Commons:Deletion requests/Files uploaded by Hyju
Out of scope - unused and unlikely to be used because they contain AI gobbledygook text.
The Squirrel Conspiracy (talk) 09:27, 6 December 2023 (UTC)
- Keep a few – Too indiscriminate. Many of those are low-quality where I may favor a delete outcome. However, AI gobbledygook text being included somewhere is (usually) not a valid rationale for deletion since it can be easily removed (and replaced if needed). It just takes a minute or so using any of the three methods explained here and the uploader should have done this before uploading. I suggest the uploader does that now for some of the images. I don't know why people upload such low-quality AI art when there is and can be much higher-quality useful images of that kind. I think File:Female barbarian from space vs giant ant.jpg and File:Ai-generated tarzanide.jpg should be kept (especially if uploader fixes the text), the rest seems deletable.
- Prototyperspective (talk) 11:32, 6 December 2023 (UTC)
- The updater fixed the first images and the second one does not have "gobbledygook text" (which is the deletion rationale) and is in use so by definition educationally valuable according to Wikimedia Commons policy. Not sure if policies matter here anymore though, this place is turning into a totalitarianism of unilateral decisions by admins supported by one or or three refuted opinions. Prototyperspective (talk) 09:10, 7 December 2023 (UTC)
- Delete all. No obvious educational use for these images, even if the text were corrected or cropped out. AI images should be generated "on demand" to satisfy specific project needs, not "on spec" or to demonstrate capability. Omphalographer (talk) 17:55, 6 December 2023 (UTC)
Deleted: per nomination & Omphalographer. --Infrogmation of New Orleans (talk) 17:34, 13 December 2023 (UTC)
The uploader is apparently confusing Commons with a web space provider. Massively AI-created images, some of them clear violations of copyrights on characters, some of them pure fantasy passed off as historical reality, some of them even just plain racist. Not one of the images can be usefully used in an educational sense. On the contrary, a large proportion of the images are problematic in the opposite direction and spread the opposite of secure education, fantasy and speculation. Here you can see once again that AI-generated images are currently simply a problem and are in no way even close to solving problems.
- File:Suggestive romance comics.jpg (added January 8) — Preceding unsigned comment added by Counterfeit Purses (talk • contribs) 16:25, 9 January 2024 (UTC)
Marcus Cyron (talk) 00:35, 27 December 2023 (UTC)
- I really didn't understand the arguments, I can even agree that some old arts are not of good quality, but I think they are all useless, I also didn't understand the supposed racism, as soon as I discovered that you can generate arts that simulate comics, I thought about putting characters blacks and Asians like black cowboys that have been rediscovered by history and are even used in fiction.Hyju (talk) 00:50, 27 December 2023 (UTC)
- Specially these files, I don't understand, too:
- Both are mythological entities in afro-brazilian Candomblé. André Koehne TALK TO ME 04:21, 27 December 2023 (UTC)
- In fact, I only generated these two after seeing one of Oshun. Hyju (talk) 09:53, 27 December 2023 (UTC)
- I really didn't understand the arguments, I can even agree that some old arts are not of good quality, but I think they are all useless, I also didn't understand the supposed racism, as soon as I discovered that you can generate arts that simulate comics, I thought about putting characters blacks and Asians like black cowboys that have been rediscovered by history and are even used in fiction.Hyju (talk) 00:50, 27 December 2023 (UTC)
- Delete Out of scope. The Squirrel Conspiracy (talk) 03:08, 27 December 2023 (UTC)
- This is not an explanation and you stated that you vote (and nothing more than voting has been done) delete for all AI images, saying Due to both the copyright and ethical concerns, I am always in favor of deleting AI art, especially when we have any non-AI generated images that don't have those concerns. I've never seen you vote anything but for delete in AI-related DRs. The listed files are too indiscriminate and are often one very few relevant to a subject, many of them are in use, and a more discriminate subset may be worth deleting even though I don't see why these clearly AI-labeled images would be such a problem. Prototyperspective (talk) 10:57, 16 January 2024 (UTC)
- Keep indiscriminate, weird arguments— most or all of them are marked as “AI” so they’re obviously not supposed to be hoaxes, none of them I’ve seen obviously violate any copyrights, there’s no obvious “racism” whatsoever, and most importantly some are in legitimate use or are clearly in scope as illustrations of less well-known genres like cyberfunk or simply usable quality illustrations of things like anime art. The nominator is seemingly just indiscriminately targeting AI art out of personal prejudice. I wouldn’t oppose nominating some of the low quality, actually OOS or redundant images but we can’t just randomly cull dozens of images based mainly on how they were made. Dronebogus (talk) 06:34, 27 December 2023 (UTC)
- Keep – Indiscriminate, per Dronebogus, multiple of these are in use, and invalid false deletion rationale. some of them clear violations of copyrights on characters these should be deleted but you didn't name them and just made up a claim which I think is false. some of them pure fantasy passed off as historical reality none of these are claimed to be historical reality and they are clearly marked as AI-generated some of them even just plain racist very strange accusation that is not true at all. Not one of the images can be usefully used in an educational sense very false, many of them are the only or nearly one images that can help illustrate a concept (not necessarily on Wikipedia and with info that it's AI-made) such as an art style, are the only artistic depictions of archaic humans, illustrate how some kinds of comics or their subjects roughly looked like, or how to use AI. problematic in the opposite direction and spread the opposite of secure education again just a claim with no link to any policy or supporting explanation or even just referring to specific images fantasy and speculation is that meant to be an argument? Despite that WMC seems to have a bias against anything that aren't photographs, there very large amounts of fantasy-related images this image by NASA is also speculation File:Artistic depiction of a NASA lunar base.jpg or widely-used File:GAX-447.jpg. Fantasy and speculation are major aspects of human culture plus can be used to illustrate concepts (e.g. major subjects of fiction) and so on. A very small number of these images may be low quality out of scope but this is a strong keep for this indiscriminate unfounded nomination of a few files among over 100 million where these images are overall probably more useful than 90%. Prototyperspective (talk) 10:15, 27 December 2023 (UTC)
- Delete I agree with the comments of Omphalographer in the previous discussion. There is no need to store everyone's AI-generated images when they can be created on an as-needed basis. The images themselves are quite nice, but there is nothing unique or special about them and no clear educational purpose. Counterfeit Purses (talk) 04:35, 6 January 2024 (UTC)
- Some of them are the only images we have of topics like clockpunk— how is that not “unique or special”? More importantly, many of them are COM:INUSE! Nobody says “I think we should delete almost all pictures of grass and trees, they can be taken on an as-needed basis”. Dronebogus (talk) 07:19, 6 January 2024 (UTC)
- @Dronebogus You misunderstood me. I meant that they were not unique or special as AI-generated images. Hyju (or anyone else) could make similar images by the thousands whenever they desired. I suspect that if everyone suddenly decided to upload their pictures of grass and trees we would having a discussion about deleting those. You've seen COM:PENIS, right? Counterfeit Purses (talk) 15:34, 7 January 2024 (UTC)
- 1. They do and so far we don't. 2. Many of them are useful regardless of how difficult you hypothesize it to be to create them. While some may be worthy of deleting, the list is indiscriminate and includes many very useful ones. Prototyperspective (talk) 16:35, 7 January 2024 (UTC)
- @Dronebogus You misunderstood me. I meant that they were not unique or special as AI-generated images. Hyju (or anyone else) could make similar images by the thousands whenever they desired. I suspect that if everyone suddenly decided to upload their pictures of grass and trees we would having a discussion about deleting those. You've seen COM:PENIS, right? Counterfeit Purses (talk) 15:34, 7 January 2024 (UTC)
- Some of them are the only images we have of topics like clockpunk— how is that not “unique or special”? More importantly, many of them are COM:INUSE! Nobody says “I think we should delete almost all pictures of grass and trees, they can be taken on an as-needed basis”. Dronebogus (talk) 07:19, 6 January 2024 (UTC)
- Delete Although I'd say I'm on the weak side due to how indiscriminate this is. Although you could maybe argue the few images of "clockpunk" (whatever that is) could be kept. I'm not really convinced that every image of "X art style" is worth having though simply because we don't have other images of it. There's at least nothing in Commons:Project scope saying anything about how images related to a certain subject become more in scope the less we have of them. Otherwise people could upload random nonsense, call it "whatever art style", and we'd be forced to keep the images purely due to their novelty.
- Plus these images are extremely easy to re-create anyway. Has anyone bothered to look online to see if there's free licensed images of "clockpunk" by actually notable artists that they can upload? If not, then that should really be the default option here or really anywhere that has a gap in coverage. Not just to upload OOS amateur artwork and then act like it's the only option out there when it probably isn't. We should really be encouraging people adding "notable" artwork related to a specific genre to Commons though. Not whatever chicken scratch that's even slightly related to it that can easily be generated by AI. --Adamant1 (talk) 07:19, 8 January 2024 (UTC)
- Why is everyone so prejudiced against an artwork because an AI made it? The general assumption is that AI, unlike literally any other medium or mode of creating artwork, needs to be kept to the absolute bare minimum. Why? Dronebogus (talk) 11:12, 8 January 2024 (UTC)
- Comment: for reference, the following files are legitimately COM:INUSE and should be kept regardless of whether anyone thinks the others are in scope:
- File:Nibiru entering a collision course with Earth.jpg
- File:Hercolubus entering a collision course with Earth.jpg
- File:Ai-generated sword and planet hero.jpg
- File:Clockpunk Leonardo Da Vinci.jpg
- Dronebogus (talk) 11:22, 8 January 2024 (UTC)
- More than these are in-use – File:AI wuxia tabletop role-playing game book.jpg for example is as well regardless of how much people think is fine to discriminate against me which I'm pretty sure it a violation of WMC policy as it's currently written. It's the only for image for an example of a major application of AI art which is art for tabletop games. Also File:Boy in cyberpunk city.jpg is in use as well as is File:Vintage space opera comic cover.jpg and probably a number more. Hopefully at least WMC policy can survive the anti-AI discrimination. Prototyperspective (talk) 12:30, 8 January 2024 (UTC)
- @Dronebogus Can you explain how it is helpful to readers to have an AI generated cartoon image of events that did not happen included in an article? Counterfeit Purses (talk) 20:18, 8 January 2024 (UTC)
- To illustrate a risk or a subject of fiction to name just two. Stop assuming WMC images are just for Wikipedia and must be included somewhere to be useful; WMC isn't just for Wikipedia. Also most of these images are not about "events that did not happen". Prototyperspective (talk) 20:34, 8 January 2024 (UTC)
- My question was for Dronebogus, about some of the specific images that they listed. A Grizzly bear might somehow get into the lobby of the Empire State Building and attack tourists. Should I ask Bing to create an image to illustrate that possibility and add it to the article on Wikipedia for Grizzly bears and the Empire State Building, or just upload it here in case it is ever needed? Counterfeit Purses (talk) 21:44, 8 January 2024 (UTC)
- You’re arguing apples to oranges. You’re using a made up, non-notable nonsense example to argue against illustrating a notable purported event that didn’t happen with a hypothetical image of it happening, which nobody would object to if a human made it. Dronebogus (talk) 07:58, 9 January 2024 (UTC)
- You think my example is nonsense? One of the images you want to keep is of a fictional planet crashing into the Earth. Counterfeit Purses (talk) 16:14, 9 January 2024 (UTC)
- A notable fictional planet. Stop strawmanning my arguments. Dronebogus (talk) 13:06, 10 January 2024 (UTC)
- Which other images are there to illustrate the concept of rogue planets crashing into planets? That is or may be a notable subject of science and science fiction. Just because in your narrow preconceived notions you couldn't think of ways it could be valuable/useful doesn't mean that it isn't. Prototyperspective (talk) 14:58, 10 January 2024 (UTC)
- There's no need to be insulting. Counterfeit Purses (talk) 17:56, 10 January 2024 (UTC)
- You think my example is nonsense? One of the images you want to keep is of a fictional planet crashing into the Earth. Counterfeit Purses (talk) 16:14, 9 January 2024 (UTC)
- You’re arguing apples to oranges. You’re using a made up, non-notable nonsense example to argue against illustrating a notable purported event that didn’t happen with a hypothetical image of it happening, which nobody would object to if a human made it. Dronebogus (talk) 07:58, 9 January 2024 (UTC)
- My question was for Dronebogus, about some of the specific images that they listed. A Grizzly bear might somehow get into the lobby of the Empire State Building and attack tourists. Should I ask Bing to create an image to illustrate that possibility and add it to the article on Wikipedia for Grizzly bears and the Empire State Building, or just upload it here in case it is ever needed? Counterfeit Purses (talk) 21:44, 8 January 2024 (UTC)
- To illustrate a risk or a subject of fiction to name just two. Stop assuming WMC images are just for Wikipedia and must be included somewhere to be useful; WMC isn't just for Wikipedia. Also most of these images are not about "events that did not happen". Prototyperspective (talk) 20:34, 8 January 2024 (UTC)
: Delete Unused, out of scope personal fan art by a non-notable artist. Commons isn't here to be a personal file host, and no I'm not open to changing my vote. So it would be cool if the people who think the images should be kept skipped the argumentative, strawman nonsense. --Adamant1 (talk) 15:20, 16 January 2024 (UTC)
- You indiscriminately vote Delete either so it's impossible to convince you – you don't need to clarify that. Clear arguments for keeping have been made such as listing specific use-cases, clarifying that lots of these are in-use, and pointing out that this indiscriminate list and the vote users don't even provide explanations for deletion or make provably false statements such as pure fantasy passed off as historical reality. Truth, policies, and arguments don't matter here; what matters is the opinions of the voters by headcount it seems. Nothing of it is fan art but you still don't seem to have even read definitions of fan art. It's not a personal file host if useful images are kept and image creators don't need to be notable for useful media to be kept. Prototyperspective (talk) 15:56, 16 January 2024 (UTC)
- Stop responding to everyone you disagree with. Especially if it's me. I could really care less about your opinion. Especially when it comes to my voting habits or the arguments I make in DRs. --Adamant1 (talk) 17:43, 16 January 2024 (UTC)
- I could say the same of you. Dronebogus (talk) 18:06, 16 January 2024 (UTC)
- You could, but I think the evidence shows that's not what I'm doing though. So maybe go find someone else to make up accusations about. I don't think Prototyperspective needs you playing defense for them anyway. Thanks. --Adamant1 (talk) 18:26, 16 January 2024 (UTC)
- I could say the same of you. Dronebogus (talk) 18:06, 16 January 2024 (UTC)
- Stop responding to everyone you disagree with. Especially if it's me. I could really care less about your opinion. Especially when it comes to my voting habits or the arguments I make in DRs. --Adamant1 (talk) 17:43, 16 January 2024 (UTC)
- You indiscriminately vote Delete either so it's impossible to convince you – you don't need to clarify that. Clear arguments for keeping have been made such as listing specific use-cases, clarifying that lots of these are in-use, and pointing out that this indiscriminate list and the vote users don't even provide explanations for deletion or make provably false statements such as pure fantasy passed off as historical reality. Truth, policies, and arguments don't matter here; what matters is the opinions of the voters by headcount it seems. Nothing of it is fan art but you still don't seem to have even read definitions of fan art. It's not a personal file host if useful images are kept and image creators don't need to be notable for useful media to be kept. Prototyperspective (talk) 15:56, 16 January 2024 (UTC)
- I’ve struck this vote since Adamant1 has already voted. I want to assume good faith but at this point, between the incivility and bludgeoning and now this, Adamant1 is really skirting a trip to the user problems board. Dronebogus (talk) 18:11, 16 January 2024 (UTC)
- Sorry about that. I didn't notice I had voted already. That can happen when a DR is mostly a wall of by a single user. Regardless, I appreciate that you struck out the redudent vote. Your free to leave me a message on my talk page if I make a mistake like that in the future and I'll fix it. I'm not really sure what your talking about in regardless to the incivility, bludgeoning, or responding to everyone who disagrees with me though. Since I didn't respond to either you or Prototyperspective when either one you voted keep even though I disagree with your reasons. It's rather hard to "respond to everyone who disagrees with me" when I'm not actually responding to people I disagree with ;) --Adamant1 (talk) 18:24, 16 January 2024 (UTC)
- Comment/question: it's not that these are badly done. I'm just trying to understand why they would be in scope. They seem to me like (AI-generated) fan art. I could imagine a reason to keep a few images of AI-generated fan art as such, but normally we consider most fan art out of scope. Why, exactly, are these different? (Obviously, any that are in use should be kept.) - Jmabel ! talk 23:06, 16 January 2024 (UTC)
- The overwhelming majority are not “fan art”. I’m not sure where this notion is coming from. There are too many files to go through why each of them may or may not be in scope but since we can’t reassess deleted files I think erring in the side of “in scope” is fair in these situations. Dronebogus (talk) 04:05, 17 January 2024 (UTC)
- @Dronebogus: let me take three, more or less at random (as random as a human can be): File:AI Fighting game.jpg, File:Man licking breast in erotic comic.jpg, File:AI manga boy.jpg. Can you explain to me why any of these three would be in scope? - Jmabel ! talk 08:03, 17 January 2024 (UTC)
- There are several images, not just drawings, but erotic photos of all kinds and even lolicon (and other comics with girls in panties, clearly they are children) on Commons, I've never seen any questions about them, I don't generate art with randomness , I know that many may not even be used, but the ones already mentioned about Hercolobus and Nibiru were to illustrate entries in Portuguese, even though they are pseudoscience, I thought they needed illustrations, there are already fanarts of Marvel characters, Elvis Presley and others deleted, but there are others types of fanarts that are there, I generated some of Nikola Tesla and even orishas, I've seen several AI fanarts and gods from all types of mythology, then I saw one of Oshun and thought there could be others. If you were to use these arguments here, a lot would have to be deleted.Hyju (talk) 10:37, 17 January 2024 (UTC)
- I've seen a ton of questions about them. They often come up in DRs and a fair number are deleted. But the fact that there are other images of questionable value on Commons is neaither here nor there. My question, since I was told things were two broad, was how any of these three images are in scope. You have not answered that. And, yes, for the record, I would like to see a lot more of this sort of thing deleted. No problem with it being on the Internet somewhere else, but in Commons' terms it strikes me as sludge. - Jmabel ! talk 18:52, 17 January 2024 (UTC)
- The first image could illustrate fighting games. The second image could illustrate W:femdom or muscle fetishism. The third could illustrate manga or superheroes in anime and manga. Dronebogus (talk) 19:03, 17 January 2024 (UTC)
- I don't believe COM:SCOPE's "Must be realistically useful for an educational purpose" can be reduced to "could illustrate something." Any reasonably representational image could theoretically illustrate something, but presumably not every reasonably representational image is in scope, or we would have to accept all user-drawn art, maps of nonexistent places, porn, etc. - Jmabel ! talk 21:08, 17 January 2024 (UTC)
- This is probably too many rather different images in one DR (on that I agree with Dronebogus) and they should probably be brought back and DR'd in batches that have more in common. However, so far my inclination would be to delete any that are not in use, and that's how I'd expect to vote on the smaller batches, although I remain ready to be shown why some particular image(s) here is (are) in scope on their merits. - Jmabel ! talk 21:12, 17 January 2024 (UTC)
- @Jmabel: I've begun doing that, below. The Squirrel Conspiracy (talk) 22:02, 17 January 2024 (UTC)
- Should this DR not then be procedurally closed as redundant and confusing? Dronebogus (talk) 01:36, 18 January 2024 (UTC)
- @Jmabel: I've begun doing that, below. The Squirrel Conspiracy (talk) 22:02, 17 January 2024 (UTC)
- I've seen a ton of questions about them. They often come up in DRs and a fair number are deleted. But the fact that there are other images of questionable value on Commons is neaither here nor there. My question, since I was told things were two broad, was how any of these three images are in scope. You have not answered that. And, yes, for the record, I would like to see a lot more of this sort of thing deleted. No problem with it being on the Internet somewhere else, but in Commons' terms it strikes me as sludge. - Jmabel ! talk 18:52, 17 January 2024 (UTC)
- There are several images, not just drawings, but erotic photos of all kinds and even lolicon (and other comics with girls in panties, clearly they are children) on Commons, I've never seen any questions about them, I don't generate art with randomness , I know that many may not even be used, but the ones already mentioned about Hercolobus and Nibiru were to illustrate entries in Portuguese, even though they are pseudoscience, I thought they needed illustrations, there are already fanarts of Marvel characters, Elvis Presley and others deleted, but there are others types of fanarts that are there, I generated some of Nikola Tesla and even orishas, I've seen several AI fanarts and gods from all types of mythology, then I saw one of Oshun and thought there could be others. If you were to use these arguments here, a lot would have to be deleted.Hyju (talk) 10:37, 17 January 2024 (UTC)
- @Dronebogus: let me take three, more or less at random (as random as a human can be): File:AI Fighting game.jpg, File:Man licking breast in erotic comic.jpg, File:AI manga boy.jpg. Can you explain to me why any of these three would be in scope? - Jmabel ! talk 08:03, 17 January 2024 (UTC)
- The overwhelming majority are not “fan art”. I’m not sure where this notion is coming from. There are too many files to go through why each of them may or may not be in scope but since we can’t reassess deleted files I think erring in the side of “in scope” is fair in these situations. Dronebogus (talk) 04:05, 17 January 2024 (UTC)
Procedural close, as this has now been broken down to sets of more closely related cases. - Jmabel ! talk 22:39, 18 January 2024 (UTC)
Per the discussion above, breaking this out into a group of smaller DRs. This group is prehistoric people: Out of scope - unused and unlikely to be used - these are purportedly prehistoric African people, but are in fact fantasy illustrations in a modern aesthetic, and have no encyclopedic purpose.
The Squirrel Conspiracy (talk) 21:36, 17 January 2024 (UTC)
- I will note that File:AI African prehistoric people.jpg is in use at AI Art Application and Improvements Handbook, but I would encourage folks to read Commons:Administrators' noticeboard/User problems#Prototyperspective and "AI Art Application and Improvements Handbook" on Wikibooks for context on that. The Squirrel Conspiracy (talk) 21:38, 17 January 2024 (UTC)
- Delete Per the nominator. Clearly out of scope fan art. I'd probably not have an issue with keeping the one being used in Wikibooks, but I think it's worth deleting in this case considering the context with Prototyperspective and The Squirrel Conspiracy's point that they are ultimately fantasy illustrations in a modern aesthetic. Not actually prehistoric African people or even accurate depictions of them. --Adamant1 (talk) 21:57, 17 January 2024 (UTC)
- Delete, concur with the above. No reason to have images of how Bing "thinks" a 1940s illustrator would illustrate prehistoric Africans. Images of subjects like this based on anthropological research or nationalist sentimentality are questionable enough. This is two more removes beyond that. - Jmabel ! talk 00:22, 18 January 2024 (UTC)
- Keep No valid deletion rationale was given and nothing suggests these images are anyhow "inaccurate". In addition there's lots of other images that may be inaccurate and a template like Template:Inaccurate paleoart could be added if it was. "prehistoric people" are not out of scope and nothing requires all illustrations to not have a modern digital art look but a 70+ years outdated barely usable "aesthetic". Instead, absurd claims pro deletion like nationalist sentimentality which is fully and entirely unrelated to these images. How do you even associate that? If it's anything it would be anti-nationalist since it illustrates African prehistoric people humans worldwide share ancestry with. Reasons and arguments don't matter anymore, it's the censorship-tyranny of unjustified subjective opinions of whatever most have voted for by very small headcount. These images should be kept since they are somewhat useful and more or less the only images of ancient and archaic humans in art when not including neanderthals.
- While we keep quite tons of bizarre images as apparently educationally valuable, few of the only images depicting ancient humans apparently isn't. Explain that to somebody. Nothing in these images is very false despite that you allege they are inaccurate and they don't have to be used. Same goes for out of scope fan art – really? You still have not looked up fan art and spam deletion nominations with these provably wrong accusations? fantasy illustrations in a modern aesthetic is not a valid deletion rationale, rather it's a point pro realistic educational value. Other paintings of neanderthals or nearly anything are also fantasy illustrations and may be wrong, just not high-resolution modern digital art. Prototyperspective (talk) 10:33, 18 January 2024 (UTC)
- @Prototyperspective: You seem to miss my point. What I was saying was that we do accept historically inaccurate images based on nationalist sentimentalilty if they are, themselves, representative of what some notable artist has painted. This ranges from File:Washington Crossing the Delaware by Emanuel Leutze, MMA-NYC, 1851.jpg (surely Washington would have had the sense to sit down when crossing a near-freezing river in a small boat) to File:Horace Vernet Napoleon Rising Out of His Tomb copy.jpg (basically a fantasy image of Napoleon) or for that matter File:Tag von Potsdam. Painting by Carl Langhorst. Hitler and Reich President Hindenburg shake hands in church interior, propagandistic depiction of Hitler's 1933 appointment as Reich Chancellor. Private auction collection. No known copyright.jpg (presumably no comment necessary). But it doesn't mean that fan art or AI art equivalents are in scope. We accept the images because they are notable either as art, or as a significant part of history, not because of what they (mis-)represent.
- Similarly, insofar as Pan-Africanist or specific African nationalist images by established artists are available in PD or free-licensed versions, they are in scope. (And I agree with something you seem to be implying here, that we should have a lot more of that from an African perspective.) But things that are generated on the fly by a user or by an AI are not an acceptable substitute, any more than if I were to draw my own fantasia of, say, Rosa Luxemburg. - Jmabel ! talk 22:58, 18 January 2024 (UTC)
- This addresses one point of many I made which is one that addressed your rationale. These images are more or less the first works depicting nonneanderthal Category:Ancient and archaic humans in art on WMC and are of high-quality with no known inaccuracies.
- Even when you dismiss laboriously skillfully created AI images as "generated on the fly" that does not reduce the educational realistic usefulness of these fairly unique images. There are so few artworks and illustrations on WMC because we miss illustrators and because artists nearly never license them this way, not any of the reasons you may think of. You seem to only consider who made which artwork but that is secondary, most images on WMC are not produced by established notable people and COM:SCOPE nowhere suggests illustrations have to be either. We're overflowing with thousands of photos of the same subject but when Hyju uploads what could be the first art on Earth depicting archaic nonneanderthal archaic humans in the Commons that is out of scope fan art on a site that keeps who knows what. The only other image on WMC depicting that is the low-resolution one on the right. Prototyperspective (talk) 23:17, 18 January 2024 (UTC)
- "No known inaccuracies"? So we are to believe that prehistoric Africans had hairstyles from the 1960s (the two at left in File:AI African prehistoric tribe.jpg) or later (File:AI cavewoman.jpg) and adornments comparable to various African tribes in relatively recent centuries (all except possibly File:AI cavewoman.jpg)? "More or less the first": well, we do have File:Homo erectus pekinensis - facial reconstruction.png which appears to be based on the actual shape of a known Homo erectus skull; File:Homo antecessor statue.jpg which appears to be similarly based on actual research; File:Homo naledi facial reconstruction.jpg, File:Homo floresiensis - facial approximation - color 1.jpg and other similar images by Cicero Moraes et. al. which also appear to be rather serious efforts at facial reconstruction, etc. I will admit that I'm surprised we don't have (or I can't find) more scientifically-based images of early humans; I've certainly seen plenty of such images elsewhere, and I'd be surprised if none of them are either in the public domain or free-licensed, but I remain of the opinion that substituting fantasy images like this in contexts where such an image would be desired is the opposite of educational. - Jmabel ! talk 04:54, 19 January 2024 (UTC)
- Why would they not have these hairstyles except for the one at the left in the first pic? And even that is possible and in addition could be solved by adding a note about this there such as the mentioned template. Why would they not have these adornments? 120 k years ago they already used marine shells for personal decoration. So nothing is inaccurate except for likely the hairstyle of one person which could be fixed in a new version of the image.
- Computational facial reconstructions are not visual arts of the type paintings belong to and photos of statues and models in museum are also sth else – it's more or less the first when excluding the image above in that sense. Whatever you call this major part of art is not the issue, maybe call it creative art. Again, that it's on WMC doesn't mean that it has to be used. Hyju was helping close a major gap in WMC and even if that is only by enabling people to become aware of what is missing and providing rough outline of what could be seen. These images are very useful and clearly in scope. Prototyperspective (talk) 18:56, 19 January 2024 (UTC)
- For the most part, I've said my piece, and don't plan on replying further on this thread. However I really do object to you making substantive edits, with no markup indicating an edit, to a comment that has already been replied to. It can make the context of someone else's remarks unclear. Also, you say "Computational facial reconstructions are not visual arts of the type paintings belong to and photos of statues and models in museum are also sth else". Yes, and those things, unlike these illustrations, are clearly in scope. - Jmabel ! talk 23:07, 19 January 2024 (UTC)
- The keywords used to generate these include “ historical art comic vintage 1940s”. That’s extremely concerning - basing these on 1940s comic books is one of the most bizarre rationales for historical accuracy I’ve ever heard being used! - Chris.sherlock2 (talk) 09:06, 24 January 2024 (UTC)
- "No known inaccuracies"? So we are to believe that prehistoric Africans had hairstyles from the 1960s (the two at left in File:AI African prehistoric tribe.jpg) or later (File:AI cavewoman.jpg) and adornments comparable to various African tribes in relatively recent centuries (all except possibly File:AI cavewoman.jpg)? "More or less the first": well, we do have File:Homo erectus pekinensis - facial reconstruction.png which appears to be based on the actual shape of a known Homo erectus skull; File:Homo antecessor statue.jpg which appears to be similarly based on actual research; File:Homo naledi facial reconstruction.jpg, File:Homo floresiensis - facial approximation - color 1.jpg and other similar images by Cicero Moraes et. al. which also appear to be rather serious efforts at facial reconstruction, etc. I will admit that I'm surprised we don't have (or I can't find) more scientifically-based images of early humans; I've certainly seen plenty of such images elsewhere, and I'd be surprised if none of them are either in the public domain or free-licensed, but I remain of the opinion that substituting fantasy images like this in contexts where such an image would be desired is the opposite of educational. - Jmabel ! talk 04:54, 19 January 2024 (UTC)
- Delete. Out of scope. Post-hoc speculation about how accurate this fantasy fan art may be to actual anthropology is not persuasive. Likewise, how faithful these generated works are to some actual artistic style is similarly arbitrary. Grayfell (talk) 00:40, 20 January 2024 (UTC)
- Delete per nom and Grayfell -- Infrogmation of New Orleans (talk) 03:07, 20 January 2024 (UTC)
- Delete I find the reasoning given by the nominator persuasive and believe they are out of scope and should be deleted. - Chris.sherlock2 (talk) 08:49, 24 January 2024 (UTC)
- I also note that if we are going for historical accuracy it is somewhat concerning to see them specify “ tribal dances ancomic vintage 1940s 4-color jungle”! - Chris.sherlock2 (talk) 08:57, 24 January 2024 (UTC)
Deleted: per nomination. --Minorax«¦talk¦» 04:14, 25 January 2024 (UTC)
Per the discussion above, breaking this out into a group of smaller DRs. This group is Cyberpunk fan art: Out of scope - unused and unlikely to be used - it just seems to be a collection of images the uploader made and thought were cool. (Note, there is one file in this series that's being used in the Afrofuturism article on pt.wiki, which I did not nominate).
Extended content
|
---|
The Squirrel Conspiracy (talk) 21:49, 17 January 2024 (UTC)
- Delete, COM:OOS. I can understand the initial excitement in regards to text-to-image artificial intelligence, but now it is possible to evaluate more rationally that these images are out of scope. RodRabelo7 (talk) 23:46, 17 January 2024 (UTC)
- Delete, COM:OOS, I cannot even imagine a reason to keep any of these. - Jmabel ! talk 00:25, 18 January 2024 (UTC)
- Delete Per RodRabelo7 and Jmabel. I to can't see a reason to keep these images. It's questionable that "cyberfunk" is even artist genre and even if it was these images probably aren't the best way for us to educate people on it anyway. If they even are one to begin with, and I don't they are. --Adamant1 (talk) 01:24, 18 January 2024 (UTC)
- Delete I don't see how these could be useful; there's lots of similar/better images showing the same; these have some misgeneration issues. Prototyperspective (talk) 10:35, 18 January 2024 (UTC)
Deleted: Closing this small subset (originally nominated 27 December), seems clear that even often staunch AI advocates are not offering defense of these. --Infrogmation of New Orleans (talk) 23:50, 19 January 2024 (UTC)
Per the discussion above, breaking this out into a group of smaller DRs. This group is Western-style comic characters: Out of scope - unused and unlikely to be used - it just seems to be a collection of images the uploader made and thought were cool. (Note, there are a few this series that are currently used in the mainspaces of sister projects, which I did not nominate).
The Squirrel Conspiracy (talk) 21:56, 17 January 2024 (UTC)
I'm taking the liberty of creating a gallery here so I and others can eyeball these without needing to click into each file in this longish list. - Jmabel ! talk 00:30, 18 January 2024 (UTC)
- Delete I'll admit these at least are fun, and the style is so simple that the AI can do it absolutely convincingly, but I still fail to see how they would be in scope. - Jmabel ! talk 00:33, 18 January 2024 (UTC)
- Delete Overly simplistic and inaccurate dipections of what the AI thinks pulp art looks like. There's already many other images on here that serve the same purpose, weren't created by AI, and don't have the issues with accuracy that these ones do. So I fail to see how these serve an educational purpose. At least more then amateur depictions of the same subjects by a non-notable artist would. If people really wanted to educate people on the art style they should just do it by uploading historical magazines or actual artwork by notable artists to do it with. --Adamant1 (talk) 01:29, 18 January 2024 (UTC)
- Keep File:Nibiru entering a collision course with Earth.jpg and File:Hercolubus entering a collision course with Earth.jpg; I’ve already established that these are in scope because they have been used constructively and illustrate notable topics. File:AI Jenny Everywhere fanart.jpg is also in scope as a unique AI-generated image of w:Jenny Everywhere. I also feel File:A scientist and an android female in a laboratory looking at a laptop.jpg is in scope because it could illustrate human-robot interaction. Delete the others as having no discernible use. Dronebogus (talk) 01:30, 18 January 2024 (UTC)
- File:Nibiru entering a collision course with Earth.jpg and File:Hercolubus entering a collision course with Earth.jpg were in use when this discussion started. I have since removed them from articles on the Portuguese Wikipedia. We don't need AI-generated cartoon depictions of things that haven't or can't happen. That doesn't help readers. I also removed File:Ai-generated sword and planet hero.jpg from an article that already had 3 images from real comics. Counterfeit Purses (talk) 03:53, 18 January 2024 (UTC)
- You keep emphasizing some imaginary policy that illustrations of things that aren’t and never will be real are OOS. I guess by that logic we should delete every image of bigfoot, ghosts, superheroes, Martians, conspiracy theories, myths and legends, fiction, notable hoaxes? Dronebogus (talk) 08:35, 19 January 2024 (UTC)
- @Dronebogus: It really depends. Sometimes illustrations of "fake" things are deleted. If want a recent example there was the Flag of Paraduin, which was deleted as OOS because Paraduin is a non-exiting microstate and there's plenty of other examples. Although your correct that "illustrations of things that aren’t and never will be real" are never going to be deleted, but I suspect that's more to do with other reasons besides them being illustrations or not real. Of course anyone can pick a random subject on here and find OOS images that haven't been deleted yet related to it as if that's some kind of indicator of something, it really isn't though. Except that no one has nominated the images for deletion yet. Honestly, I'm kind of surprised you don't know that already though, what with the time you've on Wikipedia dealing with AfDs and all. --Adamant1 (talk) 18:14, 19 January 2024 (UTC)
- "illustrations of things that aren’t and never will be real" can be in scope if the artist is notable, or they formed part of a notable work, or even without those they are genuinely representative' of a cultural phenomenon of belief in or representation of that unreal thing (e.g., on that last, a conventional representation of Bigfoot/Sasquatch such as File:Sasquatch.svg is not by a notable artist, is not part of a notable work, but it an example typical of imagery with plenty of currency in the real world). But users or AI making stuff up out of nowhere? Not in scope. - Jmabel ! talk 23:15, 19 January 2024 (UTC)
- Excellently said Jmabel. -- Infrogmation of New Orleans (talk) 23:46, 19 January 2024 (UTC)
- You’re pretending like the w:Nibiru cataclysm isn’t a real conspiracy theory? Dronebogus (talk) 05:56, 21 January 2024 (UTC)
- Even if it is, Nibiru/Planet X is suppose to be 4 times the size of earth and probably mostly made of ice due to it's distance from the sun. So it's not like the illustration is at all accurate of the conspiracy theory anyway. But it does go to show exactly how untethered from any kind of educational value these images and the reason's for keeping them are. Not only is it an image of a non-exiting planet, it's not even an accurate illustration of said planet to begin with. --Adamant1 (talk) 06:16, 21 January 2024 (UTC)
- You have a point there. Dronebogus (talk) 09:57, 23 January 2024 (UTC)
- Even if it is, Nibiru/Planet X is suppose to be 4 times the size of earth and probably mostly made of ice due to it's distance from the sun. So it's not like the illustration is at all accurate of the conspiracy theory anyway. But it does go to show exactly how untethered from any kind of educational value these images and the reason's for keeping them are. Not only is it an image of a non-exiting planet, it's not even an accurate illustration of said planet to begin with. --Adamant1 (talk) 06:16, 21 January 2024 (UTC)
- You’re pretending like the w:Nibiru cataclysm isn’t a real conspiracy theory? Dronebogus (talk) 05:56, 21 January 2024 (UTC)
- Excellently said Jmabel. -- Infrogmation of New Orleans (talk) 23:46, 19 January 2024 (UTC)
- "illustrations of things that aren’t and never will be real" can be in scope if the artist is notable, or they formed part of a notable work, or even without those they are genuinely representative' of a cultural phenomenon of belief in or representation of that unreal thing (e.g., on that last, a conventional representation of Bigfoot/Sasquatch such as File:Sasquatch.svg is not by a notable artist, is not part of a notable work, but it an example typical of imagery with plenty of currency in the real world). But users or AI making stuff up out of nowhere? Not in scope. - Jmabel ! talk 23:15, 19 January 2024 (UTC)
- @Dronebogus: It really depends. Sometimes illustrations of "fake" things are deleted. If want a recent example there was the Flag of Paraduin, which was deleted as OOS because Paraduin is a non-exiting microstate and there's plenty of other examples. Although your correct that "illustrations of things that aren’t and never will be real" are never going to be deleted, but I suspect that's more to do with other reasons besides them being illustrations or not real. Of course anyone can pick a random subject on here and find OOS images that haven't been deleted yet related to it as if that's some kind of indicator of something, it really isn't though. Except that no one has nominated the images for deletion yet. Honestly, I'm kind of surprised you don't know that already though, what with the time you've on Wikipedia dealing with AfDs and all. --Adamant1 (talk) 18:14, 19 January 2024 (UTC)
- You keep emphasizing some imaginary policy that illustrations of things that aren’t and never will be real are OOS. I guess by that logic we should delete every image of bigfoot, ghosts, superheroes, Martians, conspiracy theories, myths and legends, fiction, notable hoaxes? Dronebogus (talk) 08:35, 19 January 2024 (UTC)
- File:Nibiru entering a collision course with Earth.jpg and File:Hercolubus entering a collision course with Earth.jpg were in use when this discussion started. I have since removed them from articles on the Portuguese Wikipedia. We don't need AI-generated cartoon depictions of things that haven't or can't happen. That doesn't help readers. I also removed File:Ai-generated sword and planet hero.jpg from an article that already had 3 images from real comics. Counterfeit Purses (talk) 03:53, 18 January 2024 (UTC)
- Keep The second is useful as one of the only images potentially useful for illustrating the subgenre "Dieselpunk"; the planet collision images are some of the only potentially useful for illustrating the concept of rogue planets colliding; some of the raygun images are potentially useful for illustrating at high-quality and without magazine-cover-text pulp scifi magazine tropes and so on; most of the other images there are probably not nearly as useful and I don't see a big problem with deleting those but also I don't see why that would be needed (given which images we keep, how few images those are, how unproblematic these images are since they're properly categorized and labelled as AI-made, and because there's 100 million images with lots of TBs so a few images aren't such a problem), especially since one would have a better range of options to choose from in illustrating AI images for comics. Prototyperspective (talk) 10:40, 18 January 2024 (UTC)
- Delete, except for File:AI Jenny Everywhere fanart.jpg, per Dronebogus and contextualized by w:Jenny Everywhere. Whether or not these are visually appealing to look at is besides the point. They are visually distinct from the style of art they emulate, which makes them poor for representation. To include these as a representation of an existing art style is badly misrepresenting that style. This undermines the educational value of real examples by causing confusion and muddying the waters. It's also worth noting that these kinds of images are trivially easy to generate and so they have no value from scarcity. Grayfell (talk) 00:52, 20 January 2024 (UTC)
Deleted: with exception of File:AI Jenny Everywhere fanart.jpg per nomination & discussion. --Infrogmation of New Orleans (talk) 15:59, 25 January 2024 (UTC)
Per the discussion above, breaking this out into a group of smaller DRs. This group is Japanese-style comic characters: Out of scope - unused and unlikely to be used - it just seems to be a collection of images the uploader made and thought were cool. (Note, there is one upload in this series which is used in the article for Magical girl on He.Wiki, which I did not nominate).
The Squirrel Conspiracy (talk) 22:01, 17 January 2024 (UTC)
- Delete, COM:OOS. I can understand the initial excitement in regards to text-to-image artificial intelligence, but now it is possible to evaluate more rationally that these images are out of scope. RodRabelo7 (talk) 23:46, 17 January 2024 (UTC)
- Delete, COM:OOS. A few of these are convincing fakes (e.g. File:Women kissing in comics style.jpg, File:AI - redhead anime girl.jpg) and File:AI anime girl coloring page.jpg is even something I would call "pretty good" for what it is (except maybe the over-heavy lines around the face and diadem, and the rather mediocre eyes) but, again, I don't see what makes these fall within Commons' scope. - Jmabel ! talk 00:43, 18 January 2024 (UTC)
- Delete Per other comments. I don't really see what's educational or in scope about these. Their subpar, amateurish and inaccurate depictions of the art styles at best. And AI artwork is somehow automatically in scope simply because of the art style or technology involved. So I don't really see any other reason to keep the images. --Adamant1 (talk) 01:32, 18 January 2024 (UTC)
- Keep File:AI-generated walking girl.jpg is useful for illustrating scaling up low resolution images to high resolution reimagined images. Some of the other images could be useful for illustrating AI tools usefulness for anime and comics production, File:AI vintage manga superhero.jpg is good for that, and possibly other subjects. However most are not realistically likely useful so Delete most or all of the others. Prototyperspective (talk) 10:45, 18 January 2024 (UTC)
Deleted: per nomination. --IronGargoyle (talk) 05:32, 24 January 2024 (UTC)
Per the discussion above, breaking this out into a group of smaller DRs. This group is Misc. characters in modern styles: Out of scope - unused and unlikely to be used - it just seems to be a collection of images the uploader made and thought were cool.
The Squirrel Conspiracy (talk) 22:09, 17 January 2024 (UTC)
Once again a lot of these, so I am creating a gallery to allow rapid eyeballing. - Jmabel ! talk 00:47, 18 January 2024 (UTC)
- Delete. What is the argument for these being in scope? I sure don't see it. Most of these strike me as not even being very good for what they are (AI-generated work in genre styles), but it's not clear what would be the basis to keep them even if they constituted good work in these styles. - Jmabel ! talk 00:52, 18 January 2024 (UTC)
- Delete There's zero reason these images would be in scope and any claims to the contrary are just bad faithed gaming of the guidelines. --Adamant1 (talk) 01:34, 18 January 2024 (UTC)
- Delete I don't see how these could be useful and some like the second have misgeneration issues. Thanks for creating the gallery for glancing over the images which is useful. File:AI-generated walking girl.jpg is also nominated in the DR above so it should be excluded here (and I think it's one of the ~two that should be kept there). Weak keep for File:African American cyborg.jpg (sufficient quality etc) and Keep File:Cyborg Frankenstein.jpg – a modern cyborg reimagination of / take on Frankenstein with no alternative image of that and sufficiently free of misgeneration.
- Prototyperspective (talk) 10:52, 18 January 2024 (UTC)
- Comment The "African American cyborg" looks to me influenced by 1970s Blaxploitation, while "Cyborg Frankenstein" seems specifically the Boris Karloff version of Frankenstein's Monster. -- Infrogmation of New Orleans (talk) 18:03, 18 January 2024 (UTC)
- Good point there in regards to Frankenstein image while the incorporation of a notable genre style is only a point pro keeping. For the Frankenstein image there also is File:Frankenstein's monster (Boris Karloff).jpg and both just show the head from the same angle. It's not a very valuable image and could possibly be recreated in a way that looks less like the Boris Karloff version just to be sure. Prototyperspective (talk) 22:10, 18 January 2024 (UTC)
- Comment The "African American cyborg" looks to me influenced by 1970s Blaxploitation, while "Cyborg Frankenstein" seems specifically the Boris Karloff version of Frankenstein's Monster. -- Infrogmation of New Orleans (talk) 18:03, 18 January 2024 (UTC)
- Delete all per nom & discussion. -- Infrogmation of New Orleans (talk) 18:04, 18 January 2024 (UTC)
Deleted: per nomination. --IronGargoyle (talk) 05:30, 24 January 2024 (UTC)
Missed listing, same criteria — billinghurst sDrewth 00:55, 26 January 2024 (UTC)
Deleted: per nomination. — billinghurst sDrewth 00:56, 26 January 2024 (UTC)