There seems to be more and more activity developing in the use of Artificial Intelligence (AI) in the creation of art. For example, just a few weeks ago Microsoft announced a new tool called Microsoft Designer, powered by AI technology, including DALL∙E 2 by OpenAI, to “instantly generate a variety of designs with minimal effort”.
Even comedian John Oliver has been talking about this new innovation on ‘Last Week Tonight with John Oliver’.
In an attempt to ease the concerns of creators, Shutterstock announced a “plan to launch its AI-generated content capabilities in a manner that is responsible and transparent for its customers and contributors…. launching a fund to compensate artists for their contributions, and focusing its R&D machine on gathering and publishing insights related to AI-generated content.”
Canadian comic book creators have weighed in on the topic of AI as a tool for the purpose of making illustrations or the writing component that are the fundamental building blocks of a comic book.
“My sense as an observer is that the product of AI art is consistently terrible, but that’s not disqualifying. People have always enjoyed terrible art. What does disqualify AI art as a practice, for me, is that it’s inherently exploitative. It’s built on the labour of artists without their permission and the people using it make no effort to compensate those artists. If there were a way for artists to opt in to having their art used and to be appropriately rewarded, great! Make all the terrible art you want to make!”
– Andrew Wheeler writer of Sins of the Black Flamingo
“I think it’s really cool that these image generators are able to do what they do! I certainly don’t think they’ll be able to replace an illustrator enough to establish a look for a project or tell stories in layouts, or develop completely new visual ideas any time soon, but who knows what the future will bring?”
– Danny Gorny writer of Sleepwalkers
“I’m sure it could probably be used to create a simple webcomic where the art isn’t the main focus. For something more complex, I would think it would really struggle to be able to choose cinematic camera views, while also making sure everyone has the right expressions and clothing, AND be consistent. So yeah, I think this COULD possibly be used for single panel comics maybe? But more than that… we’re probably not there yet. It’s not like we don’t already have tools to help up, like 3D objects we can use to help us.”
– Gisèle Lagacé creator of Ménage à 3
“I see it as a great tool for concept development, but don’t see it as actual original art, as it’s not original. It’s a composite of whatever image database is used. I know many voiced their fears of it taking away jobs and I’m sure there will be some experimentation from companies to do just that, but only to see that their desired end result will not be realized.
We’re talking about people communicating the way we do with each other. It’s not just the words, but the layered meanings behind them that only people can truly understand. Yes, there are people writing prompts with artist styles and detailed instructions, but what happens when you have a client that asks it to make it more ‘cooler’ or ‘trendy’?
People on the other hand would not only know how to work with it, if they’ve worked with the same client before, know what they are referring to without the need for the client to write an essay to explain what they mean; sometimes, they don’t know what it is they’re looking for, but know what it is after they see it. This is part of the process in creative development where the communication between people erupts into a creation of comics with immersive stories and characters. We exchange ideas and add on to each other’s, building and refining a story until it becomes something special.”
– Howard Wong writer of Damned, Cursed Children
“This discussion around AI and creation of comic book art reminds me of previous discussions around how computer technology and automation would replace workers of all stripes. And workers in many cases weren’t replaced, but evolved into using those ‘automated’ tools to do their jobs. As an example, I’m sure there were concerns about Photoshop software when it was originally created and today it’s an indispensable tool many artists use. What I’m saying is…you still need an artist (or artistic skill) to use the tool and it isn’t replacing anyone.
So I don’t think image creation AI will replace artists in any shape or form. It will be a tool artists can use to communicate ideas with others (and vice versa) or help them with their own artistic creations but it won’t replace them. Will writers drop artists in lieu of using image creation AI to tell their comic book story? They can try…but I can’t imagine it would be any good if the writer isn’t an artist as well, to properly tweak as needed. And it wouldn’t be a collaborative effort…isn’t that why you do comics? That’s why you should be creating comics, in my mind. The best stories are a collaboration between writer and artist(s)…not a writer asking the artist to robotically follow what is asked of them in the story. Write in another medium if you don’t want to collaborate with a team of creators…that’s always been my comics credo. So for image creation AI, I see it as a tool that could be used for good by writers and artists collaboratively and anyone who tries to use it to remove artists from the equation will fail miserably.”
– Jack Briglio writer of Dominion Jack
“I’m concerned that these programs are “trained” by mining the work of millions of artists online. Artists who didn’t consent to having their work used this way, and aren’t getting any more for the art that’s being based off their work. Secondly, I’m concerned because forging a career as an illustrator takes years of training, honing your craft, and often personal sacrifice. And now the advent of AI art threatens to make that training and dedication moot. When a program can do the exact same sort of work you do in mere seconds, why would one bother putting in the work of learning to do it yourself? And finally, I’m concerned because I worry that the commercial usage of AI generated art feels inevitable. Like how do you avoid it? I know there’s some opposition to it now, and I don’t think that’ll entirely go away. And prestigious publication like, say, the New Yorker, aren’t going to be replacing their illustrators with robots any time soon, if ever. But there must be a lot of small-time art directors out there who are chomping at the bit at the prospect of being able to receive high quality illustration work in seconds. Without even needing to pay an illustrator! How does that sort of technology not get used? And once it starts getting used by smaller publications, well, the precedent starts creeping in….
I will say that I think this sort of technology presents a threat to commercial illustration more than it does to comics. These AI art programs can generate a series of artistic interpretations of a single prompt, but that’s the full extent. Comic illustration requires not just interpretation of a prompt, but an entire story’s worth of prompts. Every panel is a different prompt. And then the comic artist needs to put those prompts together in such a way that they read well and feel cohesive. The amount of artistic decisions that need to be made over the course of a graphic novel goes far beyond what any of these programs are capable of at present.
The only problem being, that these two artistic mediums don’t exist entirely separately from one another. A lot of comic artists get their start in commercial illustration, while others will use commercial illustration work to supplement their income while they work on their comics projects. So what happens when commercial illustration becomes a less financially viable path? Moreso, I worry that AI art could escalate a trend of devaluing artistic labor in general. Will the availability of good AI-generated art mean that illustrators are paid less? And if illustrators are paid less, will comic artists be paid less? If people start forgetting the amount of labour that creating a piece of art takes, that hurts all artists long term.”
– Josh Rosen, illustrator of The Good Fight
“There are many unknowns with the future of AI-generated illustration right now and there’s a general sense of panic about the potential futures from artists in the comics community today. However, what I can say is that unionisation is, for sure, a failsafe to protect against the eradication of artist; just look at camera operators in Hollywood and TV, robot cameras exist – that’s a reality today, but the reason they’ve not replaced human operators en masse across both aforementioned industries is because it’s a given that every class of on-set crew are unionised and the collective bargaining power of a union as broad as IATSE – as one example – is something that the studios don’t want to displease.”
– Mark Allard-Will writer of Siegfried: Dragon Slayer
“As far as I’m concerned, AI art has no place in comics. Comic books are a legacy medium, built on the hard work of hundreds and thousands of talented artists who worked hard and practiced and strove to improve themselves over years and decades. While changes in how some of that art is created have happened, such as digital art created on tablets and not on paper, it’s still hand-drawn. Using AI art in comics is a slap in the face to everyone who has created comic books over the years, and in my opinion, AI generated art should be boycotted from the comic book industry.”
– Mark Bertolini writer of Strange Case
“I just find the whole thing very upsetting. Specifically the way that the AI has been trained by being fed the art of current working artists. It’s one thing to teach it to mimic Monet or Warhol, but it’s something else entirely to teach it to impersonate artists who are very much still alive and working and dependent upon their art to support themselves. I think the developers of this AI were so excited by the fact they could do it that ethical questions of whether they should do it were ignored. I’m sure the ethical questions will be addressed eventually (it feels inevitable that there will be lawsuits, and then legislation will follow). I’m also sure that the software isn’t going anywhere, and will only become more robust. Soon enough it’ll probably just become another tool in a working artist’s toolkit. As an artist and someone who makes comics, I can see AI being quite useful. I know quite a few artists who’d be thrilled to let the AI handle adding things like backgrounds to panels, or even adding color to black and white line art.”
– Nathan Fairbairn creator of PAWS
“I’m not totally opposed to the use of AI as a tool. I can see its potential for generating ideas and helping break artists out of the go-to compositions they may fall back on too easily. Great for those times when your publisher asks for four cover concepts, then chooses none of them. As long as you’re using it as inspiration for your own future creations, rather than passing off algorithm-generated averages of other artists’ work as your own, I don’t see a problem.
When it comes to comics specifically, I know people are already experimenting with AI to generate create graphic novels at record speed. I haven’t read any of those, but I suspect they’ll be terrible. Anyone who knows anything about comics knows that they aren’t just a series of unrelated illustrations placed side-by-side. Inter-panel composition is as important — or even *more* important — than intra-panel composition. How do you prompt the software to lead the reader’s eye subtly toward the next panel? Or cue it for a suspenseful page turn? How do you tell it to constrain characters in their individual panels to suggest their sense of isolation? Or to convey the passage of time? Or any other of the thousand other detailed and thoughtful decisions that storytellers need to make on every page? Cartoonists are writers, not concept artists. AI might be able to quickly generate images, but I defy it to generate *meaning*. You need human beings for that.”
– Scott Chantler creator of Bix
“I think AI generated art is a logical step in the tech landscape as well as the artistic front, and one I hope most people won’t write off. It’s important to separate it into its own category – It’s essentially math-art. Pieces composed of rules and pre-existing concepts. A ‘what if?’ machine. It also pushes us further into conversations about sentience in AI, which we’ll be hearing more about in the next few years.
I might not have any desire to pursue AI generated art, but I do think it’s useful as a tool for visualizing concepts. If anything, maybe it’s best treated as an artistic aid instead of an artistic medium. “
– Shawn Daley creator of Samurai Grandpa
“I’m still on the fence on this. It actually might be an interesting tool, I would often look at stuff for inspiration, google images, other artists, why not have this weird voice in my inspiration bag. BUT, It might also steal from me and possibly make my line of work irrelevant – not yet though, it sucks at storytelling and art itself is a sharing of personal human experience not just cool imagery. It’s a fascinating time, forcing on us some deep questioning on creativity. I’ll have to spend more time with it to see if Ai is a friend or foe. “
– Yanick Paquette illustrator for The Incal: Psychoverse
An informal poll suggested that 88 percent of respondents would not buy a comic book or graphic novel created by AI. However, a telling comment posed the question as to whether the reader would recognize if AI was involved in anyway in the creation of the book while another indicated a curiosity ‘just to check it out’.
The subject of AI in the creation of comic books continues to generate many feelings and concerns. It’s apparent that the introduction of this method of creation and its impact on the industry has yet to be felt by all. Some creators are reluctant about making a statement until more time has passed and it’s obvious we’ll be debating this issue for many months or years to come. Nevertheless, until someone ‘throws a spanner in the works’, the use of AI in the creation of comic books will be with us for the foreseeable future.