AI image generator - for book covers???

[.matthew.]

The problem with AI-generated fantasy art, at least the stuff I've seen, is that it all screams BLANDLY GENERIC, and that's the impression it would give me of the book (which might in many cases be accurate, of course). We haven't yet reached the point where AI can possess actual imagination or original vision.

Yet reached being the takeaway here. Most of these AI are still in their testing phases and already they're remarkably impressive.

I've also just looked through my library of books (with many chosen because of their covers) and frankly, BLANDLY GENERIC also seems to apply to most of them as well. There's only so much you can do with a cover that also requires the title and author to be clearly readable.

There are quite a few excellent covers, but for the most part, I'm not seeing much that couldn't have the background done by AI. The main issue would be insuring continuity between books, though that could be arranged by saving a bunch of similar ones at the same time.

That and perhaps characters, but I'm not sure how accurate the understanding of descriptions is yet.

 

[JunkMonkey]

No, creative artists do not owe each other anything, regardless of what discipline they have. I didn't mention respect at all... but no, you can of course respect someone you admire, and there are plenty of authors and artists and sculptors that I do, but they're not owed it by default.

They are especially not owed your money if you don't want their services. You don't have to financially support creatives just because they're creatives, which is even more true if you can get similar work done by a machine for a fraction of the price. That's like saying you only buy products made by hand because Luddites don't support mechanisation.

Yep. Why buy real food when there's MacDonalds?

As for the time commitment thing, it is still a valid point. You're implying that the skill of the artist is worth more than the skill of the author because they can charge a lot more for a lot less. Now, with economies of scale, the author will make a lot more, but only if the book sells well. Paying a lot for a cover might help drive sales up, but it might also mean the book you've just self-published could lose you money.

And yes, it's true that the reason you pay mechanics so much is that they know how to do the thing and you don't, but are you telling me that you wouldn't use some automated repair shop if such a thing was ever built? A place free from the price gouging of mechanics looking to put the screws to a clueless customer? A place that could tear down and rebuild your car faster than a mechanic could even diagnose the problem?

How about the printing press? Do you think books should be written out by hand? That used to be a job too until technology moved on and effectively ran it out of business. Was this a sad day for scribes? Certainly... but it was a good day for anyone who wanted to buy a book.

Well that's a silly argument. You're conflating the method of mass production of an artwork with the creation of the artworkin the first place. Andy Warhol might have had fun with that idea but it really doesn't stand up very well. A copyist is a copyist - no matter how skilled. Vastly different from being a creative artist.

This is the exact same thing. An incredibly similar outcome for a significantly reduced cost.

There's also nothing to stop artists from incorporating this technology into their own workflow. It could help them generate ideas and sketches for clients that would give them an edge over their competition and allow them to reduce the costs to their customers, all while maintaining a similar level of income for their time.

That said, I'd put money on the majority just using it as a means to get their work done faster without passing the savings on.

Furthermore, I'm not suggesting stealing an artist's work, merely not buying it in the same way you avoid books without good covers. Why aren't you supporting that creative author? It sounds to me like you're disregarding the discipline...

As far as books go I am a punter like any other. And as susceptible to good seductive packaging as the next guy. The author maybe a brilliant undiscovered genius but my brain does not say - 'That looks like sh*t but there's probably the work of an undiscovered genius lurking behind that crappy cover; I'll buy it.' It just goes, 'That looks like sh*t.' Life is too short. There are only so many books I can read. I have to be selective.

BTW this is the last book I had to pick up because of the cover:

Isn't it great?

But to sum up, even though future artists may not get as many customers, their potential customers shouldn't be guilted into paying for something they don't need to.

Edit: My degree is in sculpting and 3D Design, so I'm hardly unsympathetic here, but at the end of the day, the customer always has the right to seek the best deal for thesmelves.

Yep. The customer does always have the choice. Do we need more books?

I had a cartoon stuck up on my studio wall for a long time - wish I knew where it had gone. It was from the Hollywood Reporter or some-such and was entitled 'How Hollywood Works'.

First panel : Producer with an angry expression banging desk - "Do it faster! Do it cheaper!"
Second panel : Confused producer holding can of film - "Why does it look like crap?"

 

[HareBrain]

I've also just looked through my library of books (with many chosen because of their covers) and frankly, BLANDLY GENERIC also seems to apply to most of them as well.

Fair point. I'm not sure when I last bought or even looked at a book where the cover itself was a major pull. In most cases the effect is neutral, and it's the title that generates the initial curiosity. For it to be positive, it would probably have to appeal to specific niche tastes I have, such as a particular drawing style. But it's easy for a cover image to have a negative impact, such as where it suggests nothing interesting about the story. In that case I think the cover would be better off just showing an abstract design that shows, say, tension. This used to be quite common.

Isn't it great?

Yup. It's attractive, works as a whole design and does a good deal of work in suggesting the story.

 

[.matthew.]

Honestly, at this point, I've sort of got snarky. If you want the TLDR, it's in bold at the bottom

Yep. Why buy real food when there's MacDonalds?

You mean, why buy real food when there are over 38,000 McDonald's? If your argument here is that better quality invariably sells more, then that's a big ole negative. And this is coming from someone who hates McDonald's.

But let's put it in a financial context. McDonald's is known for selling a lot of food to low-wage workers, is that because they're too stupid to buy better? or is it because it's what they can afford?

Same deal. If you're a self-publishing author, you might not have hundreds to spend on hiring an artist, or the time to do the research to find a good one or go through the whole process of back and forth.

Or maybe you want to get some images ready to show a real artist what you're looking for... I don't think it's easy for an artist to decode a client's words into something they'd like without a lot of work.

Well that's a silly argument. You're conflating the method of mass production of an artwork with the creation of the artworkin the first place. Andy Warhol might have had fun with that idea but it really doesn't stand up very well. A copyist is a copyist - no matter how skilled. Vastly different from being a creative artist.

In regards to the copyist, have you seen old manuscripts? They're beautiful and creative. And yet, the ability to print the written word in mass makes them almost totally redundant, because at the end of the day, fiscal concerns matter.

Plus, as we've seen, these programs are able to do better than many artists. Not all, but a lot. I've spent a lot of time looking at art and sculpture and paintings, etc, and yea, most artists are nothing special, you could even say they're generic. Now that's not a bad thing, and they certainly have trained to draw/paint that way, but there's hardly an original thought out there.

They learn the correct technical skills and often duplicate styles they like the look of, it's copying and iteration.

This isn't much different than the various AI that can just smush and tweak and create art virtually indistinguishable to all save the professionals themselves. In today's world, this makes it a matter of mass production.

You even said yourself that when this was first seen in the places you frequent, they were wow. Only after saturation and analysis did they lose their magic, and that's a place most readers would never reach.

How about the mechanic example? That wasn't a silly argument by your own standards. A mechanic is a person with the skills to do something you can't do, yet requires no creativity, only understanding. If a robot could do that cheaper, you wouldn't go paying a mechanic to fix your car.

First panel : Producer with an angry expression banging desk - "Do it faster! Do it cheaper!"
Second panel : Confused producer holding can of film - "Why does it look like crap?"

Hollywood looks crapper than ever with the biggest budgets they've ever had.


As a final note. I'm not actually suggesting that all book covers be designed this way. I'm not suggesting that an artist wouldn't be able to make a better one. What I'm saying is that if you're self-publishing and can't afford/don't want to pay for an artist, this sort of thing could help you get a respectable cover design for very little money.

Which, by the by, would allow them to sell their books to those who do judge them by their covers

 

[Swank]

The problem with AI-generated fantasy art, at least the stuff I've seen, is that it all screams BLANDLY GENERIC, and that's the impression it would give me of the book (which might in many cases be accurate, of course). We haven't yet reached the point where AI can possess actual imagination or original vision.

No argument, but the stuff I've seen that is assembled out of clip art is so horrendous that this kind of thing is preferable.

 

[Mon0Zer0]

Plus, as we've seen, these programs are able to do better than many artists. Not all, but a lot. I've spent a lot of time looking at art and sculpture and paintings, etc, and yea, most artists are nothing special, you could even say they're generic.

Interestingly, I've been looking through Artstation and vintage book covers recently and the big takeaway for me is how much better and less generic the art is than Midjourney. Even stuff that might be thought of as generic. Take this by Bill Siencewiczczczcz (i can never spell his name right).

It is the kind of piece Midjourney would do - character in front of a circle is a design trope - but this is so much more interesting than MJ. The pose. The different techniques. The composition. The dynamism. Fantastic.

or this by Boris Vallejo:

Midjourney would struggle to produce a completely imaginary monster, even with a visual prompt. And that pose? Forget it. Unless the thing exists within its database of known objects it won't create it. And trying to create the same character in each image? currently impossible.

Now that's not a bad thing, and they certainly have trained to draw/paint that way, but there's hardly an original thought out there.

That really is nonsense. There is certainly a lot of copy-catting, but each artist brings something of themselves to the artwork. Artists like Chris Foss, John Berkey, John Harris, Frazetta, Chris Moore, Bruce Pennington, Robert McGinnis, Colin Hay, Frank Kelly Freas, Moebius, Giger, Peter Mohrbacher, Laurie Greasley, Geoff Darrow, Syd Mead, Ralph McQuarrie, Robert McCall, Roger Dean, Vincent Di Fate, Vitaly Bulgarov, Angus McKie, Alfred Kesner etc. all have distinct styles that have set the visual standard for their generations.

They learn the correct technical skills and often duplicate styles they like the look of, it's copying and iteration.

That isn't true. No one artist has the exact same style, with the exception of forgerers and animators. Even then, house styles have great variation between them.

This isn't much different than the various AI that can just smush and tweak and create art virtually indistinguishable to all save the professionals themselves. In today's world, this makes it a matter of mass production.

You even said yourself that when this was first seen in the places you frequent, they were wow. Only after saturation and analysis did they lose their magic, and that's a place most readers would never reach.

How about the mechanic example? That wasn't a silly argument by your own standards. A mechanic is a person with the skills to do something you can't do, yet requires no creativity, only understanding. If a robot could do that cheaper, you wouldn't go paying a mechanic to fix your car.

What will happen is the same as book covers now. Authors without an eye for design will produce images that seem stunning - but like when photoshop first came out, everyone's will soon look the same and then only those who pay for a decent artist will have an edge in the market.

Hollywood looks crapper than ever with the biggest budgets they've ever had.

One of the problems of filmmaking is that many directors lean so heavily on technology they've forgotten the basics - lighting, editing, composition. Watching Star Trek Picard or SW: Obi-wan for example the lighting is terrible - no keying, poor separation of characters from the background, generic design, poor framing, poor blocking. These are the consequences of relying on CGI to fix on-set errors.

However, in other respects, Hollywood has never looked better in terms of realism and effects, thanks to cgi concept art. We have so much photorealistic work out there it, it goes unnoticed now to see someone jump off a small building and land on the floor causing a crater. Everything is spectacle now.

Technology introduces trade offs: CGI looks photorealistic, but the physics are wrong; we have almost-real-life 8k 60 fps movies with incredible costumes that suffer from the soap opera effect. As close as CGI can get, it never quite captures film. Increasingly filmmakers are returning to practical effects.

Technological advances happen on an exponential curve. I suspect AI art will be the same, it made a big splash at first, we'll see incremental improvements but it will never quite get there.

 

[Laura R Hepworth]

Out of curiosity, does anyone know how likely it is for an artist to actually read the book before they do the cover? I've read so many books where the art really doesn't sell the impression I got from reading it.

Not very likely as turn over time is very important. To expect an artist or cover designer to read the book they're creating for is unrealistic and even a bit disrespectful as time is money. Unless they are charging by the hour, the more time an artist/cover designer spends working on a single commission, the less they earn to be able to pay their bills. There simply is not enough time in the world to read every book they are commissioned to work on. I am a freelance artist who does cover design, we work to the brief we're given whether it comes from a publisher or from the author themselves. The purpose of a book cover is not to perfectly represent every aspect of the book. It is to, first, grab a potential reader's attention and then to convey the type of story it is in terms of genre, sub-genre, tone, and, sometimes, even theme or tropes in order to best market the book towards its intended audience. It does not, however, have to perfectly match the way someone (reader or even the author themselves) imagines certain aspects of a the story, be it a scene from the book or even character appearance. It would even be impossible to do this as everyone's imaginations work differently. A good freelance artist or cover designer will do their best to as closely as possible match what the author/publisher wants, but within the bounds of what will also better ensure the book's success. This is where good communication is important, both at the start of a commission and during it and providing an accurate brief of what the author/publisher wants the cover to be like in terms style, theme, tone, general composition, etc and any other details they deem critical is the responsibility of the author/publisher to provide the artist or cover designer that they have hired. After the initial brief, it is then crucial for there to be good communication throughout the actual creation process to guide the art direction and address anything that comes up (and things do come up) so that the final result is the best possible one.

 

[AnRoinnUltra]

I use blender for 3d model creation. Does it have a feature that can work as book cover creation?

Not specifically as far as I know, but it's great for exact text placement and the likes -plus it's got piles of plugins and dohickeys. I've played around with overlaying images and reckon it has a vectorize function somewhere ...my thinking'd be to sketch something then work on it in the software. I only just started with creative stuff so know very little, it strikes me as a tool that could do the job though.

 

[Teresa Edgerton]

I have two points here.

Number one: It is short-sighted (and yes, most big-time publishers are short-sighted in that way) to think of a cover as a tool only for selling the book to as many readers as possible. It should be to sell the book to readers who will be inclined to love it, and to do this it does need to fairly represent the actual contents—not every tiny detail, but enough that the reader who buys the book based on the cover won't be disappointed at getting something very different from what they were led to expect. If a book sells a lot of copies but to readers who will be alienated and feeling cheated because it wasn't what they were expecting (and, at the same time, the cover discourages the very readers who would have most enjoyed it) that may be nice for the publisher who is only thinking of the bottom line this particular time, but it's a failure in terms of selling the next book in the series, or further books that the author writes. It's very, very bad for building a career.

It's one reason why with most books in a series the first book earns out (and possibly earns modest royalties for years), but there is a severe drop in sales with the next book. After all, with a series the prospective pool of buyers for book two is going to be people who bought book one. You've got a more limited pool right there. And naturally whatever the cover looked like there will nevertheless be some readers who didn't love book one, so the potential readership is going to be buyers of book one minus those who didn't finish it, finished but felt like they wasted their time in reading it, and those who didn't hate but were not terribly enthusiastic about it. And as the wrong sort of cover draws in more readers who don't like the book and fewer readers who would have liked the book very much indeed, that number may end up being very low, setting up book two for failure.

Number two: If the writer is the one paying the artist for the cover art, doesn't it seem fair that the writer, as the artist's customer, should expect to be satisfied with the result? Also, in terms of the artist's time being too valuable to read books they are doing the artwork for, it can take up a lot of everyone's time, the artist's included, if the cover art is sent back to the artist numerous times for revisions. (When Ace/Berkley published my book The Work of the Sun -- fortunately book three in a trilogy so it was in a position to do limited damage to my writing career--my editor told me that they sent the artwork back five times [I as the author had no input at all, so I don't know what the first four versions looked like], and though they were still not satisfied there was no more time left because the book was in the catalogue, had already been ordered by bookstores, and pub day was looming, so they had to go with what they had at that point. It is a truly wretched cover: ugly, unappealing, and I will never know what was in the artist's mind to include a bird sitting on the head of a bust of George Washington on the cover of my medieval fantasy novel!)

 

[Laura R Hepworth]

@Teresa Edgerton Absolutely the book cover still needs to, as closely as possible, represent the book it's for, but there are design principles and genre cues that are every bit as important also. It's about blending those together to accomplish both. This is something that relying on an AI will not achieve. Sure, it can be used as part of the process, but the raw image itself will very rarely be suited for it as is. Cover art is also very different between trad pub to indie. With indie, it is even more important for the covers to hit the right marketing cues for its audience because, yes, it's not just about selling loads of copies, but reaching the right audience for the book. And, yes, the author should be satisfied with the result. Every other artist that works with book covers and book cover designers I know, myself included, work hard to make sure the author is as happy as possible with the final cover while also doing our best to ensure it is as on market for its genre as we can achieve while honouring what the author wanted. The fact that trad pub does what it wants without input from the author is one reason I will never go trad pub as an author. (I've also heard they are a nightmare to work with as an artist.) However, this also makes the issue of trad pub authors ending up with covers they don't like partly trad pub's fault because they didn't consult them. This is why communication between the artist or cover designer and the author/publisher is so important. However, the cover is not, ultimately, for the author but for the readers and there may be some compromise in terms of design or details that is needed in order to ensure the cover does its job right. An artwork may be absolutely beautiful, hit every nail on the head for the book, and yet not be a very good book cover. Book cover art is a different type of work that combines art, graphic design, and marketing. It is also possible to be too close to one's book to be objective about the cover. I'm not saying the artist hired should just do whatever they want. Absolutely not, they need to work to the brief they're given and try to tick as many of those boxes the author or publisher have as possible while also adjusting them or applying them in the best way for the book to reach the market it is aimed at. However, sometimes the briefs that are given are not clear enough (and, when this is a the case, I strongly believe the artist/cover designer needs to go back to them and clarify it; this is also why I send screenshots to client's every step of the way so that they can request changes as I work). Also, not every artist out there is suitable for book covers. It's a type of work that requires a different eye than some artists have which makes researching artists (and even cover designers) carefully before choosing one to go with very important. Bottom line, successful cover art is a team effort between the artist or cover designer (and even typographer if this is hired out separately) and the author or publisher.

 

[Teresa Edgerton]

The thing is, I'm a hybrid author. That is, all my books came out originally from major trad publishers, and the cover art for these ranged from excellent, to beautiful but misguided, to very bad indeed. But when it came to reprints of OP books for which I asked for and received the rights back, these were either published by a very small press publisher where I had quite a bit of input (I happen to love those covers) or self-published where I worked with the artist directly (mixed experiences, because the artist had ideas of his own which were totally unrelated to the book). I have also worked as an editor, in the small press, where I was the one giving the artist a description of the book and a sense of what we wanted on the cover, and sometimes it was very, very hard to get through no matter how clearly or often I stated what was needed.

So I do have a fairly broad experience, and can say from that personal experience that it is wrong to put all the blame on traditional publishers. Sometimes an artist, having done excellent work in the past, is going through an artistic period where they are so caught up in a personal vision that it blinds them to anything else and they are not interested at all in whether the end result would sell the book, whether it represents the genre, etc. (I remember telling one artist, "This would make a lovely page in an illustrated calendar, but it doesn't look like a book cover at all.") Artists, after all can be as temperamental as writers, but it is the writer whose reputation and future is riding on the success of their collaboration, so it makes sense to me that indie writers would be drawn to the idea of an AI program that would give them more control over the final result than they may have had in the past working with artists. Especially when their choice to go indie in the first place was because they wanted more control than they would have with any publisher.

Sure if we could all afford to hire a Michael Whelan or David Cherry, I imagine we might love our cover art, have satisfying experiences working with the artists, and sell a hundred thousand books. (Two of my best book covers were done by Steve Youll, who actually phoned me with questions, bless his heart!--but I wasn't the one footing the bill in that instance.) But let us be realistic, the artists that most indie authors can afford are not always good and not always cooperative. At least with the AI program you could fuss over the art as long as it takes to get an attractive and appropriate result—asking a human artist to make that many changes would undoubtedly strain their patience beyond all bounds.

 

[Laura R Hepworth]

@Teresa Edgerton Oh, I'm not putting all the blame on trad publishers at all, but it is equally wrong to put it all on the artist as well. It can fall apart from either side (or both) which is exactly why I stressed communication between all parties involved being super important. Some artists may also be a fabulous artist, but completely wrong for doing book covers as it's a specific application of art and involves other principles and factors than what they have been trained for. This is also why some indies go to cover designers and not necessarily artists as these are actually two different things and, with cover designers, they have specifically studied book cover art/design and cover trends to be more aware of the nuances of cover art that might otherwise escape an artist used to just creating whatever they want. Although there are some cover designers that are also artists and aren't just working with stock imagery, but they have taken the time to learn the skills for both.

AI does have an appeal, but there are design issues and other factors that will require a human mind to consider and overcome. I am hugely fascinated with some of the AIs out there. Some, however, have major legal concerns affiliated with them and I cannot in good conscience recommend them. Of the AIs I've tested out and researched (which is a lot), there is only one that I have found that I have legal confidence in as regards to the issues that have been raised over them. This is another reason I urge extreme caution with adopting AI for commercial use as some of the AIs out there may be in violation of copyright laws. MidJourney is the only one I have confidence in and I do see it as being an extremely useful tool for art direction; just not, necessarily, as the final art. Many authors have tried to create their own covers using programs like DAZ with poor results because readers didn't like the artificial look of CG. It's why, even now, when a cover designer uses DAZ they will generally give it a paintover. AI art is similarly distinct from the illustrated covers done by artists or the photomanip ones by cover designers and, with the attitudes against AI art that I have seen from numerous corners, there is a very real possibility of it running into the same problem as obviously CG covers.

Believe me, I am well familiar with the concerns of cost and why this would be an even bigger appeal. I got into doing cover design because I knew I couldn't afford to hire out and, having an art degree, I was confident that I could create the cover myself. It's how I came to fall in love with digital painting and become a freelance illustrator. Some artists and even cover designers do get very costly, but there are many that are much more reasonable options for an indie, even a new one, to go to. Pre-made covers are also a very good, lower cost option for an author on a tight budget (though I fully understand how they may not be as appealing an option as a custom cover). I am also highly supportive of other authors taking the time to learn the skills required for good cover art, but, at the same time, it is a specific skill set and not everyone is going to be able to do it well.

I totally understand the uncertainty or awkward feeling of going back to an artist/cover designer for revisions as I feel the same way about approaching my formatter every time I want to make an edit to something. And, yes, revisions do need to be within reason and not endless, however, as a freelance artist, I have spent, and am willing to spend, months on a cover to get it to where it needs to be and where I know the author is happy with it. It only becomes an issue for me when the revision requests begin to go around in circles; is only down to small details that won't affect the representation of the book, but where, to make the change, might actually negatively impact the cover; or the author suddenly completely changes their mind on everything and wants to start all over with a totally different idea. I can't speak for every artist and cover designer out there and I'm sure there are some that are frustrating to work with, however, I am a part of multiple book cover designer groups and the general consensus there has been that we'd rather have the author come back to us for a request to tweak something than to go away unhappy. We just also want the author to be open to discussion as to why a change might not be a good change to make. It is also possible to overwork something and completely ruin it.

Someplace where I do see an AI being of great help is for an author taking an AI image they've made to an artist/cover designer along with their brief for them to use for art direction. It may help to increase understanding between the author and artist as to what they're after for more accurate results faster and with fewer revisions. Any extra info or visuals that an author can give their artist or cover designer at the start is always appreciated and hugely helpful. Even if it's just a crude stick figure drawing it is immensely helpful as a starting point.

Anyway, my two very long cents as someone that's been on both the author and artist end of things. You can take it or leave it as you will.

 

[.matthew.]

all have distinct styles that have set the visual standard for their generations.

That isn't true. No one artist has the exact same style, with the exception of forgerers and animators. Even then, house styles have great variation between them.

A little contradictory there. You listed a fair few excellent artists who defined a standard, then claim all these other artists have their own style. I'm not unfamiliar with the art world, and while there are plenty of 'unique' artists, there are tens if not hundreds of thousands of those who simply draw-by-numbers. I wish I could do that, so it's definitely not an insult, but there is a stupendous amount of artists who all do generic sameness.

What will happen is the same as book covers now. Authors without an eye for design will produce images that seem stunning - but like when photoshop first came out, everyone's will soon look the same and then only those who pay for a decent artist will have an edge in the market.

As I mentioned above, I'm not an uninitiated observer in this, and yes, I know not every generated image will be a winner, the point of AI is that you can hit it up for hundreds or thousands of images in quick succession, picking out and refining from the best options.

Also, virtually all book covers currently are done by artists (except the self-published homemade ones of course) and a huge number of them are absolute garbage, so hiring an artist isn't a guarantee of a good cover - leaving you out the money for nothing.

Technological advances happen on an exponential curve. I suspect AI art will be the same, it made a big splash at first, we'll see incremental improvements but it will never quite get there.

It doesn't have to quite get there. For the price, it can be good enough. That's my whole point. Of course, there are amazing cover artists, but there are also a lot of bad ones... and whichever option you pick there, it's going to cost you a lot of money or a lot of sales.

At least with the AI program you could fuss over the art as long as it takes to get an attractive and appropriate result—asking a human artist to make that many changes would undoubtedly strain their patience beyond all bounds.

Exactly this. The Midjourney one seems to let you do exactly that - plug in a bunch of parameters, then refine or generate similar from your favourites. Plus, AIs can't get angry or spiteful or take you to court for some legal disagreement.

Someplace where I do see an AI being of great help is for an author taking an AI image they've made to an artist/cover designer along with their brief for them to use for art direction. It may help to increase understanding between the author and artist as to what they're after for more accurate results faster and with fewer revisions. Any extra info or visuals that an author can give their artist or cover designer at the start is always appreciated and hugely helpful. Even if it's just a crude stick figure drawing it is immensely helpful as a starting point.

Think I mentioned that in one of the earlier comments, but it's a point worth making for sure. I don't know how much an artist would charge to do the initial sketches and stuff, but unless it's pennies, the author would save a lot of time and money by using an AI to get the concept art out of the way.

 

[JunkMonkey]

Some of the great cover art out there:

Pulp Covers

The Best Of The Worst

pulpcovers.com

The not so great cover art out there:

Good Show Sir - Only the worst Sci-fi/Fantasy book covers

Good Show Sir! Awful, terrible and simply the worst Science Fiction & Fantasy book covers for your enjoyment! Send us your terrible covers so we can document the carnage to our eyes!

www.goodshowsir.co.uk

and Sturgeon's Law:
'Ninety-percent of everything is crud.'

 

[clarissajaneen]

While AI covers may seem like an easy option, many readers will actively avoid buying books with AI cover art, whether because it signals to them that the work is amateur and/or because most AI art runs off of art stolen from real artists who have not consented to the use of their work.