Photographers Push Back on Facebook's 'Made with AI' Labels Triggered by Adobe Metadata. Do you agree “‘AI was used in this image’ is completely different than ‘Made with AI’”?
I think every touch up besides color correction and cropping should be labeled as "photoshopped". And any usage of AI should be labeled as "Made with AI" because it cannot show which parts are real and which are not.
Besides, this is totally a skill issue. Removing this metadata is trivial.
DOD Imagery guidelines state that only color correction can be applied to "make the image appear the same as it was when it was captured" otherwise it must be labeled "DOD illustration" instead of "DOD Imagery"
People are complaining that an advanced fill tool that’s mostly used to remove a smudge or something is automatically marking a full image as an AI creation. As-is if someone actually wants to bypass this “check” all they have to do is strip the image’s metadata before uploading it.
No - I don't agree that they're completely different.
"Made by AI" would be completely different.
"Made with AI" actually means pretty much the exact same thing as "AI was used in this image" - it's just that the former lays it out baldly and the latter softens the impact by using indirect language.
I can certainly see how "photographers" who use AI in their images would tend to prefer the latter, but bluntly, fuck 'em. If they can't handle the shame of the fact that they did so they should stop doing it - get up off their asses and invest some time and effort into doing it all themselves. And if they can't manage that, they should stop pretending to be artists.
I think it is a bit of an unclear wording personally. "Made with", despite technically meaning what you're saying, is often colloquially used to mean "fully created by". I don't mind the AI tag, but I do see the photographers point about it implying wholesale generation instead of touchups.
or... don't use generative fill. if all you did was remove something, regular methods do more than enough. with generative fill you can just select a part and say now add a polar bear. there's no way of knowing how much has changed.
ai denoise, ai masking, ai image recognition and sorting.
hell, every phone is using some kind of "ai enhanced" noise reduction by default these days. these are just better versions of existing tools than have been used for decades.
Can't wait for people to deliberately add the metadata to their image as a meme, such that a legit photograph without any AI used gets the unremovable made with ai tag
Seriously though, "AI" itself is misleading but if they want to be ignorant and whiny about it, then they should be labeled just as they are.
What they really seem to want is an automatic metadata tag that is more along the lines of "a human took this picture and then used 'AI' tools to modify it."
That may not work because by using Adobe products, the original metadata is being overwritten so Thotagram doesn't know that a photographer took the original.
A photographer could actually just type a little explanation ("I took this picture and then used Gen Fill only") in a plain text document, save it to their desktop, and copy & paste it in.
But then everyone would know that the image had been modified - which is what they're trying to avoid. They want everyone to believe that the picture they're posting is 100% their work.
We've been able to do this for years, way before the fill tool utilized AI. I don't see why it should be slapped with a label that makes it sound like the whole image was generated by AI.
This isn't really Facebook. This is Adobe not drawing a distinction between smart pattern recognition for backgrounds/textures and real image generation of primary content.
I don't think that's fair. AI wont turn a bad photograph into a good one. It's a tool that quickly and automatically does something we've been doing by hand untill now. That's kind of like saying a photoshopped picture isn't "good" or "real". They're all photoshopped. Not a single serious photographer releases unedited photos except perhaps the ones shooting on film.
Even finns photographers touch up their photos, either during development by adjusting how long they sit in one or the chemical processes or by using different methods of shaking/mixing processes and techniques.
If they enlarge their negatives on photo paper they often have tools to add lightness and darkness to different areas of the paper to help with exposure, contrast and subject highlighting. AKA. Dodging and burning which is also available in most photo editing software today.
There are loads of things to do to improve developed photos and been something that has always been something that photographers/developers do. People who still go with the "Don't edit photos" BS are usually not very well informed about photo history and techniques of their photography inspirations.
The image looks like OP cherry picked some replies in the original thread. I wonder how many artists still want AI assisted art to be flagged as such.
EDIT The source is also linked under the images. They did leave out all the comments in favour of including AI metadata, but naturally they're there in the source linked under the images.
It's unreasonable to make them illegible for no good reason; you could've included them as-is, possibly in multiple, smaller images. It's also far more common to just share a link rather than an image post, as we'll have to see the link anyway.
I didn't see the source, though, I've updated my comment for that.
I believe AI art should definitely be labeled to minimize people being mislead about the source of the art. But at the same time the OP on the Adobe forums post did say they used it as any other tool for touching up and fixing inconsistencies.
If I were to for example arrange a photoshoot with a model and they happened to have a zit that day on their forehead of course I'm gonna edit that out. Or if I happened to have an assistant with me that got in the shot but I don't want to crop in making the background and feel of the photo tighter I would gladly remove that too. Sure Adobe already has the patch, clone and even magic eraser tool (Which also uses AI, that might or might not mark photos) to do these fix-ups but if I can use AI, that I hope is trained on data they're actually allowed to train on, I think I would prefer that because if I'm gonna spend 10 to 30 minutes fixing blemishes, zits and what not I'd much prefer to use the AI tools to get my job done quicker.
If the tools were however used to rigorously change, modify and edit the scene and subject then for sure, it might be best to add that.
Wouldn't it be better to not discourage the use of editing tools when those tools are used in a way that just makes one's job quicker? If I were to use Lightrooms subject quick selection, should it be slapped on then? Or if I were to use an editing preset created with AI that automatically adjusts the basic settings of an image and further my editing from that, should the label be created then? Or if I have a flat white background with some tapestry pattern and don't want to spend hours getting the alignment of the pattern just right as I try to fix a minor aspect ratio issue or want to get just a bit more breathing room on the subject and I use the mentioned AI tool in the OP.
Things OP mentioned in his post and the scenarios I mentioned are all things you can do without AI anyways it just takes a lot longer sometimes, there's no cheating in using the right tool for the right job IMO. I don't think it's too far off from someone who makes sculptures in clay uses an ice scream scoop with ridges to create texture or a Dremel to touch up and fix corners. Or a painter using different tools and brushes and scrapers to finish their painting.
Perhaps a better idea would be if we want to make the labels "fair" there should also be a label that the photo has been manipulated by a program in general or maybe add a percentage indicator to see how much of it has been edited specifically with AI. Slapping an "AI" label on someone because they decided to get equal results by using another tool to do normal touch-ups to a photo could potentially be damaging to ones career and credibility when it doesn't say how much of it was AI or in what reach, because now there's the chance someone might be looking for their next wedding photographer and be discouraged because of the bad rep regarding AI.
Indeed, if the AI was trained based on theft it's neither right on their part or ethical on mine.
I did some searching but sadly don't have time to look into it more but there were some concerning articles that would suggest they have either used shady practices to get their training data or users having to manually check an opt out box in the app settings.
I can't make an opinion on it right now before looking into it more but my core argument about using AI itself in this manner, even if that data was your own on your own trained AI using allowed resources, I still believe somewhat holds.
I'm not sure of the complaint, is the tag not accurate? If you use AI to make something are you not making it with ai? Like if I use strawberry to make a cake would the tag made with strawberries be inaccurate?
Like I failed to see the argument, if you don't want to be labeled as something accurate don't use it otherwise deal with it.
I do think it's a problem when 100% of people seeing "made with AI" will assume the entire thing is AI-generated, even if all you did was use AI for a minor touch-up. If it's really that trigger happy right now, I think it'd make sense for it to be dialled down a bit.
The complaint the photographer is making is that it's an actual photograph where a small portion is made or changed with AI.
They list expanding the edges of the image to change the aspect ratio, and removing flaws or unwanted objects etc.
Removing flaws and objects at least is a task that predates modern computers - people changed the actual negatives - and tools to do it have improved so much a computer can basically do it all for you.
I think people should just say how they modified the image - AI or not - since airbrushed skin, artificial slimming, and such have been common complaints before AI manipulation, and AI just makes those same problematic things easier.
The biggest use of AI in my editing flow is masking. I can spend half an hour selecting all the edges of a person as well as I can, or I can click the button to select people. Either way I do the rest of my edits as normal.
I agree pretty heartily with this metadata signing approach to sussing out AI content,
Create a cert org that verifies that a given piece of creative software properly signs work made with their tools, get eyeballs on the cert so consumers know to look for it, watch and laugh while everyone who can't get thr cert starts trying to claim they're being censored because nobody trusts any of their shit anymore.
Bonus points if you can get the largest social media companies to only accept content that has the signing and have it flag when signs indicate photoshopping or AI work, or removal of another artist's watermark.
The opposite way could work, though. A label that guarantees the image isn't [created with AI / digitally edited in specific areas / overall digitally adjusted / edited at all]. I wonder if that's cryptographically viable? Of course it would have to start at the camera itself to work properly.
There are ways to secure signatures to be a problem to recreate, not to mention how the signature can be unique to every piece of media made, meaning a fake can't be created reliably.
To appease the artists worried about "fake" art somehow replacing the "real"art, while the big social somehow profits. They just didn't think leopards would eat THEIR faces...
Did they just e.g. remove a passing car from the background*, and will tags on some images lead to untagged fake images being trusted more? Oh this fun new world we’re in.
*as someone else pointed out, if it was a minor edit, was the underlying technology using legit training data or unlicensed stuff
It’s exaggerated but it gets the point across: I too would like to know if AI tools were used to make even part of the image.
There’s a reason any editing is banned from many photography contests.
If they want to make a distinction between “made using AI” and “entirely AI generated”, sure. But “made using AI” completely accurately describes an image that used AI to generate parts of the image that were inconvenient in the original photo.