Category talk:AI-generated images

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Where in Wikimedia projects should these be used?

[edit]

Does there really exist a use case for this kind of image? I believe that an encyclopedia does not need imaginings of how historical figures or events might have appeared, but perhaps somewhere in Wikimedia these can find use. I remain skeptical of the utility, however. —Anon423 (talk) 19:54, 30 January 2023 (UTC)[reply]

  • I think there are many. The most notable one is as example of AI art in articles about AI art or relating to it. Other examples where such could be useful are articles about art styles and art genres.
  • "imaginings of how historical figures or events might have appeared" currently are not viable use-cases and I don't think anybody suggested that they are. There is no need to debate whether there are or could be some exceptions to this because currently nobody created such images to upload them here and to add them to articles, if people do I'd suggest to discuss it there and at a minimum make it very clear in the image description that the image is AI art.
  • Moreover, if you don't like the use of an AI art image somewhere it could be replaced with something better if it exists and is in the commons.
Prototyperspective (talk) 21:14, 30 January 2023 (UTC)[reply]

License tags

[edit]

Shouldn't all images here have {{PD-algorithm}}? I see a lot of cc-licensed images although that shouldn't be possible in most jurisdictions. MB-one (talk) 12:20, 21 September 2023 (UTC)[reply]

Template:Prompt now available

[edit]

Use it to record what prompt was used to create an AI work. Cheers, {{u|Sdkb}}talk 07:37, 21 December 2023 (UTC)[reply]

Maybe a script could move the prompts to this new field. However, it implies that only one prompt is used. That may be the case for most images here that include prompts but that's not necessarily the case. One can use ten different prompts for one image such as making one then using the image in img2img with a new prompt repeat x times. Just sth to keep in mind. Maybe that field could be automatically be added if Template:PD-algorithm is used or somehow via the Upload Wizard. Thanks for creating the template. Prototyperspective (talk) 11:16, 21 December 2023 (UTC)[reply]
Yeah, this is definitely something that I anticipate we'll need to build out/adjust over time. For now, I just wanted to get the ball rolling. The overall view is that it's better for us to have prompts recorded in a more structured way (maybe we could also have a field in StructuredData?), rather than just put into the description text. {{u|Sdkb}}talk 16:12, 21 December 2023 (UTC)[reply]
Good! ^_^ JPxG (talk) 22:12, 26 December 2023 (UTC)[reply]