We’ve already seen A.I. ‘artwork’ beginning to infest a lot of media, including physical items like print clothing, posters, mugs, and even colouring books. Take this post as an extension of one I wrote not that long ago, where I wondered about how A.I. could aid in translating books and what the potential caveats are.
Once again, I have to reiterate that I am very against artificial intelligence. I don’t engage with it, I’ve never used ChatGPT, and I wish WordPress would stop nagging me to try its A.I. writing tools for my blog posts. Actually, let me try it below. Here’s what it had to say with the prompt ‘write about how A.I. can create book cover art’:
AI presents an innovative avenue for creating book cover art through its ability to generate unique and visually captivating designs. By leveraging machine learning algorithms, AI can analyze existing cover art trends, genre characteristics, and target audience preferences to produce compelling visuals that resonate with readers. Furthermore, AI can offer personalized recommendations based on the specific themes and narratives of the book, ensuring a customized and relevant cover design. With AI's capacity for rapid iteration and exploration of diverse styles, authors and publishers can efficiently explore a wide range of options to find the perfect cover art that complements and enhances the book's content. This intersection of artistic creativity and technological advancement has the potential to revolutionize the process of crafting engaging and meaningful book covers.
There you go, the A.I. says that A.I. is a positive for this specific scenario, and it also did it in American English. Who would’ve thought?
Jokes and the scariness of how quickly that machine wrote something vaguely competent aside, what it said wasn’t at all aligned with my own concerns about A.I. and its involvement in illustrating book covers.
If you look around this website, you’ll notice I mention covers a lot. I can, and often will, judge a book by its cover. I also try to seek out specific editions or prints if I am incredibly fond of their covers, although I rarely care enough to go all-in on those pursuits. If these things make me a shallow reader, then so be it. I just enjoy the aesthetics and role of cover art.
And this is where my concern lies. A.I. is no doubt getting better with time, and is increasingly able to create more convincing images when it comes to realistic shots, alongside better grasping how to make stylised artwork and take heed of ‘mood’ and ‘tone’ prompts. Despite this, I don’t really see it becoming widely involved with the vast majority of books, so you cover artists and photographers can breathe a sigh of relief.
Effective book covers get a lot across, and A.I. can’t be trusted with the importance of that first impression we make by gazing upon the cover.
If it’s an autobiography, then it’s as simple as putting the author’s picture on the cover. If it’s a novel, then you need to effectively convey the mood through colour, details, and layout. How each publishing house does this is down to them. Some follow a format and some change it up on a book-by-book, or author-by-author, basis.
A.I. lacks the ability to understand what is pleasing to the eye. It can certainly attempt to make something it has been prompted to conclude is pleasing, but it can never actually be influenced by our criticism and emotional response to whatever it outputs.
If a publisher wants a photograph for their book cover, they’re going to licence that photograph one way or another. Equally, A.I. isn’t perfect at being consistent, so you can’t get it to create uniform cover art collections for a range of novels. Therefore, I don’t see how it’s worthwhile for publishers to try to generate photographs using artificial intelligence. It is probably more hassle than it is worth to try to prompt, then tweak, the stylised images it generates to be something consistent and pleasing to the eye. Even if it’s more expensive, it would be less bother to hire an illustrator who can follow the guidance and requests of the publisher.
There’s also huge risks involved, both legally and morally, when using A.I. images for commercial purposes. Readers aren’t going to be amused if publishers start cutting costs by generating tacky cover art. You lose something when you realise an image lacks humanity, and books are supposed to be one of the most raw and visceral ways we share our emotions, thoughts, and imagination with others.
Equally, who does the publisher credit for the art or generated image? Themselves? The A.I.? The company behind the A.I.? Everyone and everything the A.I.’s image generation algorithm was ever trained on?
That’s not to mention the likely outcome of someone suing for copyright infringement should the A.I. generate illustrations remarkably close to that of a commercial artist or organisation. That’s a slippery and unclear legal battle that could go badly over the sake of a five-by-seven-inch book cover.
For publishers, it’s too much of a mess to justify, and that’s why I think we’re safe on this front. Don’t get me wrong, publishers are a business, and almost any business will be greedy and lazy when given the chance to shortcut and sack people. Cutting costs is the fetish of every business and corporation.
But the publishing industry, especially regarding novels, is viewed as one of the few that still holds genuine passion. It has not been monopolised and can’t really be, nor are these publishers actively trying to screw one another over.
Books are written and books are published. The same works can be published by different authors, even within only a year or two of each other, not to mention publishers who are based in other countries and handling the release of a translated work. We have to assume that the time and effort given to each and every book would mean corners aren’t going to be cut on the cover art front. Some books quite literally consist of a solid colour face with some text (title and author) printed on it.
It’s more complex than you think it is. The more you ponder the issue, the more you have to go down rabbit holes of what is cover art? How many readers care about covers? How important are these things to the people involved in making them? And so on.
But, for now, I think we are fine. And the longer it stays that way, the better. The man-hours involved in the refinement and release of each book, alongside the high quality standards involved in all aspects, mean errors rarely slip through as a result. Competent and respectable publishers won’t want to place those standards in jeopardy to save a few thousand on designing a cover. However, the risk of this changing will always remain.
Leave a Reply