DJasonFleming
Movies • Books • Writing
No, AI art is not stealing food out of anybody's mouth
April 24, 2023
post photo preview
Original MidJourney creation by D. Jason Fleming, 2023, CC BY-SA 4.0

I publish public domain books. And the fact is, I put in a fair amount of work on each book. 

True, there are some people who just copy and paste books from Project Gutenberg, slap them together in "MEGApack" collections, and call themselves publishers. With a very few exceptions, that is not what I do.

I find books that are not yet on Gutenberg, and more often than not that have never been turned into properly-formed (and proofread) ebooks at all, anywhere. I take the OCR text generated from page images, clean it up, proofread it, and convert it into proper ebook format. And that's after I do the research necessary to ensure the book is in the public domain.

Some of the books I post are lucky to sell ten copies. A few books have never sold one.

Thus, not only am I a one-man operation, but I truly cannot afford to invest much (beyond my own time) into any given book. 

Up to last summer, the covers I put together were kludged together in the GIMP from public domain artwork, public domain photographs, or freely-usable art or pictures, with additional manipulation from me to get them to kinda-sorta look like the books they graced were in the genres they were in.

Then I bought a MidJourney subscription.

Even at the beginning, when the artwork was very abstract and sometimes quite horrifying, it was a superior alternative to manipulating public domain images of only vague relation to what the books contained. This was new artwork, looked, sometimes, really good --- the "horrifying" played into the Eerie October covers very well indeed --- and the more I learned to work with Midjourney, the more I can make artwork that looks distinct from any other book covers on the market.

So, to make this perfectly clear, I literally cannot afford to pay artists for new artwork for these book covers. Some books may only make ten dollars in a year. A few may not make any money. Even licensing old artwork is going to be too expensive (and too much of a pain to do, frankly, in terms of time investment finding an old piece of art, tracking down the artist or rights owner, and working out a license deal; again, I'm a one-man operation, so I cannot do that every week, for each new book I put out) for it to be practical on old public domain books.

Midjourney, on the other hand, allows me to generate literally hundreds of unique pieces of art per month, at a cost of about thirty dollars each month. 

This robs nobody. I was never going to be able to buy the services of any artist to do this task, certainly not at the scale of a new cover every week. MidJourney improves my product, and hurts nobody.

What about the artists whose styles you steal? some will ask. I have several answers to that, from the general to my own personal rules.

First, you cannot steal somebody's style. Style is not legally protectable, any more than ideas can be copyrighted. You can only legally protect specific works of art with copyright. Doing something in the style of whomever is no more plagiarism than making a gangster movie with slick camera work, long tracking shots interspersed with time-fracturing editing and a fourth-wall breaking coda is "ripping off" Martin Scorsese's Goodfellas. Derivative, sure, but the specific work is the important thing. And if you make a movie derivative of Goodfellas, nobody is going to stop considering Goodfellas a masterpiece.

Second, there are a vast number of artists who are dead, from whom Midjourney takes cues. You can't steal work from the dead, they simply cannot create new work. (As an aside, you can get some very interesting results using Vincent van Gogh as a style reference when creating science fiction art.)

Third, at my personal level, while I do use living (or recently-deceased) artists as style references, my habit is to apply their styles to genres they never have done art for.

For example, an artist named Leonid Afremov passed away in 2019. When I was still on Twitter, I followed his official account, because his artwork was unique --- he did paintings using only a palette knife, no brushwork at all. They were generally exceedingly bright and colorful, and somewhat abstract representations of urban settings.

MidJourney for a while turned out things that looked like his style, but were almost totally abstract. Now it does the "colorful", but leaves aside the palette-knife style for more realistic figures and objects. I use him as a style reference for westerns (so far, only the new cover for William MacLeod Raine's Ironheart has been published, but more are coming, as I will probably use him for covers in a long series), because he never did such a setting, and the results, when they work, are extremely pleasing.

Is this stealing from him? Of course not. He's dead, and the art generated is unlike anything he ever actually made. Is it stealing from his studio, which continues to create paintings in his style? Just as much as the studio is stealing from the original, perhaps.

As with any new tool that makes something previously difficult easy, AI art generators will be abused. That's just human nature. But is it harming anyone? Once someone starts actively forging a living artist's work, in that case, yes.

But short of that? No. It's just another tool. How the tool is used is up to each user.

community logo
Join the DJasonFleming Community
To read more articles like this, sign up and join my community today
0
What else you may like…
Posts
Articles
Well, this is going to make the con a bit "interesting"...
post photo preview
Inspiration

Here is the Table of Contents for an issue of Western Story Weekly from 1932.

Can you see what's funny and inspiring about it?

No?

The first three authors are all one author. Max Brand, Peter Henry Morland and George Owen Baxter were (just a few of the) pen names of Frederick Schiller Faust. That magnificent so-and-so was not only one of the best pulp writers, he wrote so stinking fast that he could take up more than half the issue of a weekly pulp, and do it on the regular.

post photo preview
Free Culture Art

I generate a lot of AI art for potential book covers. Much of it will never get used, so I'm sharing things here that I have no plans for, under the Creative Commons Attribution-ShareAlike 4.0 (https://creativecommons.org/licenses/by-sa/4.0/ ) International License. One a day, every day, for as long as I feel like it.

(Cross-posted to Minds (https://www.minds.com/newsfeed/1591570775834365956 ).)

post photo preview
post photo preview
Promises and Expectations
or: Steven Spielberg Ought To Know Better

Steven Spielberg's decline as a filmmaker has been sad to witness.

It was, perhaps, inevitable. When he was at the height of his powers, he got no respect from the "important" people, because he was commercially successful. I would have hoped that his success with Schindler's List might have satiated his craving for prestige and respect.

But ever since, his career has evinced a strange set of motivations: chasing even more prestige and respect from the elites (Amistad, Munich); experiments in craft that seem to have no genuine heart, a quality that his earliest films had more than almost anybody else's (AI: Artificial Intelligence, Warhorse, The Terminal); and movies he made for personal reasons (Saving Private Ryan, Catch Me If You Can, Bridge of Spies). 

While some of these movies are quite good, not one of them really measures up to what anyone in Generation X thinks of when they think "a Steven Spielberg movie".

(I know that Saving Private Ryan is beloved, but the film is a mess; to be sure, it is a mess with brilliance in it, especially the opening sequence of the Normandy landings, but still very much a mess.)

Now we have Disclosure Day (which I've not seen, and likely will not), and not only is it a dud, it is a clumsy dud. The title alone demonstrates the failure of the film. The title is the climax, and thus a "spoiler".

That's akin to making Jaws with only one shark attack, at the very end of the film.

If Mel Gibson had made The Passion of the Christ exactly as he did, but titled it The Resurrection, that would be the same sort of failure.

Or, in internet meme terms, it's the Surf Dracula problem in feature film form.

When you title your film Disclosure Day, you imply that the named day will happen in the first act, and the rest of the film will deal with the effects of it on whichever characters we follow. When you make that promise to the audience, but deliver instead a two and a half hour chase film that only ends with the disclosure, with no exploration of the consequences of it, you have cheated the audience and confessed that you no longer have the imagination that was once your stock in trade.

Spielberg used to understand how to set audience expectations, and how to deliver on the promises made to the audience. He was the definitive crowd-pleasing director. I think his taking to heart the disdain that the literati have for clarity and popularity poisoned his soul. Which is a shame.

Read full Article
post photo preview
A Gift of Fire
In which the editor geeks out a wee bit

In the late 1990s and early 2000s, I was in my 20s, and eBooks were (mostly) not a thing. I haunted book stores and used book stores, and even then, finding novels that suited my tastes was rough, as the tides in traditional publishing were already going toward exclusion and gatekeeping.

I bought anything that looked "cool", with a decided prejudice toward 1980s paperback SF art, which is how I stumbled across K.W. Jeter, among many others. But finding freedom-oriented fiction of any kind was a real fight against the odds.

Except I had a secret ally.

I have been an online creature since 1993. And somewhere around 1995 or 1996, one way or another, I stumbled across the website for the Libertarian Futurist Society, and their lists of Prometheus Award winners and nominees.

Those lists were how I was finally able to dig into Poul Anderson, when previously I had bounced off of his books even though I was sure I should have been loving them. (Well, there was also an insight into his writing style that I began to grasp when I read Harvest of Stars and "No Truce With Kings" simultaneously, but that's a different story.) They were how I discovered the works of authors like J. Neil Schulman, Victor Milán, Victor Koman, and many, many others I might not have found on my own. They were how I found Ira Levin's best novel (and, interestingly, the only one never adapted into a film) This Perfect Day, which I rate above both Orwell and Huxley in terms of totalitarian dystopias.

Sure, there are plenty of writers who have been nominated, or won, whose writing I don't care for. But over the years, the Prometheus nominee list has been a reasonably reliable guide to finding books by people who either are not insane, or are insane in fascinating ways. (And, bit of inside baseball, it has been the only science fiction award that has reliably lead to increased sales for the winners, and this has been true for well over a decade.)

One excellent writer who won, before I ever got to know her, was Sarah A. Hoyt. My love of Darkship Thieves led, through a chain of circumstances too complicated and prolonged to lay out here, to her adopting me as her little brother, and fatefully informing me that I am a good editor. (Seriously, I had no idea.) I've now been her main structural editor for eleven years, have done indie editing on and off for something like eight years, and have been Raconteur Press's lead novel editor for as long as they've been doing novels, just over a year and a half or so.

So, yeah, this is all her fault. 

Because something happened this year that I literally never imagined.

You see, the Prometheus Award Finalists for this year were announced in April. Five books made the cut, out of a total of fourteen nominations.

And I am the editor of three (!!!) of them.

I do not know if this has ever happened before. I suspect it has, simply because of the relatively narrow list of publishers willing to take on liberty-oriented fiction for most of the lifetime of the award. But I am gobsmacked that the first year anything I am associated with got nominated for anything, three (!!!) of "my" books all got nominated for the only award I care about.

I am giddy. 

Read full Article
post photo preview
Another example of why Hollywood is dying

Several years ago, Shout! Factory acquired the rights to all of Roger Corman's catalog, including his New World Pictures library.

Shout! Studios has made use of this to remake... Deathstalker.

Before I even get to my rant, problem number one is that the actor playing Deathstalker is sixty years old, and looks at least five years older. Sure, he's in great shape, and a skilled martial artist. But in no way is he believable as a warrior-barbarian, and he's a terrible choice for a lead in what they probably intend as a series, because he's only going to get older. (Unless they follow Deathstalker tradition and recast the lead role in every sequel.)

Deathstalker has no deep lore, no canon of stories or novels to adapt, just four cheesy original films which had little or no continuity, all shot in South America on the cheap to make cash from the coattails of Conan.

The only really appealing things in the original films are a plethora of unclad or scantily-clad ladies, and a certain doofy charm. This trailer lacks either. (Given the filmography of the director, the film itself may be a different matter.)

But getting a trash knockoff IP wrong in a remake isn't what set me off about this trailer.

It's the fact that they are remaking a trash IP, instead of adapting actual sword and sorcery literature. There is a lot of it, and some very good stuff is in the public domain. Even considering what's still in copyright, getting the film rights for quite a few good books or stories would not be expensive.

Gardner Fox wrote some highly entertaining Conan pastiches in his Kothar the Barbarian series, which were paperback originals, are in ebook now, and since Fox is long dead and not at all a household name, it seems likely that buying the film rights wouldn't be terribly expensive.

Henry Kuttner's excellent Elak of Atlantis stories are all in the public domain, so they would cost nothing in terms of rights. 

C. L. Moore's Jirel of Joiry is a female sword and sorcery protagonist, and all of those stories are public domain. 

The late, lamented Tanith Lee began her career with the astonishing s&s novel The Birthgrave, and it or her other early s&s books could probably be bought for a small (by Hollywood standards) price. 

Should I mention Fritz Leiber's Fafrhd and the Gray Mouser? Never adapted to my knowledge, some of the early stories are public domain, and the ones that are not should not be expensive to acquire.

And that's what set me off. There is an enormous amount of quality material that Hollywood could be adapting. Not just in sword & sorcery, but in science fiction, fantasy, and genre fiction in general. So many masters have had little or nothing adapted, the mines are rich in ore.

But no. If it's not Philip K. Dick, Frank Herbert, or something that's been remade and ripped off to death, Hollywood won't touch it. They either want to put their own stamp on things, twisting them all out of recognizeability, or just rehash a rehash that was rehashed a dozen times already.

 

Read full Article
See More
Available on mobile and TV devices
google store google store app store app store
google store google store app tv store app tv store amazon store amazon store roku store roku store
Powered by Locals