Jump to content
  • 0

About NFSW AI gens in the card maker


cr47t

Question

Some time after the original No AI thread, a user reported in a status an instance where the card maker's current AI generator made a NSFW image where no such detail was requested... and after procrastinating for some time, I thought as the maker of the original no AI thread, and someone concerned with the future of YCM, that I'd make a thread about it.

To be honest, I hadn't even thought of this as a possibility when I made the previous thread, but given most of the site's users are here for the card maker (not so much the forum) this could be a problem for YCM going forward. The current forum rulebook, specifically in rule 4, details a "zero tolerance policy" for "explicit pornography and graphic content (will be penalized with a permanent ban)" - how do we square this precedent with the AI and its not-very-context-considerate mode of operation?

In addition, as @Rayfield Lumina illustrated in the status replies, given the presumed average age of people who use the card maker regularly, this could get them in trouble with others outside the forum, possibly prompting one of those people to take it up against the forum with an outside authority. No, I'm not some prude puritan who goes out of his way to fake outrage over things like this, and while I don't think the site should be beholden to such people either, they've got a way with people who actually make the decisions - not necessarily YCMaker or any actual site mods or admins, but supposing a higher up who is in charge of web infrastructure that YCM relies on decided to cut us off... ( just to be clear I don't know anything about whether they would, I'm just guessing here)

@Rayfield Lumina also pointed out that we don't know much about the AI in the card maker - what it trains on, what it cites in the card to come up with the prompt, etc. - that we don't have much of a negative prompt to tell the AI to not do specific things, and there is probably no reliable safeguard against this kind of thing happening except to either disable the AI altogether, dig into the model's training data and dig out the NSFW one by one (probably too demanding of a task for anyone here) or switch the model entirely. In the previous AI thread YCMaker said that "Once Adobe Firefly is commercially available, I'll look into switching into that" - looking it up now, it's hard for me (someone out of the loop with the ai hype) to tell if it's implementable yet, based on the top result in a google search, so if someone could inform me/us further that would be nice.

In the meantime I am making this thread to urge @YCMaker to take some kind of action about this issue. Preferably something big or moderately impactful to slow this issue down before it manifests on the forums (like by some spammer, bots, or other troublemaker) but even something small would be appreciated for now (for example, some disclaimer on the AI dropdown on the card maker page that forbids the posting of such images to the forum and/or rules out a guarantee of avoiding the NSFW images, etc.) Personally I think the safest option is to disable AI at least for users who haven't been around for long, but it's your call.

Thanks,

-cr47t and the YCM forum

P.S. TO USERS: Please keep the conversation civil - if all you have to say is attacks or strawmanning of other people in the discussion, we'd rather you not do it.

EDIT 9/12/2023 - Please stay on topic

Edited by cr47t
amended note at bottom of post to not overblow the issue and let it speak for itself or something idk
  • Like 1
Link to comment
Share on other sites

15 answers to this question

Recommended Posts

  • 3
24 minutes ago, Horu said:

Admittedly, after seeing the status, I tested it. Producing the image, even intentionally, takes at least 5+ tries in the generator. While I understand the concern, you do have to understand that when an image generator uses pretty much all of google as a reference to compile something, NSFW images are likely to happen on occasion.

If using it means "NSFW images are likely to happen" then that's exactly the problem isn't it?

  • Like 1
Link to comment
Share on other sites

  • 2
2 hours ago, Zamazenta the OS-Tan Fan said:

But that is like giving a free pass infect artists don;t even want you using their art DRM is never a good thing but if they have to DRM their images than they will.

That is the entire point of DRM. Google protects your intellectual property. Although, enabling DRM prevents documents greater than 128MB from being sold on the playstore. But yeah, ultimately your work is protected.

1 hour ago, TheImmaculateDreadLordCowCow said:

If using it means "NSFW images are likely to happen" then that's exactly the problem isn't it?

This is a correct assessment. This is the point of testing.

Link to comment
Share on other sites

  • 2
On 9/10/2023 at 5:18 PM, Zamazenta the OS-Tan Fan said:

Yeah I agree with you that it can generate some offensive stuff but AI Art in General feels unethical because it is trained on the works of many talented and frustrated artists that still don't like it some people love to defend this bullcrap but if you get into the law and ask many artists yourself they will give you angry responses I'm not trying to hate on it because its AI I hate it for a different reason that it infringes on the right of many people's work.

People are shrilling AI like there is no fucking tomorrow and what kind of long term side effects can this lead into the only reason companies want to push it so they can do bad and greedy stuff just ask @Raikoh130 and he will give his take on it also have any of you seen the movie Skynet or The Terminator many of us that watched them should be worrying about the upcoming and pending doom on are dumb and stupid human society not like things are already hell politically on both the Internet and in real life and technically with  social media addictions but I feel as a Gen Z my generation are being dumbed down by these fucking companies such as META formerly know as Facebook before the Zuck rebranded it in 2021, Google, Microsoft, OpenAI, Apple, Amazon (company) and others will all have a hand in destroying our society it has already been destroyed by them just imagine the amount of damage this can do in the future 10-20 years from now and we our all ignoring the warning and red flags like fucking retards you all are it just frustrates me that there are a whole bunch of fucking ignorant retards that will defending this bullshit I like now I know how dumbed down many people are by these fucking companies and the other drama with the politically stuff is just a distraction from the real dangers in this world I'm also a big hater of Smart Phones even despite being Gen Z I despise those motherfucking pieces of shits they are just a fucking distraction and just fucking dumb down people but these companies will attack you if you say something bad about them its all in the name of profits and companies want to spend less money and pay less and replace humans with AI/Robots in the work place I think you should all watch that episode of the where Mr. Burns replaces all of his human employs with robots, except Homer Simpson https://en.wikipedia.org/wiki/Them,_Robot it's a good episode to watch its not far from the truth at all you can go read the Wikipedia article than we have a movie like WALL-E it gives off a whole dystopian vibe if you haven't watched the movie before I think you should give it a good watch you'll like the movie and it even teaches you some things and has a message about the future of our society if where not careful https://en.wikipedia.org/wiki/WALL-E But yeah that was my entire rant on AI and why I just hate the thought of AI @cr47t can also explains his take by quoting this post.

                  

Zama. Next time, don't make a wall of text ranting. You're coming off more heated than necessary, and even using certain words that should be discouraged. Plus, having more than two "fucks" in your wall of text kills impact. Think before you post

  • Thanks 1
Link to comment
Share on other sites

  • 1
1 minute ago, Horu said:

That is the entire point of DRM. Google protects your intellectual property. Although, enabling DRM prevents documents greater than 128MB from being sold on the playstore. But yeah, ultimately your work is protected.

This is a correct assessment. This is the point of testing.

Yeah imagine having to resort to DRM I may be anti-DRM but I wouldn't mind if when they post it has DRM that prevents AI ART Databases from using it. Also DRM is never a good thing you know the reason why I and many people hate DRM is because it is always used in the wrongest of ways and used in an unethical way like Video Game Anti-Cheat that starts-up with the Windows NT Kernel and run on the user's Windows System with Kernel level access which is ring 0 or DRM that prevents you from using what ever kind of OS you want imagine if Microsoft someday decides to make HP, Dell, Asus, Acer, Lenovo, MSI, etc to put DRM on your Machine by putting it directly into the UEFI which is the firmware of your motherboard/device that prevents you from running/installing any kind of Operating System besides Windows *cough* *cough*  Microsoft not wanting you to use Linux on a Windows Laptop or Desktop that is what I'm afraid of oh ho maybe I shouldn't be giving them any ideas @Horu but DRM can protect artwork from being stolen and trained on AI but DRM can also be used for evil like locking down user's devices and punishing them if they try to modify it any way shape or form *cough* *cough* *cough* *cough* *cough* *cough* *cough* Apple.

Link to comment
Share on other sites

  • 0
Just now, Horu said:

Admittedly, after seeing the status, I tested it. Producing the image, even intentionally, takes at least 5+ tries in the generator. While I understand the concern, you do have to understand that when an image generator uses pretty much all of google as a reference to compile something, NSFW images are likely to happen on occasion.

But that is like giving a free pass infect artists don;t even want you using their art DRM is never a good thing but if they have to DRM their images than they will.

Link to comment
Share on other sites

  • 0
16 hours ago, Zamazenta the OS-Tan Fan said:

Yeah imagine having to resort to DRM I may be anti-DRM but I wouldn't mind if when they post it has DRM that prevents AI ART Databases from using it. Also DRM is never a good thing you know the reason why I and many people hate DRM is because it is always used in the wrongest of ways and used in an unethical way like Video Game Anti-Cheat that starts-up with the Windows NT Kernel and run on the user's Windows System with Kernel level access which is ring 0 or DRM that prevents you from using what ever kind of OS you want imagine if Microsoft someday decides to make HP, Dell, Asus, Acer, Lenovo, MSI, etc to put DRM on your Machine by putting it directly into the UEFI which is the firmware of your motherboard/device that prevents you from running/installing any kind of Operating System besides Windows *cough* *cough*  Microsoft not wanting you to use Linux on a Windows Laptop or Desktop that is what I'm afraid of oh ho maybe I shouldn't be giving them any ideas @Horu but DRM can protect artwork from being stolen and trained on AI but DRM can also be used for evil like locking down user's devices and punishing them if they try to modify it any way shape or form *cough* *cough* *cough* *cough* *cough* *cough* *cough* Apple.

Here's the thing. DRM already exists on google and you have a choice of wether or not to use it. Google's DRM plugin is called Widevine. It's essentially a Digital Rights Management tool that allows you protect your work. That being said, Microsoft can't force companies like HP, Samsung, MSI, etc to be Windows exclusive because, at the end of the day, Microsoft has no rights to the computers that their OS is installed on since they didn't create nor build it. Also, DRM only applies to digital works. What you're referring to is if Microsoft were able to get these major hardware companies to sign a contract, which isn't very likely happen.

Link to comment
Share on other sites

  • 0
2 hours ago, Zamazenta the OS-Tan Fan said:

Yeah but Google wants to DRM Websites with WEI so this could give Microsoft the stupid ideas since these OEMS bow down to Microsoft making it so when people buy these OEMs computers they can't oh let's just say install Linux effectively Hard Locking you into shitty Windows that would be hardware/UEFI/OS DRM the DRM is just saying "We're Anti-Linux".

Anyway, back on topic. We need to figure out a solution where the card maker doesn't produce NSFW images.

  • Like 1
Link to comment
Share on other sites

  • 0
On 9/12/2023 at 9:50 AM, cr47t said:

@Horu 

Before I say anything else I'd like to share thanks for testing this, given it takes 5+ tries and only 6~ generations permitted per day for an account, I think we can rule out more casual users going for such details on purpose.

The scenarios I brought up in the OP where someone would try to punish the site for it would probably be if a younger user gets it on accident and it gets the attention of the parents, especially if they're the context-negligent "i let the internet do my job for me" kind. (Note: Despite such ipad parents tending to make the news a lot and appear in content a lot too, I don't think these kinds of parents are anything of a plurality in today's society, thankfully.) I don't want anyone getting into trouble because of something they couldn't control (the AI results) that was provided by us, and it could possibly have downstream effects on the site (especially if they manage to mobilize an authority) which are things I am trying to avoid with this petition/post,

I have to wrap it up now and get to other things so if I was unclear or if I overlooked another consequence of this part of the AI in the cardmaker please let me know (in a respectful manner) so I can clear things up when I have the time again

Testing is important. Although, YCMaker did say that he was going to switch to Adobe's generator when it got released. But since this is the temporary option, it is important to know how many tries it takes to make certain things happen. But yeah, running this casually, you shouldn't run into anything NSFW very often. That being said, I tested using prompts/descriptions along the lines of what would commonly appear on a ygo card. I'm fairly certain that is you outright give a NSFW prompt, you'll get a NSFW result. Aside from that, casually entering a prompt has mimimal risk of giving NSFW results.

  • Like 1
Link to comment
Share on other sites

  • -1

Admittedly, after seeing the status, I tested it. Producing the image, even intentionally, takes at least 5+ tries in the generator. While I understand the concern, you do have to understand that when an image generator uses pretty much all of google as a reference to compile something, NSFW images are likely to happen on occasion.

Link to comment
Share on other sites

  • -1
1 hour ago, Horu said:

Here's the thing. DRM already exists on google and you have a choice of wether or not to use it. Google's DRM plugin is called Widevine. It's essentially a Digital Rights Management tool that allows you protect your work. That being said, Microsoft can't force companies like HP, Samsung, MSI, etc to be Windows exclusive because, at the end of the day, Microsoft has no rights to the computers that their OS is installed on since they didn't create nor build it. Also, DRM only applies to digital works. What you're referring to is if Microsoft were able to get these major hardware companies to sign a contract, which isn't very likely happen.

Yeah but Google wants to DRM Websites with WEI so this could give Microsoft the stupid ideas since these OEMS bow down to Microsoft making it so when people buy these OEMs computers they can't oh let's just say install Linux effectively Hard Locking you into shitty Windows that would be hardware/UEFI/OS DRM the DRM is just saying "We're Anti-Linux".

Link to comment
Share on other sites

  • -1

@Zamazenta the OS-Tan Fan if you have any concerns with DRM and AI that don't have to do with the specifics of the problem raised in the original post, please feel free to make a new thread about it somewhere else (such as General subforum for example) so we can have a dedicated conversation about it there. (Or if you think DRM could negatively affect the site make a post about it in Q&H, or Comments and Suggestions, etc. so it has its own thread)

My point is, this is a thread about the cardmaker's AI producing NSFW images and the effect it can have on the site, not a thread about AI outside of the cardmaker, not a thread about DRM, etc. and since this is a potentially serious issue I expect people to stay on topic. I'm only going to say this once here, and if this derailing digression continues for one post more I will reach out to a site mod.

@Horu 

On 9/10/2023 at 7:15 PM, Horu said:

Admittedly, after seeing the status, I tested it. Producing the image, even intentionally, takes at least 5+ tries in the generator. While I understand the concern, you do have to understand that when an image generator uses pretty much all of google as a reference to compile something, NSFW images are likely to happen on occasion.

Before I say anything else I'd like to share thanks for testing this, given it takes 5+ tries and only 6~ generations permitted per day for an account, I think we can rule out more casual users going for such details on purpose.

The scenarios I brought up in the OP where someone would try to punish the site for it would probably be if a younger user gets it on accident and it gets the attention of the parents, especially if they're the context-negligent "i let the internet do my job for me" kind. (Note: Despite such ipad parents tending to make the news a lot and appear in content a lot too, I don't think these kinds of parents are anything of a plurality in today's society, thankfully.) I don't want anyone getting into trouble because of something they couldn't control (the AI results) that was provided by us, and it could possibly have downstream effects on the site (especially if they manage to mobilize an authority) which are things I am trying to avoid with this petition/post,

I have to wrap it up now and get to other things so if I was unclear or if I overlooked another consequence of this part of the AI in the cardmaker please let me know (in a respectful manner) so I can clear things up when I have the time again

Link to comment
Share on other sites

  • -2

Yeah I agree with you that it can generate some offensive stuff but AI Art in General feels unethical because it is trained on the works of many talented and frustrated artists that still don't like it some people love to defend this bullcrap but if you get into the law and ask many artists yourself they will give you angry responses I'm not trying to hate on it because its AI I hate it for a different reason that it infringes on the right of many people's work.

People are shrilling AI like there is no fucking tomorrow and what kind of long term side effects can this lead into the only reason companies want to push it so they can do bad and greedy stuff just ask @Raikoh130 and he will give his take on it also have any of you seen the movie Skynet or The Terminator many of us that watched them should be worrying about the upcoming and pending doom on are dumb and stupid human society not like things are already hell politically on both the Internet and in real life and technically with  social media addictions but I feel as a Gen Z my generation are being dumbed down by these fucking companies such as META formerly know as Facebook before the Zuck rebranded it in 2021, Google, Microsoft, OpenAI, Apple, Amazon (company) and others will all have a hand in destroying our society it has already been destroyed by them just imagine the amount of damage this can do in the future 10-20 years from now and we our all ignoring the warning and red flags like fucking retards you all are it just frustrates me that there are a whole bunch of fucking ignorant retards that will defending this bullshit I like now I know how dumbed down many people are by these fucking companies and the other drama with the politically stuff is just a distraction from the real dangers in this world I'm also a big hater of Smart Phones even despite being Gen Z I despise those motherfucking pieces of shits they are just a fucking distraction and just fucking dumb down people but these companies will attack you if you say something bad about them its all in the name of profits and companies want to spend less money and pay less and replace humans with AI/Robots in the work place I think you should all watch that episode of the where Mr. Burns replaces all of his human employs with robots, except Homer Simpson https://en.wikipedia.org/wiki/Them,_Robot it's a good episode to watch its not far from the truth at all you can go read the Wikipedia article than we have a movie like WALL-E it gives off a whole dystopian vibe if you haven't watched the movie before I think you should give it a good watch you'll like the movie and it even teaches you some things and has a message about the future of our society if where not careful https://en.wikipedia.org/wiki/WALL-E But yeah that was my entire rant on AI and why I just hate the thought of AI @cr47t can also explains his take by quoting this post.

                  

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Answer this question...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...