1 month, 1 week ago
There's no magic bullet. This isn't going to suddenly fix everything overnight, but I'm committed to running this as many times as necessary until we can operate the site with a reasonable amount of effort for volunteers. I don't want to spend 2 hours a day of my life deleting spam, before we even get to tickets with actual content issues. We're solving a problem here that we should have addressed in 2017 or 2018, but the wherewithal wasn't there.

From a technical and feature standpoint, Inkbunny is still ahead of sites like Furaffinity in terms of features, even with the poor state of development over the last half a decade, excluding mobile. And while I'm committed to working on features, a full mobile rework of the site is definitely not in my wheelhouse. My hope is that as staff responsiveness improves, as the burden of running the site improves, we can put more time into improving the site as well. Because right now, it basically takes a ton of everyone's free time to barely stay afloat on daily issues, without making any forward progress.
MagyarMilo
1 month, 1 week ago
Try Opera Mini. Turn on Text Wrap. It makes it pretty mobile-comfortable.

You can doubletap a comment, and it fits the screen.

Might be inspiring.
R3DRUNNER
1 month, 1 week ago
" I don't want to spend 2 hours a day of my life deleting spam, before we even get to tickets with actual content issues.

for the record, my comment isn't meant to disparage the hard work you mods do. between bulk AI uploads, spam, harassment, etc... i don't envy any of you, and you definitely DO need more mods. not disagreeing there.

" From a technical and feature standpoint, Inkbunny is still ahead of sites like Furaffinity in terms of features, even with the poor state of development over the last half a decade, excluding mobile.

again, i agree, inkbunny is ahead of other fandom sites. furaffinity will probably never get a meaningful feature update, twitter (if you can count it) is actively trying to kill over dead. sofurry is facedown in a kiddie pool. there aren't a lot of other "good" traditional fandom art sites, and that's helped inkbunny out A LOT. but competing platforms always creep in and furaffinity isn't who you need to keep an eye on.

" My hope is that as staff responsiveness improves, as the burden of running the site improves, we can put more time into improving the site as well.

this is a good point and not how i was thinking about it. from what i've seen of the AI slop before i blacklisted the tag, there was a loooot of mistagging and stuff blatantly violating the TOS. handling that has taken thousands of hours of modpower and for that, i do appreciate all the work you guys do.

" Because right now, it basically takes a ton of everyone's free time to barely stay afloat on daily issues, without making any forward progress.

i've been on the site for a long time and i agree with this. it's been a combination of 'life support' and 'treading water.'

i truly do wish you guys all the best, and, like i said, overall i do still enjoy and plan on continuing to use the site. try to keep afloat for the time being.
AutoSnep
1 month, 1 week ago
" From a technical and feature standpoint, Inkbunny is still ahead of sites like Furaffinity in terms of features

Yes and no. There're several key features FA has that IB doesn't: proper search (negatives, operators etc.), mobile layout, folder hierarchy, stories being possible to read... 😆 InkBunny wins in everything else, but these are still important.
Kadm
1 month, 1 week ago
Aren't stories on FA just dumped in the description, if they work at all? I've literally never had someone say they prefer the writing experience on FA.

I've heard a lot of asks about our searches, and some of them probably aren't heavy lifts. I've already made a few minor tweaks on the backend to search queries, so it's possible I could put time into improving the search page in the future.
AutoSnep
1 month, 1 week ago
Stories on FA are separate blobs of text. Like, without a separate tiny frame on a page with tiny scrollbars where you can barely see two paragraphs. 😆

(To be fair, FA in particular does have issues with formatting (it expects UTF-8 TXT and produces garbage otherwise), this is why some people post stories in description. But it's because FA doesn't explain it well enough.)

Stories look like huge continuous text on virtually every website. FIMFiction, SoFurry, FurAffinity, ArchiveOfOurOwn etc. Sometimes the text is collapsible/expandable, sometimes width and font are configurable, but what InkBunny does nobody else does.

Let's be real: somebody reading a story with one hand doesn't want to click on page numbers with a mouse. 😁 Pressing space is the ultimate limit to what a one-handed reader is willing to do to scroll a page. And even that is painful, because with the tiny scroll area, pressing space is required more often.

And I'm not even mentioning all the issues with reading stories on mobile. All the zooming in and out, trying to get the frame aligned with the screen, misclicking once leading to the page scrolling to the footer... I understand you don't use IB on mobile phone, but trust me, the pain of reading stories on InkBunny is way above even the pain of trying to hit 10-pixel-wide button with a fat finger. 😆
VarraTheVap
1 month, 1 week ago
Huh, what's the problem with IB on mobile? I use it all the time like that ._.
TailBit
1 month, 1 week ago
- small icons and buttons everywhere, font size too small to read on a lot of stuff, which you would have to zoom in to read or click (page navigation, like buttons .. even typing in the textfield is too small and makes it real hard to move the marker)
- comments displays with 2 different font styles which I suspect depend if it is posted from pc or mobile, while pc keep the same format on both.
- journals only using 70% of screen width, image descriptions only 60%

It is just very clear that this page is not made for a mobile fit, and that is one thing I really hope can come a fix for in the future.

Heh, they could just ask for some web dev help like this xD
VarraTheVap
1 month, 1 week ago
Honestly if those artists wanna go somewhere else, so be it.
No site out there will ever satisfy everyone. How IB differs from FA and how both differ from DA.. and how DA differs from ArtStation etc. shows this very well.

Am glad that IB isn't trying to make changes just for the sake of it. Don't fix what ain't broken.

EDIT: Didn't realize I answered to this same post before :D
Peppercorn
1 month, 1 week ago
"A large amount of our current workload centers around the moderation of AI generated works."

How can you moderate AI generated works? What does that even mean? This sentence and the one after it feel entirely unrelated to the point of this message and only serves to bring further confusion to the AI topic as a whole. Its vague wording seems to imply that AI is not going to be moderated, but made even more represented than it already is. Considering it's been 2 years since the technology has evolved to the point it has and not a single (good) model has been created through ethical means clearly demonstrates the people that use it have no intention on using it for altruistic purposes. Art needs to be protected now more than ever, and in a community where creativity is one of the largest cornerstones of this fandoms, I'd expect the overseers of this site to agree.

Not that I'm expecting a pity party, but the very existence of AI art fills me with anxiety. My livelihood and fear of people losing sense of what art is supposed to be has weighed on me ever since the decision to welcome AI generation came to IB. This site needs moderation, yes. But I don't feel that it would fix the current problems that are already muddying it.
Kadm
1 month, 1 week ago
There are basic rule violations that need to be moderated, users that need to be warned, tickets that need to be handled. Each one of these interactions is a non-trivial amount of time for a person. A lot of it is not fun, and it is wearing on a person.

We need more hands to spread the work out to allow staff to focus on issues and improvements to the site moving forward. A lot of that burden is notifying and correcting AI users so that they comply with our guidelines. It's important to us that they comply, so other users can choose to block them and be relatively assured that it works.
CookieFoxBrandon
1 month, 1 week ago
I'm a bit confused on this, you guys are calling for more community moderators primarily because of the big influx of AI drivel right? Wouldn't banning or otherwise limiting AI in another fashion make it easier and require less community mods overall? e621 has a site instance specifically for AI stuff, wouldn't shunting all AI stuff to a site made for it in mind make it far easier for you guys to moderate?
AutoSnep
1 month, 1 week ago
It's akin to cutting off your arm to get rid of the pain. 😁

It's also likely to result in FA's situation of myriads of users lying about not using AI. And it'll get worse over time — it's already pretty hard to notice signs of AI usage in many cases.

e6AI doesn't allow cubs, by the way, plus the website is half-abandoned. With them still having e621's policies (two week wait after registration, submissions getting removed after 30 days if not manually approved), in addition to lack of interest from administration on improving and managing the website, it's pretty hard to recommend e6AI as the solution.
Kadm
1 month, 1 week ago
Shunting AI off to another section or site would require non-trivial developer time that simply isn't available. We don't have the resources to stand-up and run a second website like Inkbunny. So we're handling it in two ways for right now:

1. A project to help improve the clarity of AI rules, both for moderators (to make it easier to enforce) and for users (to make it easier to comply and understand warnings).

2. Additional staff members. Regardless of what happens to AI, more staff will be valuable to ensure timely responses. We're not going to recruit people that only know about AI. We're going to recruit people that are suitable to be moderators and some of them may have expertise in AI.
Mylen
1 month, 1 week ago
It fills me with anxiety too, not just because of the blatant theft and lack of ethics, but because the idea of large amounts of the population losing media literacy down the line is a very dangerous concept. People are getting stupider and stupider as time goes on, not the other way around. Within the microcosmic confines of Inkbunny though, I personally don't understand what the point of having AI is. Before it became a thing, site moderation and updates were already a longstanding issue. It is causing more problems, at least enough for it to be mentioned that moderating it needs more time and people. There was never a vote, or asking the people of this community, that is supposed to be for artists and creativity, what they thought the best way to handle this was. None of it makes any sense to me. Feelings aside, why have AI one this site if it's such a logistics issue?
Balmung
1 month, 1 week ago
You shouldn't worry too much. I experiment with it and still spend a bit too much on commissions. You've been on my list of people to get a picture from some day and that hasn't changed.

Doing AI well takes fiddling with the software, experimenting, buying hardware or renting it, redoing things many times until they come out right... it's far easier to just pay somebody and get a finished picture later in many cases.

I'm sure you'll be just fine.
Azerio
1 month, 1 week ago
*grabs popcorn and watches this decend into ai debate chaos*
Kadm
1 month, 1 week ago
Definitely not our most active news item of the last few years. Thankfully, since I get all the notices for this one.
Vexio
1 month, 1 week ago
I wonder if some of the ai moderation could be automated. I know a good chunk of ai art generators store tag data in the image so if you load up the image it fills in the tags again. So if you set new images uploaded to be checked for tag data you could have the ai generated tag automatically applied maybe? Obviously if the image is heavily edited in an art program after generation the tag data might not show up but it might help trim the tide of ai art under review by staff. I dunno just an idea.
Kadm
1 month, 1 week ago
I mention this a few other places, but stuff like that requires development and test time, and the way things are now it's just not viable. Hopefully in the future we can improve things, but this is the best path forward for sustainability.
Xennos
1 month, 1 week ago
That would be an ideal solution. But like you said, so far it's probably best to make do with what we can. So, good luck with the recruitment.
DangerDoberman
1 month, 1 week ago
All AI generated nonsense shouldn't be allowed.
Sk8erBoiTy
1 month, 1 week ago
Well hope ya find some good people for it.
ManaAraxis
1 month, 1 week ago
Does this mean you will be cracking down on AI "art". It's a plagiristic blight.
Kadm
1 month, 1 week ago
We're not changing what we enforce, but this will help with timely enforcement of our current rules, both regarding AI and other rules.
ManaAraxis
1 month, 1 week ago
It's just been proven AI art relies on pre-existing art to base the AI programs to generate art. I might as well post other artists art pieces and call it my own then. Rules contradict each other.
VarraTheVap
1 month, 1 week ago
That is not how that works. One cannot even prove that a specific artpiece is in the dataset.
The image gens rely on existing art the same way human artists rely on it: By having learned from it.
IceAgeChippies
1 month, 1 week ago
Everything I know about AI can be written in 12-point type on a postage stamp, so in light of that, I just can't see how AI can draw anything on its own.

Like, suppose I asked the AI to draw 'Two Care Bears Hugging'. That's an intentionally tricky request because the anatomy of Care Bears will not allow them to hug if the subjects are drawn 'on-model' and not drawn to the far left-right of each other (respectively), with arms/'necks' stretched in such waythat they couldn't if kept to model.

An artist can draw this---it's been done, but (again) the artist has to rely on some cheats to do it, to create the illusion.

If the AI is to draw 'Two Care Bears Hugging' from scratch, it would need to understand both Care Bear anatomy, how the human eye sees things, as well how to trick the eye into believing what it's seeing. AI isn't human---it has its own non-human  'eye' to 'see' with as well its own non-human 'mind' through which to 'think'.
How can the AI parse 'good' from 'bad' attempts at drawing 'two care bears hugging' (if, indeed, it's perusing human uploads from the internet), without understanding or seeing the scenario from/with a human's perspective?
VarraTheVap
1 month, 1 week ago
Am not quite sure what your response has to do with what I wrote.. But here
"two care bears hugging seen from afar" with "close up" as a negative prompt done with the "anthro" model of DA's Dream-Up:
https://www.dropbox.com/scl/fi/quikfhtkky1dcb0i1z47a/tw...
https://www.dropbox.com/scl/fi/ikz1wmw79wgtg1g7p8w8o/tw...

The negative prompt was needed because otherwise it makes closeups like here but that's a quirk of the DA model: https://www.dropbox.com/scl/fi/wuk8i2opbk0q5cgger32u/tw...
IceAgeChippies
1 month, 1 week ago
And it failed on all three attempts :3
AutoSnep
1 month, 1 week ago
" If the AI is to draw 'Two Care Bears Hugging' from scratch, it would need to understand both Care Bear anatomy, how the human eye sees things, as well how to trick the eye into believing what it's seeing. AI isn't human---it has its own non-human  'eye' to 'see' with as well its own non-human 'mind' through which to 'think'.

The problem with your silly test is that whatever complex prompt you can come up with which will show the limitations of AI of today, will become obsolete in 6 months, and you'll need to invent something even more complex.

I don't want to deal with Care Bear LoRAs and whatnot, so here's a simple test on DALL-E 3, which is currently top tier publicly accessible base model:
https://pixeldrain.com/l/RB4g1GM4

I don't understand exactly what restrictions on anatomy you're talking about, but I'd totally expect a human artist drawing hugging short-armed bear plushies like this.

I could give you hints on what current models actually fail at, and you'll have your "proof of AI incompetence" be valid for another 6 months, but what is the point? We, the AI directors/enthusiasts/developers, do know the limitations of the technology. We do know that it currently fails to match human's understanding of art.

What you're missing is that it gets better. AI art started with this:
https://upload.wikimedia.org/wikipedia/commons/d/d0/%22...
Two years ago, it looked like this:
https://media.craiyon.com/2023-06-24/d51655c2a8204aa68d...
And what it looks like now, you can see if you search for AI on InkBunny.

Do you see the issue with your logic? 😁
IceAgeChippies
1 month, 1 week ago
Maybe, but that's a hypothetical. I'm talking about here and now.

I have no way of knowing whether the AI made those images from scratch or not ...which was the point I was trying to make: AI farms existing art and rearranges it somehow to make a new image, but cannot CREATE art. Like, AI cannot (so far as I know) be given only a rotation sheet of a character and then be told to draw the subject in any pose, without farming the internet for subjects drawn in similar poses (and the AI would have no opinion of its own results).

So, no, I don't (but I can see you read comments/text about as well as Google Translate does). :3

Ending correspondence here. I'll read your reply (if you have one), but I'll likely not respond to it.
AutoSnep
1 month, 1 week ago
The problem with your belief is that it's unfalsifiable. Nothing can prove to you that AI "understands" something. AI fails — "haha, AI dumb"; AI succeeds — "AI just copied it, haha, it's dumb." An unfalsifiable belief is unscientific by definition.

There's a much better test proposed by Alan Turing. Instead of focusing on unfalsifiable statements and beliefs, you can just test whether a human can tell a computer and a human apart. This is scientific because it can actually be tested.

The test as it was originally proposed (text-only interaction) has essentially already been passed for naive testers. But it's easy to expand the test area to adversarial professional testers, to different modalities etc.

At some point you will no longer be able to come up with a test which AI will fail more often than human. Visual art is quickly getting there. In the realm of LLMs (large language models), the lack of good tests is already a huge problem.

" Like, AI cannot (so far as I know) be given only a rotation sheet of a character and then be told to draw the subject in any pose

It totally can. It'll fill in the voids with the most relevant information it has, just like humans. Some researchers are even working on recreating full 3D models from incomplete photo sets.
DanteAffinityXD
1 month, 1 week ago
That's not really quite true, and besides, our GPUs have the power of about 10 cockroach brains XD. What they are isn't how humans draw art, so much as how humans hear noise inside of a blowing fan at night. They're denoisers that hallucinate images out of noise, but the way they remove that noise is via prompt (control net, other images, ect), which is a bit of extra information. With sufficient training images (millions) the models start to experience overlap in the training data. It has a ton (tens of thousands) of images tagged "skyscraper", "boy", "cat", "house", "car", "angry"... and so on, so that during training, "memorizing" any image is no longer a good strategy. Instead, given that the same prompt could be any number of images, including those it might not have seen yet, it must figure out how to remove noise in a way to matches those prompts and hallucinate something with enough "cat"ness that it might fit into the set of all the "cat" pictures" it has seen or might ever see. It gets to do this over many different tokens and when we create a prompt, we're giving it a new combination that is probably not in any other image in the data set. That said, it tries to denoise the image in a way that best matches these because of the flexibility of the neural net. This results in really crazy fun stuff, because that denoising process can often hallucinate our tokens in surprising ways. Try to have to render a "murder of crows" or a "school of fish" and you might get more than you bargained for because the other tokens have been trained to make an image that is "murder" like and "crow" like, or "school" like and "fish" like. Meanwhile, humans have a very specific syntax and can convey complex structures via language, so we're less likely to take that description in such an odd direction. Given that both LLMs and Image Generators are both actively advancing fast, however, I suspect future iterations might be better adept at understanding human language.
VarraTheVap
1 month, 1 week ago
If I google GPU and cockroaches I get instructions on how to clean out bug infested PCs XD Do you have the source for that comparison? o:

Yes you are about right. However the things is, do we really know how we humans do bring things to existence? Or that it is that different? I would honestly call my fantasy also "noisy" and when I'm thinking of a specific object (classic "imagine an apple" challenge), that noise manifests into the apple.

Words having totally different meanings depending on context is indeed a weakpoint - and actually a reason why language models like ChatGPT is sometimes used to prepare a prompt for an AI art generator. ChatGPT definitely can take context and even slang into account.
The future will indeed bring a lot...
DanteAffinityXD
1 month, 1 week ago
It was some back of the envelop calculations I did the other day, let's check my math! You might be right, I may be way off here, but given the variety of sources, you could likely get way different answers. I am presuming equivalence of flops, like I might imagine Kurzweil's stuff goes by. A lot of numbers are tossed around for the human brain, but commonly I've seen things like 10^14~10^17 Flops. Some people seem to go way way higher and I suppose if you presume it's a quantum computer you might scoff at such low estimates. In any event, I then make another assumption, which is that the the Flops scale linearly with the number of neurons (but maybe I should use synapses?), so if you half the number of neurons, you half the flops too - and by some math, you can the estimate how many neurons a given GPU is if you know how many flops it can do.

The 4090 (the consumer grade flagship of NVIDIA's space heater) can supposedly do about 82.58Terraflops.
Kilo~10^3
Mega~10^6
Giga~10^9
Tera~10^12

So assuming 8.6*10^10 neurons in the human brain, then we have... math time with big numbers!
82.56*10^12 flops / (10^17 flops / 8.6*10^10 neurons) = 71,001,600 neurons

That's closer to a mouse, or 70 cockroaches. Looks like I was off, thanks for making me check, I love watching these numbers over time because it's kind of exciting. And that's presuming we're really smart, if we opt for the lower end and go with 10^15, a 4090 is about 7E9, or going into primate territory. I'm not confident to say that my GPU is in primate territory yet, that seems a bit too far for me. But then, maybe we just don't have the right software, evolution has had a lot longer to fine tune all these things with a bit of firmware from the get-go after all XD and most of those neurons are probably heavily related to doing stuff like telling my heart to beat, ect. ect. Though, given the interconnectedness of all this, that makes you wonder how much stuff like know how to breath impacts knowing how to draw. Really crazy fun stuff!

Thanks for correcting me! And feel free to correct me some more :3.
VarraTheVap
1 month, 1 week ago
Fascinating o:
Am surprised that one can really make such a relation between biological neurons and what's necessary digitally.
Interestingly enough, there is some research out there in the direction back towards analog computers and especially AI tasks may be a frontier of that (shown in the last chapter):
https://www.youtube.com/watch?v=GVsUOuSjvcg

After all currently what most AI services do (for processing as well as training) is it takes an analog input, digitalizes it (e.g. via a camera), then processes it digitally through a system that is heavily inspired by something biological (the brain which is analog) and when necessary, converts that back to analog for us to see/hear.
So it does sound logical to try getting rid of the digital steps.
AutoSnep
1 month, 1 week ago
" What they are isn't how humans draw art, so much as how humans hear noise inside of a blowing fan at night. They're denoisers that hallucinate images out of noise, but the way they remove that noise is via prompt (control net, other images, ect), which is a bit of extra information.

You're right that diffusion models don't reflect how artists work, however the models are a hyper-specialized and optimized part of human's way of imagining details in the art. When you look at a sketch, your brain reconstructs the whole image from the existing parts; you can squint your eyes and imagine what it'd look like when finished — and when you're squinting, your brain does rely on external knowledge like what concepts you want to put into the image. 😉

Imagination is hallucination. Hallucination is imagination. 😎 Your brain implements a pretty similar hallucination procedure to let you imagine things out of thin air.

The architecture can be extended. For example, you can train models specializing on perfecting compositions before the denoising model takes over (hasn't it been done before? 😆). And we already have models which replicate various specialized thoughts: how a face looks, how shading on a reference image is done, what the pose is etc.

As we go forward, we'll replicate more and more pieces of artists' thoughts. It won't match biological brains because of vast differences in the hardware, but the models will surely get close and closer to human artists over time. Before computers completely take over humans in every single aspect. 😁

" Try to have to render a "murder of crows" or a "school of fish" and you might get more than you bargained for

The language layer of current image models is much much dumber than the current top-tier LLMs can handle. It's essentially a compromise. It's possible to run image models on your mobile phone, but you need $10,000 worth of hardware to run top tier language models. They're that far apart.

And even those top-tier LLMs are still a small part of human's thinking — they're merely the intuition part, the talking without thinking part.
Sloss
1 month, 1 week ago
Would you want a coder for anything?
Though if InkBunny's backend is in Swish, I'll pass-
Kadm
1 month, 1 week ago
That's not what we're looking for right now, but it could definitely be in the future. It depends on what you're interested in doing, but I'd take a note (as I already have) of people that are interested.
weblurejoltik
1 month, 1 week ago
What is Inkbunny built on, anyway?
Kadm
1 month, 1 week ago
Inkbunny is primarily in PHP, utilizing Postgres. Little bits of other things sprinkled in.
eeveefan
1 month, 1 week ago
yeesh i come on to read this and i see alot of people chattering over the ai stuff.
kadm i gotta hand it to you, you really know how to start a conversation lol.

soon as i read that i instantly thought "how many comments below are going to be about ai?" and i was right.
i mean it is a big topic but common.. -w-
RobbyBunny
1 month, 1 week ago
Wishing you all the best with future moderation. I hope you're able to find the additional support members you're looking for and that some good candidates rise quickly to help take the strain off of you all.

I'd love to do my part and help keep things running smoothly, I love the site and community it hosts afterall, but my bumpy history as a user I'm sure precludes me from making a successful application. So good luck!
KevinSnowpaw
1 month, 1 week ago
on one hand I love inkbunny, Ive a deep respect for rules and laws and I want to contribute more to the site...



On the OTHER hand...


Drama is toxic! and the stuff the mod team needs help with is mostly telling people bad news XD

but such is the life of a moderator.


I'll consider tossing my name out there I..really do love you guys.
CuriousFerret
1 month, 1 week ago
If you toss your name in the hat, I toss mine in as well.

Be an excuse for my to purchase a proper laptop.
KevinSnowpaw
1 month, 1 week ago
Im thinking about it not that my personal choice to offer matters in the end it's entirely up to the vetting process but... in all honesty...It is a LOT of work, and a lot of extra stress and toxic problems XD I love IB ive got some moderation experience, it bears consideration, but it's not a choice I wish to make lightly.

though honestly the majority of the role sounds like it will be auditing AI art. XD
CuriousFerret
1 month, 1 week ago
Makes sure tags get used, responding to tech complaints.

What A.I. I've seen so far makes a good effort to tag and explain it with their terms they use for generation.

But I might have missed other problematic posts out there.

And I'll have more free time coming this summer as my work requirements are easing up.
KevinSnowpaw
1 month, 1 week ago
if you go for it just remember, you and I both, have to remember to check our strong feelings at the door. Ink bunny mods kinda gotta take more shit then FA mods and remain impartial.


that being said Im sure if you go for it and your approved that your assistance will be welcomed, the site can allways use more help im sure,
CuriousFerret
1 month, 1 week ago
Enforcing standards clearly posted is easy enough.

It not attempting to change minds over TOS, but implementing it.

Exchanges should be short and to the point.
KevinSnowpaw
1 month, 1 week ago
Agreed ink bunny staff has had a history of being impartial and patient even in situations they might be tempted to bring the hammer down. Its something I respect.
CareBear
1 month, 1 week ago
The nepotism is real.
VarraTheVap
1 month, 1 week ago
Huh? If there were nepotism involved here, they'd not make this public call xD
Kadm
1 month, 1 week ago
Do you have some suggestion for how the selection of moderators should work? Let me assure you that in my case, people that I'm friendly with does not play a large factor in my opinions on it. There are plenty of people I'm friendly with who would simply make operating this site more difficult if they were given an ounce of responsibility. The goal should be as smooth and integral an operation as possible.
VarraTheVap
1 month, 1 week ago
Actually good point.
Over several years I asked friends to be moderators of my discord server.. and sadly often ended with troubles and about half not being friends anymore q_q' Surely not only their fault, but I eventually learned the lesson...
It just doesn't work like in the movies xD Good that you are already steps ahead!
darkblane257
1 month, 1 week ago
*He clap to Chippy*
thecooler
1 month, 1 week ago
All I really want is for AI to be regulated its all.
pointyball69
1 month, 1 week ago
dang and I'm fairly new to the site.
MaDrow
1 month, 1 week ago
It seems IB has started something ||=(:3c

After ~4 hours of this announcement, E621 has also put up an ad for staff recruitment.
weblurejoltik
1 month, 1 week ago
No doubt a false front to distract people while NotMeNotYou and his goons inject more friends to grow their clique 🤗
Or maybe they'll just add more sockpuppet accounts, like they did with Cinder
MistahToonCatUwU
1 month, 1 week ago
IM DOING MY PART
Autorun
1 month, 1 week ago
being a moderator is a big responsibility. and not everyone can become one.
LoneWolf
1 month, 1 week ago
You don't have to be a engineer to understand that a car doesn't belong on the sidewalk.

Get rid of the AI content.
Minagatari
1 month, 1 week ago
Okay
Spinel
1 month, 1 week ago
Suena interesante la propuesta, pero sin algun tipo de remuneracion economica por pequeña que sea seria algo dificil, y dedicarle tiempo obligatorio a moderar teniendo el criterio adecuado para hacerlo bien, nuestros trabajos regulares a veces nos dejan cansados o sin animos de seguir haciendo otra cosa, cada vez que entro a la pagina me lo paso bien, entiendo el ingles aunque a veces se me queda corto en algunas maneras de expresarse (sinonimos, antonimos, regionalismos, etc.), por eso prefiero el español.

Lo mejor es seguir mi camino natural, etiquetar de vez en cuando y hacer comentarios tontos o ilarantes como lo tradusca mejor el traductor, para alegrarme el dia y de paso a uno que otro cristiano mas (persona), no todos entramos aqui para estarle jalando el cuello al ganzo, bueno si, para que digo que no pero fue una etapa hace años, me quede por que hay arte bueno y bien riko aunque ya no me la jale tanto vendo contenido furry y algunos otros si se quedaron ciegos de tanta paja y ya no pueden moderar e_e
CottonCaramel
1 month, 1 week ago
i would have a lot of time on me... but sadly i have to earnmoney... so i have to stick to comissions instead being a moderator, srry
Natepikaboy
1 month, 1 week ago
Destroy the AI.
AI isn't the perfect employee for anything honestly... it isn't a proper human that is responsible for work or whatnot
BubbleCat
1 month, 1 week ago
Hello I have a question. Are community manager discussions held on places like Discord and/or Telegram? Just curious since I might be interested, but I don't use telegram in case that' was a requirement.
BubbleCat
1 month, 1 week ago
Whoops Sorry, community "moderators" is what I meant to say.
Kadm
1 month, 1 week ago
Inkbunny has it's own self-hosted discussion area we use for management and operation of the site. It's not dependent on a third-party service and is not platform specific.
Andorason
1 month ago
I really dont understand the fear of AI Art...
It will not "steal" your Jobs, It will not Rule the World.
If you ask me, it maybe would be better if there would be a "Single Image a day"- Rule and in Case of AI ART Just Images which are looking good instead of "wierd ugly fingers and limps where they dont belong."
The Fear against AI is the same Fear Humankind had since Industrialisation.
SomeAIArtist
1 month ago
Took me a while to post here because I wanted to ruminate on this subject first, but I think I know what I want to say, and hopefully some mod is still watching this so maybe they can answer:

Why are AI artists subject to extensive extra rules and scrutiny?

Is there a rational answer to that, that doesn't involve the many easily debunkable myths surrounding AI art?
Kadm
1 month ago
Because those were the guidelines that we were comfortable with in order to allow AI at all. We could have simply said "No AI submission allowed", but we did not. We discussed and made rules that alleviated the majority of our concerns. There was never a world in which we were going to allow AI to be unrestricted completely.

We don't need any more rationale than that. It's our website to operate, and people post because we allow them to post so long as they comply with our rules.
SomeAIArtist
1 month ago
Thank you for replying, and thank you for allowing AI art at all. I always thought InkBunny was the superior furry art site, in performance and features, so I'm glad you allow us to post our stuff here.  ❤️

I understand it's your site, and you make the rules. No complaints there. I don't believe myself entitled to anything here.

I was just curious as to the thought process that led to the stringent rules. I can't help but feel that the extra rules add to the burden on InkBunny's staff, and to the divide between art types (AI vs. everything else). Personally, I would love to see that divide lessened to no more than a technicality: Some people do AI art, some people do art with pencil and paper, some with Photoshop, and none is better or worse than the other, and here they are just different choices for artistic expression.  😊

...but hey, some folks still say digital art is not real art, so it might be a while before AI art stops being seen as the boogieman...  😉
Kadm
1 month ago
It's been two years since we wrote the rules, and I've got COVID right now, so forgive if this is not the most coherent post that I've ever made.

There isn't one singular answer as to why we wrote what we did. We discussed the various areas that eventually became the rules, and what's published is the result of that. There was no single driving though behind them. For instance, the rules against utilizing artists names in prompts are intended to be relatively in line with our existing guidelines on derivative works and tracing. We don't want you to be able to closely replicate specific artists, and so we crafted the rule in the manner best suited to discourage that.

Tagging rules are actually just generally in line with the existing tagging rules, and if anything I wish we simply enforced all of our tagging guidelines more, but that's something that initiatives like the one in this journal are designed to fix, along with software improvements.

Things like disclosing prompts and other telemetry involved in the process while disallowing commercialization are intended to ensure that AI works remain firmly in the hobbyist or enthusiast realm. This isn't really rooted in any of our other rules, but looking back I definitely see it as a positive because we're still not sure how the legal landscape is going to turn out for some of these technologies. The restrictions in place should make it easy enough should there need to be changes in the future, and preventing active monetization avoids the issue of us worrying that we're depriving people their livelihoods if we choose to make a change.

Little bit rambly, but as I said, I'm quite ill at the moment. Hopefully that offers some perspective, or at least makes sense to you. I don't know that I agree with the idea that lesser restrictions would somehow lessen the animosity towards some AI users. We want people to be able to avoid AI works if they want to. That at least is extremely important to us. We don't want AI works to simply blend in with other works.
SomeAIArtist
1 month ago
Oh no, please get some rest! COVID is no fun! I've had it once last year and it sucked! ...still haven't got back my sense of smell or taste.  😟

Thank you so much for explaining to me. And even more for doing it while sick! It does make a lot of sense, yes. I didn't outright disagree with the rules before (just felt there may be a few too many), but now hearing your reasoning I actually agree with them!

So thank you all for making what seems to me like some pretty level-headed decisions concerning AI art on the site.
(I told you, InkBunny is the superior site  😉 )

Get well soon!   ❤️
AutoSnep
1 month ago
Step one on the path to getting out of the "second-class citizen" situation is getting the general public to accept AI art. I think, the "social norm" of hating AI art comes mainly from reactionary artists (mainly freelance) and the "twitter" mob that supports them blindly. As long as that remains the case, we're stuck in this weird spot. 😐

Let's not forget that popularity on FA and IB is defined by how much porn you draw. 😁 This already moves consumers towards accepting AI — at least in private, where they don't need to adhere to the perceived social norms.

What we need is artists. And the entertainment industry. The industry will probably come first, with a lot of pushback, but it'll win eventually. As long as we have enough artists on our side, there won't be any avenue left to fight against AI in any way. And that's when the second-class citizen status will go away. 😎

I personally work towards getting the public acceptance of AI art by doing collaborations with digital artists. 😉
artiedragon
1 month ago
Since a lot of this adresses AI, I wanted to know what InkBunny's stance on more complex AI works is. Specifically, I'm working on a Unity game which will use a lot of AI images, and possibly music. I wanted to showcase my work here, and have thus far since the game is still early in development, but considering the quantity of prompts which will appear in even a single page, it seems unrealistic to disclose the bulk of it. Likewise that it includes AI music (from suno.ai, which does not allow using artists in prompts). Does that mean that I can't showcase my work, or is there a way to figure things out together? I'm more than willing to work with you guys towards something that works for everyone.

Hope you get better soon, I'm sure dealing with Covid along with all of this is a pain as it is.
Kadm
1 month ago
We allow the upload of complex workflows as a JSON file on the same submission as the work it covers. I have a hard time imagining a workflow so big that it's too big for that.

AI generated audio should follow all of the existing guidelines for other AI works. For instance, suno.ai doesn't seem like it's model is open or available. This probably precludes you sharing audio generated by it.
artiedragon
1 month ago
I understand, thank you. The problem is more that a game video is made of dozens of images, so asking to continuously list the entirety of all of that is unrealistic, the game itself will likely have hundreds if not thousands of images between characters and backgrounds, which would be a monumental task to keep track of every single prompt. Especially since I share videos and it's a card game so every card, every background and character, effects if I use AI, etc.

But I understand your position, thank you for your swift response.
New Comment:
Move reply box to top
Log in or create an account to comment.