D&D General Would It Matter To You if D&D Books Were Illustrated by AI Instead of Humans?

[MoonSong]

As an artist, speaking from experience, there is no universal "artistic process," every artist has their own unique process.

If an individual finds inspiration from an AI, so be it, there is nothing stopping them from using that creative process to make art.

If an artist was classically trained, so be it, but just because they learned to do it a certain way does not mean that it is the best way, or the only way of doing things.

Some artists use visual aids or other methonds that some might consider unconventional. Tim Bradstreet, a veteran of the RPG and the comics industry, uses live models for all of his pieces, using the photograph, lighting, and his own artistic process to create a work of art.

Heck, your inspiration can be a stain on the wall... I misunderstood your point, when you said basis I thought you meant basis as in the first sketch that brings the layout, composition and so on. Not inspiration. Though I would question a bit if AI "art" was the only inspiration -because it seems like it would be increasingly narrow and cookie cutter-.

 

[Art Waring]

Though I would question a bit if AI "art" was the only inspiration -because it seems like it would be increasingly narrow and cookie cutter-.

I absolutely agree, "AI" art as it is at the moment can only "create" pieces that are derived from preexisting works.

 

[EzekielRaiden]

I'd love to have an AI that did the first colloring of artwork. The first layer of color is a very time consuming process and it is normally done by assistants. I don't have the budget for assistants, so having an AI do it would help me work very quickly.

Sure, that sounds like a useful tool for an artist to use. An ordinary person wouldn't get much benefit out of an AI that does the base flat colors. If the AI is designed to be flexible and to make it very easy to adjust individual colors once it generates a file, that would be even better, a tool for quickly and efficiently completing a tedious but necessary task, like a washing machine does.

The problem, of course, is that it would be expensive to make such a tool. Much more expensive than to make something like DALL-E, where it's just trained on a crap load of images and produces whole pieces at once, rather than just doing that one step. And, of course, it would be much more profitable to make DALL-E type tools and sell the tool (or art) to folks who want instant art, rather than to build this tool that only a small number of low-budget clients would be interested in buying in the first place.

Which is what makes this whole thing so frustrating. The incentives specifically lie in the area of doing crappy things to real human beings, like getting all your art from an AI for pennies on the dollar compared to what you'd get from actual artists.

 

[EzekielRaiden]

I absolutely agree, "AI" art as it is at the moment can only "create" pieces that are derived from preexisting works.

The current technology being used cannot ever truly create anything new. It is very literally only recombining elements it has been taught exist. It can combine those elements in surprising ways, but it cannot invent new elements to put into the composition. If you ask it to give you something it has truly zero training to cover, it will give you gibberish. E.g. if 2024 D&D introduces a brand new race called fnords, which have two krangles on their forehead and fluorescent purple skin, then an AI trained to make D&D character portraits will simply be incapable of producing that race.

You would need to create an AI that actually handles semantic content, not just syntactic content, for it to be able to truly invent new things.

 

[Art Waring]

The current technology being used cannot ever truly create anything new. It is very literally only recombining elements it has been taught exist. It can combine those elements in surprising ways, but it cannot invent new elements to put into the composition. If you ask it to give you something it has truly zero training to cover, it will give you gibberish. E.g. if 2024 D&D introduces a brand new race called fnords, which have two krangles on their forehead and fluorescent purple skin, then an AI trained to make D&D character portraits will simply be incapable of producing that race.

You would need to create an AI that actually handles semantic content, not just syntactic content, for it to be able to truly invent new things.

Not to mention, that the AI will constantly need to keep recombining elements from existing artists, and the more artists that leave the pool of available working artists, the less it will have to compose AI art from.

ATM it relies completely on the preexisting work of living artists. If it pushes artists out of their fields, it will look like its working, up until the rubicon is reached where "AI" can no longer bite from artists who no longer exist.

Its pretty scary for artists who rely on their work to keep food on the table.

 

[Hexmage-EN]

An online distributor called One Bookshelf that DriveThruRPG partners with has now added the following to its product standards guidelines, requiring that products with art created by AI must be labeled as such.

Third-Party Tool- and AI-Generated Images

All product listings that feature art generated by a third-party source such as Inkarnate or Dungeondraft, or an AI-generation tool such as ArtBreeder, Midjourney, etc. are required to utilize the appropriate identifying filter (found under "Creation Method" in the Format section of title filters).

Titles containing any art rendered by AI-generated tools must also display the following disclaimer in their product description:

This product contains assets that were procedurally generated with the aid of creative software(s) powered by machine learning.

Titles that do not comply are subject to removal from the marketplace. Repeat offenders may have their publishing permissions revoked.

I wonder if part of the reasoning behind this is One Bookshelf wanting to cover themselves in case of future legal developments regarding AI generated images so that they can quickly identify things sold through them using it.

 

[EzekielRaiden]

This was the most diTerlizzi-esque thing I could manage with my afternoon's experience with MidJourney:
View attachment 260098

I don't know if it filters out requests for the style of specific living artists, or if it just couldn't find enough of his art to get the style. But since I feel his planescape art has obvious Arthur Rackham influence I specified that instead, and just requested the most DiTerlizzi of all subjects, an elf girl relaxing in skimpy armor.

But fantasy artists aren't going out of business anytime soon. AI art can, after a few tries manage art of human (and human adjacent) subjects in a fantasy setting, albeit usually with some complications about items carried by them which it really can't figure out the rules for (the objects this lady had in her lap make no sense). An artist skilled enough to do touch-ups and corrections could get a lot of mileage out of it for such things. BUT even the most basic of fantasy creatures are too varied in the representations it's searching through to get satisfying output (although I think the output is very useful for brainstorming or creating rough versions to be redone by a proper artist). It just can't figure out the rules for things that only exist in people's imaginations, which is a huge part of what is needed for D&D books.

Behold, the results of my attempt to use similar means as those I used above to create a picture of a dragon flying over a fantasy city (these were some of the better results):

View attachment 260099

Yeah, this is pretty typical. Now, in fairness, the easily accessed (read: free) AIs are usually inferior models rather than the best in the business, so it can be tempting to draw strident conclusions. But it really does have some issues with a variety of stuff people would often want to see, dragons and other monstrous creatures being a primary example.

An online distributor called One Bookshelf that DriveThruRPG partners with has now added the following to its product standards guidelines, requiring that products with art created by AI must be labeled as such.



I wonder if part of the reasoning behind this is One Bookshelf wanting to cover themselves in case of future legal developments regarding AI generated images so that they can quickly identify things sold through them using it.

Almost guaranteed to be a CYOA clause, yeah. Honestly it's a smart move on their part.

 

[Art Waring]

I wonder if part of the reasoning behind this is One Bookshelf wanting to cover themselves in case of future legal developments regarding AI generated images so that they can quickly identify things sold through them using it.

That is exactly my concern, because this is newly emerging technology the laws have not had time to catch up, They are covering themselves from any future liabilities from AI art and its legality in a commercial product.

If you use AI art, the laws may change at a later time, which could potentially harm creators and their work.

 

[Umbran]

You folks are talking as if AI results are magic, and fall out of the aether for free.

There are a bunch of very talented people behind those AIs - I would expect they need to be very talented at code, as well as at art. And they've had to work very hard to get the results they do.

A person who is talented at art, using their skills to produce art... is an artist, no?

Edit: and before anyone gets at me about how their art is based on other art - I will direct you at the idea of copyright. Copyright has a limit so other folks can eventually base other art upon a given work. Licensing is similar. So, we are consuming art based on other art all the time.

 

[Sacrosanct]

Almost guaranteed to be a CYOA clause, yeah. Honestly it's a smart move on their part.

Not the way they worded it. Including things like Inkcarnate? Their phrasing implies that anything digitally assisted needs to have that disclaimer, including anything made in photoshop, InDesign, ProCreate, and others. Which is ironic, because they themselves push templates for you to use, using those tools.

They are going to end up 95% of their products with that disclaimer. Not very well thought out at all.