Sleazoid44 said: as Elmer Fudd might say. Only a few of us are posting, a big drop in reactions (replies or likes). Has the novelty wore off?
I had also noticed that. I think for sure the novelty has worn off, but I like to keep going, pushing what Bing will permit, and finding ways to get past the filtered pictures.
It may also be because so many are now able to just go to Bing and create their own images. When this group first started, I was amazed and came here to see what people were creating, and now I still do, but because we can create our own, we don't need this group as much as we perhaps once did.
That's my theory, anyway. I hope it doesn't die right out, but I think there are some who still enjoy seeing other people's AI creations here.
I've picked up some tips here and had some luck generating some pie pics yesterday. I seemed to have little to no control over coverage amount, but I still got some decent ones.
Today, no luck getting coverage at all. I did get a bunch of good "anticipation" pics, though.
Apologies/thank you to whoever I stole the printed t-shirt idea. Think of it as a tribute.
There is too much repetitive posting. There are only about three people posting 90% of the threads here, sometimes several times a day. It's all very similar and people have already seen it.
The one person who likes and comments on everything hasn't logged on in a few days. With him about, if you posted 20 images you got 20 likes. With him not about, you don't get so much engagement.
For many, the novelty has just worn off. Or they are just uploading images to their profile.
- there is very little actual prompt sharing happening compared to other AI boards, so there's little to no collaboration / emulation etc;
- technology has been stagnant for around a year now, so there's less to explore and try compared to a year ago;
- content that everyone makes is kinda tailored to their own preference (like for me the blue catsuit or wrestling outfit), so people will stick to one ideal scenario and use it all the time, but it becomes less relevant to other users over time.
I think things will pick up once again when short videos become better and/or the model becomes better at analyzing complex prompts.
There is too much repetitive posting. There are only about three people posting 90% of the threads here, sometimes several times a day. It's all very similar and people have already seen it.
I am guilty of posting a lot here, but never thought of it as being a bad thing. I can cut down but that leaves less content. As far as it being 'very similar', I'm running a food show with new episodes using different foods. In my mind, that provides some variety. Also each episode includes clothed wam, in-the-face wam, nude wam and group scenarios, some with sexual overtones and most with just fun being promoted in sexy clothes. I feel this provides something for most tastes.
I'm not about to post photo sets of things I have no interest in, but I think I'm providing some variety.
Do you have any suggestions regarding my content in particular? I'm open to change and have even put polls up to get user input.
There is too much repetitive posting. There are only about three people posting 90% of the threads here, sometimes several times a day. It's all very similar and people have already seen it.
I am guilty of posting a lot here, but never thought of it as being a bad thing. I can cut down but that leaves less content. As far as it being 'very similar', I'm running a food show with new episodes using different foods. In my mind, that provides some variety. Also each episode includes clothed wam, in-the-face wam, nude wam and group scenarios, some with sexual overtones and most with just fun being promoted in sexy clothes. I feel this provides something for most tastes.
I'm not about to post photo sets of things I have no interest in, but I think I'm providing some variety.
Do you have any suggestions regarding my content in particular? I'm open to change and have even put polls up to get user input.
In general, what people are looking for is something that they haven't seen before. You start new threads very often, and your images often have a similar tone or theme which has by now become very familiar (it's probably going to be half a dozen women in bikinis in a giant inflatable pool or on a beach, or a business woman in a dreamlike food factory).
I'm not saying the images are all bad (and it would be hypocritical to say obsessing over the same theme is bad), but what you post is a lot of the same sort of thing very often, and people aren't going to have the sustained interest to go to every new thread you post, or follow every new part of every story.
I really don't want this reply to sound harsh or that I'm telling you to stop posting, please believe I'm trying to be constructive when I say this. Perhaps if you posted less frequently and more selectively, you would get more interest when you did post. Instead of a ten thread story with lots of similar images illustrating each part, why not try one thread with the single best image illustrating each part?
More prompt sharing and collaboration needed, for sure. Frustration is a huge barrier to adoption here. I struggled and got nothing for a while, then I found this post by thereald: https://umd.net/groups/i/the-splat-the-brat-guide-to-synthetic-sl Immediately, I started getting pleasing results consistently. My enthusiasm skyrocketed, and I'm eager to learn more.
Making it easier for someone to get started will bring in more people, which means more ideas and variation. Increased prompt sharing and collaboration between more people will drive innovation and creativity.
Maybe a pinned beginners guide or tips and tricks post would help. Anything is better than digging through 1000 posts looking for a couple of lines of relative text.
I still really enjoy all the posts here, but find that the site has either slowed down quite a bit lately or the images have gotten larger, so I click on fewer and hit like on even fewer - but it's only because I get impatient waiting for the images to load
Hi all, I have lots of images produced over the last 6 months or so, its been a personal project self expression and partly mental health, there has been trial and error but I have lots of good images of a particular style and certainly some themes that have not been done. I'd ideally like to earn a little bit for my art so I'm currently in the process of setting up a store here on UMD.
Some interesting thoughts, ones that stand out for me:
Bobographer said: ...but I like to keep going, pushing what Bing will permit, and finding ways to get past the filtered pictures...
This is the main thing for me I enjoy making the images and exploring new technology that my pc can handle anyway, so I'll keep going as long as Im enjoying it and sharing images I think others will like.
thing said: I hope not, I'm just getting interested!
This is awesome there will be new people joining all the time
thereald said:the novelty has just worn off.
I think this might be a thing with Bing because it doesnt change, if I get bored with outputing the same thing I try and make a better model or LoRA with different effects.
kortanklein said: - there is very little actual prompt sharing happening compared to other AI boards, so there's little to no collaboration / emulation etc;
- technology has been stagnant for around a year now, so there's less to explore and try compared to a year ago;
- content that everyone makes is kinda tailored to their own preference (like for me the blue catsuit or wrestling outfit), so people will stick to one ideal scenario and use it all the time, but it becomes less relevant to other users over time.
Alot of this resonates, prompt sharing could be better, Id be able to make use of the bing generations for training lol Id happily share my prompts but without the associated models or LoRA they aren't going to be very useful.
Technology is always improving, look at some of the stuff being made with Flux (dropped in past few days) and Kling, (Flux is a brand new model and performs better than Bing in alot of ways unfortunately it can't even someone wet yet...), I dont think this is an issue except its use for niche areas.
Preference is always going to be a thing but variety is the spice of life thats a good thing, post your stuff I'm sure there's audience
In regards posting alot I'll post what I'm proud of, I used to post in seperate threads but they just got lost in the landslide of Bing when that arrived so I just post in the giant AI thread now, and try and keep my gallery updated. The reduction is probably just a sign that the Bing boom is subsiding, it was much quieter before Bing and hopefully it will remain busier than it used to. Just post what your proud of, if it needs its own thread for interaction put it in a thread, if they're stand alone images put them in the mega thread.
thereald said: [ (it's probably going to be half a dozen women in bikinis in a giant inflatable pool or on a beach, or a business woman in a dreamlike food factory).
I don't know when I last did bikini images, and the food factory ones were finished long ago. I'm on a theme with a cooking show now.
IPerhaps if you posted less frequently and more selectively, you would get more interest when you did post. Instead of a ten thread story with lots of similar images illustrating each part, why not try one thread with the single best image illustrating each part?
I will do that. Fewer photos of each part. As for fewer parts to a thread, there is no way to caption photos, so when the story advances, I create a new post so I can add text. I don't know of any other way around this.
I admit that some of my more recent posts have been HUGE with the number of images. I'll cut down and just pick the best ones when I do the next one.
As for prompt sharing, my own saved prompts stop working after a short while, and so I tend not to keep them. (other than if I'm working on a series)
The food TV show will run for a while, but I'll try to change things up to keep it more interesting. Without feedback such as yours, it's impossible to know what people like or dislike. So thanks for taking the time, and also for being sensitive about it. Much appreciated!
kemistry said: ... might be a thing with Bing because it doesnt change,
I'm seeing some "evolution" on Bing: I get fewer backward pies than before, for example.
OTOH prompts that were reliable stop working or I'll get 4 good pictures in a row then the Dog for the rest of the night.
I am guilty of posting "too many" of soaking wet women in formal dresses and gloves because that is my main interest but also because I get blocked by Bing with anything new. Halter-top dresses are OK now (weren't always) but I can't get halter-top blouses. I go back to what works and try to push Bing but often its like a dog I had as a boy, it sets its feet and refuses to move.
To be honest, once everyone started running their images through Krea, I kinda lost interest in posting. Krea is amazing, but it's an extra step that takes a lot more time to upload, enhance, and download the image before I can sort out my best or favorite images. A frustrating thing about Krea is that it will take a normal-looking five-fingered hand and add or remove fingers, or it creates limbs or appendages out of random things, turning an image I really liked and thought turned out really well on Bing/Nightcafe/Seaart into a more realistically textured nightmare.
I got frustrated with Krea's distortions, so I mostly stopped using it. And since it seemed that the vocal people wanted the images posted to look more realistic (as in less-plastic-y, Krea-enhanced faces), I stopped posting altogether. I'm a perfectionist when I share things, and "perfection" meant trying to keep up with the trends. It does seem like the Krea trend has kinda died down though, so maybe I'll start posting my un-enhanced images again.
I use the same type of outfits (ballgowns/eveningwear, long gloves, tiaras, etc) in a lot of my images, and that has been spoken out against quite regularly. Not my outfit preferences specifically, but the repetitive images in general.
I stopped reacting as much because I felt I was spamming people with likes, and also because it was kinda overwhelming to open a thread and find 50+ images that all look very similar to each other. I've found myself only clicking on a few thumbnails, then hitting Back to leave the thread. The threads I like most tend to get ignored and pushed down the page by the more active posters, so a lot of times, I just scan the forum for something that looks interesting, then lose interest and go do something else.
I love AI and use it for image generation and text roleplay pretty much everyday. That also means I have probably close to 10,000 images now that I've generated, and a couple hundred conversations on ChatGPT, Character.ai, and Seaart. I want to share images, but sorting out my favorites is a bit daunting. And I want to edit some of my conversations into stories, but I have so many to choose from that I can't pick one to start.
So... I guess the reasons I don't post or react much anymore are people's criticisms (both general and aimed at specific people) kinda scaring me off from posting, the extra work to enhance the images or edit the conversations, and just the overall feeling that what I've made just isn't all that impressive or interesting to anyone but me. There are only a few people posting or commenting on formalwear images, so that makes me think they just aren't that popular.
Also, I'll hop on the bandwagon here: the lack of prompt-sharing and the trend toward keeping good prompts secret out of fear they'll be found by the developers and blocked or whatever is probably a big reason for the decreased activity. The collaboration kinda stopped and everyone's just posting their own things. There's not a lot to talk about when its just images and story threads.
I hope this doesn't sound too negative. I know I get discouraged easily and tend to keep things to myself, especially things I'm proud of, but I really do like this community.
Kabe22 said: To be honest, once everyone started running their images through Krea, I kinda lost interest in posting. Krea is amazing, but it's an extra step that takes a lot more time to upload, enhance, and download the image before I can sort out my best or favorite images. A frustrating thing about Krea is that it will take a normal-looking five-fingered hand and add or remove fingers, or it creates limbs or appendages out of random things, turning an image I really liked and thought turned out really well on Bing/Nightcafe/Seaart into a more realistically textured nightmare.
I got frustrated with Krea's distortions, so I mostly stopped using it. And since it seemed that the vocal people wanted the images posted to look more realistic (as in less-plastic-y, Krea-enhanced faces), I stopped posting altogether. I'm a perfectionist when I share things, and "perfection" meant trying to keep up with the trends. It does seem like the Krea trend has kinda died down though, so maybe I'll start posting my un-enhanced images again.
I use the same type of outfits (ballgowns/eveningwear, long gloves, tiaras, etc) in a lot of my images, and that has been spoken out against quite regularly. Not my outfit preferences specifically, but the repetitive images in general.
I stopped reacting as much because I felt I was spamming people with likes, and also because it was kinda overwhelming to open a thread and find 50+ images that all look very similar to each other. I've found myself only clicking on a few thumbnails, then hitting Back to leave the thread. The threads I like most tend to get ignored and pushed down the page by the more active posters, so a lot of times, I just scan the forum for something that looks interesting, then lose interest and go do something else.
I love AI and use it for image generation and text roleplay pretty much everyday. That also means I have probably close to 10,000 images now that I've generated, and a couple hundred conversations on ChatGPT, Character.ai, and Seaart. I want to share images, but sorting out my favorites is a bit daunting. And I want to edit some of my conversations into stories, but I have so many to choose from that I can't pick one to start.
So... I guess the reasons I don't post or react much anymore are people's criticisms (both general and aimed at specific people) kinda scaring me off from posting, the extra work to enhance the images or edit the conversations, and just the overall feeling that what I've made just isn't all that impressive or interesting to anyone but me. There are only a few people posting or commenting on formalwear images, so that makes me think they just aren't that popular.
Also, I'll hop on the bandwagon here: the lack of prompt-sharing and the trend toward keeping good prompts secret out of fear they'll be found by the developers and blocked or whatever is probably a big reason for the decreased activity. The collaboration kinda stopped and everyone's just posting their own things. There's not a lot to talk about when its just images and story threads.
I hope this doesn't sound too negative. I know I get discouraged easily and tend to keep things to myself, especially things I'm proud of, but I really do like this community.
Bing doesn't excel at realism but it does excel at composition, you can get bing to get multiple subjects interacting realistically with each other and come up with really original scenes, and if people would share the prompt to go with them they can actually be a really valuable resource.
That is to say keep posting even if its not 'realistic' because the scene itself is just as important.
I don't make anything AI but I was frequently commenting and liking here. Have mostly stopped. And it's because the content creators here stopped posting sexual images (as in, see-through clothes, nudity, etc.)
I don't know if the public AIs got stricter but 3-6 months ago there were transparent clothes, tits, nipples, even some pussy. Now it's just gunge on clothes with the same plastic expression on the face.
I know that some SD AI models can do it (the nudity is amazing on some other non-wam forums) so not sure why y'all have stopped here.
Eg Bobographer you had an amazing set of pics with chocolate-soaked ruined white tight tops with no bras here, all those girls with pokies in the chocolate
Later sets seem to just be plastic food on opaque clothes.
I know we all have our individual kinks but I feel like "ruined clothes" and "the human body as seen through ruined clothes" is a pretty common part of this fetish. One of the hottest videos I've ever bought is Jayce and Ariel's "Wardrobe Raid 3" where they gunged and ripped four or five outfits in a row, including a long yellow formal gown that goes see-through and clingy from getting soaked in sloppy food before they tear it to shreds. It was an expensive video to make and almost nobody does that kind of content anymore since Ariel left the biz. Could AI do it?
I feel like there'd be a lot more interest here if people used SD models that didn't have any restrictions, vs chasing the rare sexual images that make it through the increasingly-sophisticated Bing content filters.
mlt said: I don't make anything AI but I was frequently commenting and liking here. Have mostly stopped. And it's because the content creators here stopped posting sexual images (as in, see-through clothes, nudity, etc.)
I don't know if the public AIs got stricter but 3-6 months ago there were transparent clothes, tits, nipples, even some pussy. Now it's just gunge on clothes with the same plastic expression on the face.
I know that some SD AI models can do it (the nudity is amazing on some other non-wam forums) so not sure why y'all have stopped here.
Eg Bobographer you had an amazing set of pics with chocolate-soaked ruined white tight tops with no bras here, all those girls with pokies in the chocolate
Later sets seem to just be plastic food on opaque clothes.
I know we all have our individual kinks but I feel like "ruined clothes" and "the human body as seen through ruined clothes" is a pretty common part of this fetish. One of the hottest videos I've ever bought is Jayce and Ariel's "Wardrobe Raid 3" where they gunged and ripped four or five outfits in a row, including a long yellow formal gown that goes see-through and clingy from getting soaked in sloppy food before they tear it to shreds. It was an expensive video to make and almost nobody does that kind of content anymore since Ariel left the biz. Could AI do it?
I feel like there'd be a lot more interest here if people used SD models that didn't have any restrictions, vs chasing the rare sexual images that make it through the increasingly-sophisticated Bing content filters.
Personally, I lose interest in WAM once the clothes come off or are unrecognizable. It's the same with faces. To me, WAM is all about the clothes and their wearer's enjoyment of the situation. If neither are visible, then I move on to something else.
Stable Diffusion is a much bigger investment of time and research than Bing or other popular generators. I used it for a while when SD2 (I forget the version name) first came out, but after a while I ended up uninstalling it because, while it didn't have many restrictions, it also wasn't that good at prompt adherance without a bunch of addons. It was significantly more complicated to use than the other generators I was using at the time. Quite a lot of the images I was getting with SD were many-limbed nightmares with grotesque faces. I'm sure it's a lot better now, but so are all these other image generators that don't require a lot of knowledge and experience to use to their fullest capabilities.
I'm relatively tech-literate, but I also don't really want to put that much work into training Loras and learning how to use SD just so I can have a little more freedom with the prompts. As it is, my tastes are so tame that I can get what I want from Bing without the dog blocking me too much. All Bing requires is creative wording, and I'm good at that.
mlt said: Eg Bobographer you had an amazing set of pics with chocolate-soaked ruined white tight tops with no bras here, all those girls with pokies in the chocolate
Later sets seem to just be plastic food on opaque clothes.
I create all of my images with Bing. Sometimes, I get lucky with Krea enhancing the boobs, even the southern regions, but usually not. My new "Mudslide - The Movie" series has more boobs showing than other recent sets. This is also my preference, but trying to make it happen with AI is frustratingly difficult.
I've been on vacation for the past few weeks, that said I've got little interest trying to coax Bing through it's safety features. Prior to going on vacation I posted the this thread https://umd.net/groups/summary-of-the-state-of-ai-generator-sit but it seems there has been little interest in exploring other dalle3 image creators whilst I was away.
My end goal was always to try and generate sufficient image sets to train a larger SD model, with SD3 being so disappointing, I thought I may have to give up on that but FLUX is looking real promising. I need to spend some time catching up with the last 3 weeks's development but my focus is going to be on unresrtricted dalle3 and flux training.
Kabe22 said: Personally, I lose interest in WAM once the clothes come off or are unrecognizable. It's the same with faces. To me, WAM is all about the clothes and their wearer's enjoyment of the situation. If neither are visible, then I move on to something else.
I'm with Kabe22 on this.
But my main point is that I was a lurker here for more than ten years before I started dabbling with AI. I have very particular tastes that it seems that most people here don't share. I get much more response to my images on instagram than I do here, so I'm not motivated to post much.
messg said: I've been on vacation for the past few weeks, that said I've got little interest trying to coax Bing through it's safety features. Prior to going on vacation I posted the this thread https://umd.net/groups/summary-of-the-state-of-ai-generator-sit but it seems there has been little interest in exploring other dalle3 image creators whilst I was away.
My end goal was always to try and generate sufficient image sets to train a larger SD model, with SD3 being so disappointing, I thought I may have to give up on that but FLUX is looking real promising. I need to spend some time catching up with the last 3 weeks's development but my focus is going to be on unresrtricted dalle3 and flux training.
Your post was insightful, however one of the things you didn't cover was how to start using OpenAI services with Azure, and stuff like how to change filter settings. I've noticed you're posting results with High filter settings and this is getting me curious. Are you doing this through Azure? Are you using your own script to use the API, using someone else's script, or are you finding other sites that use the API? The scripts I've seen are fairly trivial, and just use your Bing cookie to interface with Bing with no extra functionality added. It looks like you're doing something else. Like, how did you get approval for High filter settings?
skirtpie said: Does Instagram expose your fetish to the world?
I've considered posting on Instagram but I'm worried about it suggesting the account to everyone in my phone contacts, Facebook friends, former work colleagues, family members, etc.
Although if ever there was a time to use "#splat the #brat", it's now
messg said: I've been on vacation for the past few weeks, that said I've got little interest trying to coax Bing through it's safety features. Prior to going on vacation I posted the this thread https://umd.net/groups/summary-of-the-state-of-ai-generator-sit but it seems there has been little interest in exploring other dalle3 image creators whilst I was away.
My end goal was always to try and generate sufficient image sets to train a larger SD model, with SD3 being so disappointing, I thought I may have to give up on that but FLUX is looking real promising. I need to spend some time catching up with the last 3 weeks's development but my focus is going to be on unresrtricted dalle3 and flux training.
Your post was insightful, however one of the things you didn't cover was how to start using OpenAI services with Azure, and stuff like how to change filter settings. I've noticed you're posting results with High filter settings and this is getting me curious. Are you doing this through Azure? Are you using your own script to use the API, using someone else's script, or are you finding other sites that use the API? The scripts I've seen are fairly trivial, and just use your Bing cookie to interface with Bing with no extra functionality added. It looks like you're doing something else. Like, how did you get approval for High filter settings?
That is a wide sweeping question...
There are scripts for Bing and have been since day 1, I wrote my own back when the filters were much more relaxed as it allowed bypassing the prompt length etc as well as auto saving results etc. It still works of course but the filters makes it annoyingly useless. Obviously with scripting, you can be creative with timers, loops, multiple sessions etc.
AS an end user, Dalle3 though can be accessed directly either via OAI or Azure. For azure, you need to have an azure account set up and request access to the OAI services. You need a registered business for this. Once you have access, you can tweak the filters to an extent although you still can't get full unrestricted. You can either then use their playground to generate manually or write your own scripts to run through the API. It should also be said that you pay per generation so it isn't cheap to produce large volumes.
Other sites may use OAI or Azure APIs for their image generation. Sites like Coze, CiCi and others I mentioned do this however you have no control over the filters. If you are lucky, you may find a site that has little or no filtering applied however don't expect this to last long as it generally an oversight during their development and will be fixed. Cici lasted about 6 weeks before being turned up.
messg said: AS an end user, Dalle3 though can be accessed directly either via OAI or Azure. For azure, you need to have an azure account set up and request access to the OAI services. You need a registered business for this. Once you have access, you can tweak the filters to an extent although you still can't get full unrestricted. You can either then use their playground to generate manually or write your own scripts to run through the API. It should also be said that you pay per generation so it isn't cheap to produce large volumes.
Thank you for the reply, really appreciate it. I might look further into costs and requirements for this.