The Coffee Pot Book Club: Join The Coffee Pot Book Club on #WriterWednesday as cover designer Cathy Helms talks about AI in Book Cover Art #HistoricalFiction #WritingTips #WritersLife

AI in Book Cover Design?

Good thing – bad thing?

by Cathy Helms of Avalon Graphics

PART ONE

AI - The TECHNICAL BIT

Since the subject keeps popping up regularly in publishing industry circles, I'd like to offer my thoughts and experiences with AI (Artificial Intelligence) in the book cover designing area of publishing. The subject of AI often stirs strong opinions from everyone; either adamantly for or strongly against it. I do not condone the use of AI for writing novels, though, just to be clear. So, below, I share my own perspective in regards to AI in book cover design. Food for thought to those of you on the fence about AI or who might wonder how it impacts cover design.

For anyone that has been living under a rock in recent years, AI is a software program that uses a generative AI model to create images from a user's text inputs (known as prompts), usually within seconds. The results (pictures) vary in style depending on the software's capabilities, but they can typically be rendered in any style you request: photographic, illustrative, 3D, cinematic, modern, Victorian, and so on.

Initially, I was enormously excited about the potential of what I could get out of AI – how it could save me hundreds of hours editing and manipulating images to come up with finished composites for my designs. But like most everyone around the globe, I was also concerned about how and where the AI servers would data mine their sources. I certainly did not want a 'search engine' locating my digital art online and using it as a base for a new 'design' for someone else. Up until AI came along, that would have been grounds for a lawsuit against another designer for altering a commissioned design created by the original artist (no matter if it were digitally created or hand drawn). Yet, I see the many ways that AI can save me a massive amount of time by increasing efficiency in my workflow.

Scenario:

It never fails – a client chooses a royalty free stock image (or two or three) that they like overall... BUT... here is a list of edits they request, and that list is usually long.

Non-designers have always viewed Photoshop as a miracle worker of sorts:

Can't you just Photoshop that?

Yes, but it will cost you $x

Oh that's too much!

Well, it has taken X amount of training and X number of years for me to learn how to do that, so I am going to charge accordingly.

Clients commonly want extraordinary design work on the cheap. So, in comes AI to help meet those needs without exacting a ton of labor time by designers hunched over their keyboards, dressed in hoodies and sweats, alone in their dark caverns, drinking copious amounts of coffee, churning out magical Photoshop creations at all hours of the night.

Well, maybe the working conditions are not quite that extreme, but you get the idea.

How AI helps the graphic designer is by cutting down the labor time in editing all of those requested changes. Adobe has built in a great new tool with 'Generative Fill' (using Adobe Firefly AI), which allows the user to extend a background landscape, the sky, trees, and other supportive type features to an existing stock photo wherever needed. Without Generative Fill in Photoshop, one would spend hours duplicating, stamping, painting, and altering the canvas to achieve what Generative Fill does in a single click, or three.

It isn't perfect, and I still often need to patch up a few AI mistakes, but it is a massive production time saver for me. Not to mention that I am already paying to use it as a part of my subscription to Adobe Creative Suite as well. If I need to add a texture to my design, one click on 'Generative Fill' with a few keywords as a guide, I have it. Otherwise, I spend valuable time searching for texture resources online or making my own from scratch.

Adding texture, grain, dirt, etc, to enhance overall compositions is a longtime practice of most designers to give compositions depth and continuity. Most online texture resources are free for commercial use already. So, having AI generate it for me isn't a hill I'd choose to make a stand on for artists out there who already provide those resources for us to use without a commission.

The main case for and against the use of AI, as I see it through the lens of a designer, is the way we create the primary subjects in our compositions.

Do we continue to only acquire licensed stock images / illustrations and spend the majority of our time altering those stock resources to achieve our final product? Or do we use this new fairly effective and time saving tool that is AI in combination with stock resources to reach our goal faster?

The answer is yes. I fully expect many graphic designers to hold firm to the old ways and steer clear of AI... for now. Yet I personally find that using AI to generate parts and pieces that I need is a massive time saver, and so much easier than scouring stock libraries for days on end just for one small element required for the final design.

Can a designer get AI to give them an entire cover design in a single click?

No. Not without plenty of oddball flaws in the image, that is.

Can AI effectively and accurately produce typography for book covers?

Again, no. Absolutely not.

I have spent time trying to get good typography out of AI without any results that I would deem proper industry standard typography practices. And it has taken AI a while to learn to spell words correctly – glad the programmers found a way to do it recently through a programming key. But again, I have yet to see an AI engine that can properly kern (space) letters. And AI certainly does not know how to properly lay out a balanced design and prepare it for publication. So, graphic designers are certainly not out of a job. Yet.

But in the future? Who knows.

Is the use of an AI image generator legal?

Adobe says yes. Many traditional artists say no. I have mixed feelings about it myself. While I play with one of the top AI generators online, I do not use it commercially unless a client requests that I use it. I have had a couple of clients actually ask me to use AI elements already. And I will oblige them after explaining what we know, at present, about the legalities of AI generated images.

Below is how Adobe explains the legalities of their use:

Yes, Adobe does compensate photographers, specifically those who contribute to Adobe Stock, for using their images to train their AI model, Firefly; this compensation is provided through a "Firefly Contributor Bonus" which is based on the number of licenses generated from the images used in training the AI model.

Key points about Adobe's compensation for Firefly usage:

Source of images:

Adobe primarily uses images from their Adobe Stock library to train Firefly.

Contributor agreement:

Adobe Stock contributors are considered to have already agreed to terms allowing Adobe to use their content for training AI models under their existing license agreement.

Bonus system:

Adobe provides a specific "Firefly Contributor Bonus" to eligible Adobe Stock contributors whose images were used to train Firefly.

Calculation of bonus:

The bonus is calculated based on the total number of licenses generated from the contributor's images used in the training dataset.

Read Adobe's statement HERE!

A note for self-publish authors using KDP
(and possibly other publishing platforms):

Amazon has a box to tick if AI is used - and this does include the cover design. If the entire cover artwork was generated using an AI engine, then the author must tick that box. The author will not be penalized for using AI, but must declare it for tracking purposes. Here is the wording directly from KDP's website:

Artificial intelligence (AI) content (text, images, or translations)

We require you to inform us of AI-generated content (text, images, or translations) when you publish a new book or make edits to and republish an existing book through KDP. AI-generated images include cover and interior images and artwork. You are not required to disclose AI-assisted content. We distinguish between AI-generated and AI-assisted content as follows:

AI-generated: We define AI-generated content as text, images, or translations created by an AI-based tool. If you used an AI-based tool to create the actual content (whether text, images, or translations), it is considered "AI-generated," even if you applied substantial edits afterwards.
AI-assisted: If you created the content yourself, and used AI-based tools to edit, refine, error-check, or otherwise improve that content (whether text or images), then it is considered "AI-assisted" and not “AI-generated.”
Similarly, if you used an AI-based tool to brainstorm and generate ideas, but ultimately created the text or images yourself, this is also considered "AI-assisted" and not “AI-generated.” It is not necessary to inform us of the use of such tools or processes.
You are responsible for verifying that all AI-generated and/or AI-assisted content adheres to all content guidelines, including by complying with all applicable intellectual property rights.

Personally, I limit my use to Adobe's built-in fill and extension AI tools which pull from Adobe's royalty free stock library (Adobe Stock) if needed to do the job.

For instance: the contributing photographers on Adobe's site are compensated if I ask Photoshop to add a bird in my design. So, my covers do not require the author to check KDP's box to indicate an AI-generated cover design. My cover designs could contain AI-assisted elements (if I use an editing tool in the design), but I never use a fully AI generated 'image' for a cover design.

As far as the rest of the AI generators out there, some are free and some charge membership fees to use their product. If you are curious about the AI generators, here is a list containing most of them:

MidJourney (I use and pay for commercial use)

Adobe Firefly (I use and pay for commercial use)

Dall-E 2 (ImagineArt)

Stable Diffusion (OpenArt)

Generative AI by Getty Images

Coming soon:

Part Two - The Practical Bit