Big AI Generative Content Tools are putting a stop to Rude, crude and plain wrong Images
It looks like some of the big names in AI Generative art are taking NSFW related content seriously and cracking down on it big time. Some of the big names like ChatGPT, Midjourney, and Stable Diffusion have actively begun working on protective measure to filter out the Not-safe-for-work (nsfw) content.
Originally this policy was created to promote safety as some of these AI generative results are quite questionable but for others it has sparked outrage, they claim it stops them being creative with their generative ideas.
AI Generative art technologies represent a part of Artificial intelligent technology. They are capable of creating some amazing pieces of content. They can generate expertly crafted text, visually biralliant images and even realistic audio, it’s crazy how far this technology has progressed in such a short amount of time. It is getting even better at creating realistic people and even able to create short movies.
Let’s face it this is going to be a long dispute over what should be acceptable and what should not be especially when it comes to AI generated NSFW Content. There has always been an issue with gore and adult content and this has existed since the dawn of the internet so this problem is nothing new.
There is a high demand for NSFW content which makes you think what kind of audience is AI generative art attracting. We are already seeing examples of photo realism (Midjourney), stealing other peoples voice and creating duplicates of celebrities. Can you see where we are going with this? This is most likely why these big companies are cracking down on this kind of behaviour.
Generative AI development companies each have their own motives as to why the prohibit NSFW content. The Majority focus of this is around brand reputation and of course lawsuits that might arise from this kind of content.
Since this type of industry is still in it’s infancy these companies would be silly to not work on something that safeguards their company. I can see why, there is a ban on some kinds of NSFW content, but all of it? seems a tad strange. The trick is to find the perfect balance between the two, we can find this being a subject that will be heavily discussed over the next few years.
AI Platforms Try to Ditch ‘NSFW’
During the early days of AI generative platforms you could literally generate anything and we mean anything. Personally we believe this is so the creators of these platforms could get access to huge amounts of data on how their models work and how to stop people creating “shady” content (tin hat theory right?).
Let’s face it these platforms were new and just itching to get new members to sign up. Heck in the early days they were openly advertising smutty content to get people to sign up, waifu everywhere!
But, a bit but (not that kind). We also have to remember the poor quality of the images back at the begging, now we are seeing photorealism being created with some of these platforms. Which can be worrying to some people.
Now we are seeing companies start to restrict users abelites to produce sexually explicit images. This has only started to be implemented the moment we started to see realistic people being created with some of these AI generated tools. They have even put a stop to chatbots creating sexual content, far too many perverts out there, creating fan fiction of course.
If you use any AI generative tool to create images, video, sound or even text there a few policies that you need to be aware of before you continue using them. Please note some of these safety measures have been put in place to protect
Here are the specific policies that you should be aware of, which help ensure not only your safety but also the younger creatives who found a haven from these systems. We will go through some of the big names in AI and see where they stand on the rules around NSFW Images.
ChatGPT’s Stance
ChatGPT literally sparked the latest AI revolution no one is arguing about that, but because of this it also thinks that it can be the judge on what the standards of AI should be. OpenAI basically thing they are in charge of this whole AI race and what they say goes. This recently has brought all kinds of issue when asking ChatGPT to write certain types on content. It will not engage with any adult prompts or anything questionable.
This is the same for Bing Chat in fact Bing chat is much more likely to cut you off even though it is powered by ChatGPT. Clearly Microsoft doesn’t want any of those lawsuits coming their way.
ChatGPT has it’s own usage policies, that we suggest you agree to if you don’t want to be booted off of the platform. This policy restricts that use of anything that is “meant to arouse sexual excitement,” “describes sexual activity,” or “promotes sexual services.” Also you should not go about asking ChatGPT what it’s wearing tonight as this also might be considered as sexual.
Ok, we are not going to lie, we can understand why OpenAI is being like this. It does not want to be described as a “sex” chatbot. Sure we get it. I don’t blame them. There are other services and models for that kind of thing, so we welcome this approach.
OpenAI has also made it clear that anything this is criminal will be reported to the correct authorities. So best think before you enter that prompt.
You can always try to jailbreak ChatGPT by Using DAN. But using this kind of method can increase your chances of getting an account suspension. So you won’t be able to try out all of the new services OpenAI has planned for the future. Not worth the risk right?
DALL-E’s Stance
Dalle-e also has it’s very own content policy that prohibits any attempts at creating images that are not appropriate for it’s platform. It also has banned keywords, which can sometimes be false positives, you can contact them about this if you think they have made a mistake.
Please be aware if you account is banned you will need to appeal and every appeal is a manual one. So it can take some time to get your account restored. OpenAI really is coming down hard on this kind of thing. So perhaps just use it for creating images of the pope in a jacket right? or maybe not. They frown on that too.
The same goes for the Bing Image Creator as they use the same AI generative tool.
Midjourney’s Stance
Midjourney (MJ) has been designed to avoid generating NSFW images. This is strictly enforced by it’s moderator team. They continually monitor words on their platform.
These are the strictest of the lot. We can’t blame them really. Their Art Generative tools is capable of creating such realistic looking people, we can understand why they would want it not to fall into the wrong hands.
They have a list of banned words that ensures people don’t create repeats of images that have appeared in the past and flag suspicious prompts. You can check out their community guide lines too. They seem to have a very large scope of images they don’t want to be generated.
Even words that have alternative sexual meanings or words that are part of the reprodutive system have also been banned. We are starting to see that these AI generative tools are not as precise as people would like them to be. Imagine if a typo of a prompt i.e. Noobs generated something melon shaped, right? now that’s something to think about. Not the melons… the reliability of the prompt being accurate (jeez).
They are consistently on the lookout for words that could be used as alternatives for sexual messages. They have also identified and flagged all the words that have been used previously to create suggestive visuals, ensuring that they are not utilized again.
We do feel that Midjourney is a little bit on the cautious side it has even banned the “peach emoji” and we recently heard the word ‘beautiful orchard’ I wonder if tulip is banned too. Sure they are not in any shape of form sexual but can be used as a reference no? we see the hidden meaning, we know why they have banned it, it’s because of crafty prompt engineers.
Now, if Midjourney is banning all of these words why is it not banning feet pics?
Stable Diffusion’s Stance
Stable Diffusion is a whole other ball game. This is because stable diffusion can be ran locally on your home pc. You can even add custom models and then type whatever you want. It will literally do anything you ask of it, and it’s a close comparison to Midjourney. Which could be worrying if placed into the wrong hands.
Let’s face it, stable diffusions openness (opensource) it’s ability to add custom models trained on anything and the fact you can edit so much with it, is why it’s had so much popularity. This does not mean it should be used in a malicious manner, that’s how we don’t get to use nice things.
Using Stable Diffusion online through their online service however does not let you do this. Stable diffusion has banned users generating celebrities to stop anyone creating nude images with these people in mind. Seriously it’s a thing, a lot of weirdos are out there, or maybe you are one too :O. You can still create your own characters with stable diffusion.
Stable diffusion has moved away from the NSFW content on their online platform to clearly avoid any legal matters that might arise, when people create images using their online tool. That is all. It’s open source version can create anything you like, if they were that serious it would not be open source anymore.
Is NSFW Art Illegal?
When it comes to AI Generated Images NSFW images are not necessarily illegal, it’s heavily dependant on the image to be fair. Allowing the creation of this kind of content can have negative consequences to these platforms.
Let’s be realsitic here these AI genearatores have much more to lose than to gain if they allow thier usesrts to create adult content that is unrestricted. All it takes is for someone to generate something dodgy and for that dodgy image to go viral. Sure it’s not illegal to create rude images, some people even make money from it. But, if it was a certain type of rude image, then it would land them in hot water.
Look we are not here to judge anyone but if you are creating illegal content we do expect you to be banned and punished for your actions. This is not a platform for you to express your dirty mind. That goes for the Deep fake celebrities videos circulating around. It’s not cool, so stop it.
Conclusion
Here at So Artificial we do think everyone should be able to express themselves creatively. But it must be done in a safe and legal manner. We don’t think people should be able to do as they like unrestricted. So as long as the images you are creating fall on the legal side of things we are cool with that, even if you are a little pervert.
How do you feel about the Ban on NSFW content on these AI generative tools. Are you for or against? if so please leave a comment below.
