Author: Joe Vasicek

Joe Vasicek is the author of more than twenty science fiction books, including the Star Wanderers and Sons of the Starfarers series. As a young man, he studied Arabic and traveled across the Middle East and the Caucasus. He claims Utah as his home.

In defense of AI art & AI writing

If Andrew Tate wrote a book about how to make your wife or girlfriend into your slave, would he be within his rights to demand that no woman reads that book without his consent?

Brandon Sanderson was inspired to become a fantasy writer when, as a child, he read Dragonsbane by Barbara Hambly. Sanderson is now worth some seven or eight figures, while Hambly, who is still alive and still writing, struggles to pay her bills*. Should Hambly be entitled to a portion of Sanderson’s earnings, for inspiring him to become a fantasy writer?

Every mother who has ever lived gives tremendously of herself to her children, even if only in the physical act of giving birth. Should mothers have a legal claim on their children, for monetary compensation for all of the sacrifices they make?

These might seem like crazy questions, but when you consider them in the context of the ethical arguments about AI art and AI writing, they really aren’t. They illustrate just a few of the unintended consequences of the regime that many disgruntled and resentful creators are arguing for, when really what they want is a world in which AI doesn’t exist.

One of the most difficult parts of being a creator is putting your work out into the world and letting it go. At that point, you really have little control over what it does and how it impacts the world. Many artists who labor in obscurity dream of making an impact on the world, not realizing that success—even artistic success—can be far more devastating and traumatic than obscurity. After all, just ask Rachel Zegler about that now.

I’m not saying that artists shouldn’t be paid for their work. Certainly they should be paid—and certainly there are valid ethical concerns with how AI is disrupting art and literature. But unhinged people who rant online about how AI is “stealing” artists’ work, or how it is “plagiarizing” writers’ books, simply because the LLM’s training data includes free online content (much of which was posted online by said artists and writers)—I don’t think those people really care about the ethical nuances of the debate. I think they just want to force us all to go back to a world where generative AI doesn’t exist.

Did David Weber steal from Star Trek when he wrote the first Honorverse novel? Did John Scalzi steal from Robert A. Heinlein and Joe Haldeman when he wrote Old Man’s War? Did Terry Brooks steal from Tolkien? How about George R.R. Martin?

Where exactly is the line between the “stealing” that should get you thrown in prison, and the “stealing” that people wink and nod at when they say that good artists copy and great artists steal? And how do we know that we’ve drawn the line in the right place? Would we have worse art, or better art if Star Wars had gone into the public domain in the 80s or 90s? Would artists be making less money, or more?

I don’t have the answers to these questions, but I ask them because I think they are worth considering. And I think that most of the artists who think they have the answers are really just acting out of fear.

Will AI outright replace artists and writers? Will it make it impossible for artists and writers to make a living? I remain skeptical, though I acknowledge that there are some ways in which AI art appears to be doing exactly that. For example, I’ve been playing around with OpenAI’s new image generator, making some cover mock-ups, and I’ve been very impressed. But I will still seek out James at GoOnWrite.com for my covers, because he has a much better eye for this sort of thing, and my sales data reflects that his covers sell more of my books than my own covers do.

Should writers and artists expect to be paid whenever their art is used to train an LLM? Aside from the impracticality of enforcing such a law, I don’t think that we should—at least, not for general training data. Fine tuning is a different matter. If an AI is going to be fine-tuned to write in my particular style, I think I have a right to be recompensed for that—and I’d be willing to license that right for a reasonable fee. Perhaps this is a path that artists could pursue as well. But demanding that every AI company pay every artist for training their LLMs is kind of like Barbara Hambly demanding that Brandon Sanderson pay her a portion of his earnings. Likewise, whenever artists or writers demand that their intellectual property is excluded from the training data, it smacks to me of the first question with Andrew Tate and his hypothetical book.

I will admit that I’m biased in favor of AI, since for the last two years I’ve been working to incorporate it into my own creative process. But I’ve been doing this out of a recognition that these things we call “writing” or “making art” is going to change because of these new technologies. In a world saturated with AI, will it still be possible to make a living as an artist or a writer? Yes, I believe it will, but at the same time, I believe that our conception of what it means to be an “artist” or a “writer” will almost certainly change. That’s why I’ve chosen to embrace these tools, rather than fight them—and why I think my fellow artists and writers should as well.

*At CONduit 2010 in Salt Lake City, Barbara Hambly was the guest of honor, and in her keynote address she talked about her struggles to pay her bills with writing. I assume that things haven’t changed much in the years since then, though I would be delighted to learn that I’m wrong.

Five things I did at work last week

  • I passed 10% of the revised AI draft for The Soulbond and the Sling (about 16k words).
  • I started working on Bloodfire Legacy again, making some progress on the human draft.
  • I made the covers for Thanks for Reading! my forthcoming compilation of author’s notes.
  • I wrote the book description for Thanks for Reading!
  • I ran a promotion for Genesis Earth.

New Blog Series: Fantasy from A to Z

Several years ago, back when “the blogosphere” was still a thing, I participated in a 30-day blogging challenge, where I went through the alphabet talking about various aspects of science fiction. Not only was it a lot of fun, but I also made a book out of it, which I later revised and made into a reader magnet exclusively for my newsletter subscribers.

Well, it’s been a few years, and I’ve since decided that instead of being a science fiction writer who occasionally writes fantasy, I want to be a fantasy writer who occasionally writes science fiction. But in order to make that pivot, I need to change my newsletter magnet into something that isn’t explicitly about science fiction.

So what I plan to do is write a series of blog posts exploring the fantasy genre from A to Z. I probably won’t do it in less than thirty days, because 1) life with small children is pretty crazy, and 2) I want to put some serious thought into it. But I will probably have them all up by the summer, at which point I’ll turn it into an ebook and make it my new reader magnet for the newsletter, replacing Science Fiction from A to Z.

So that’s the plan. Are there any particular letters you’d like to see? I need to brainstorm ideas for the tropes and aspects of fantasy that I’d like to cover. I have a few ideas already, like “M is for Magic” and “E is for Epic,” but I’m open to more.

Just for fun, when I asked the wordpress AI assistant to come up with a featured image for this post, this is what it came up with:

Not too bad, though if you look closely enough, it’s obviously AI art. But it should be fun to come up with a cover image for this one. I’ll do that first.

Should I split my epic fantasy series into two trilogies?

So I’m working on the first book in a new epic fantasy series, called the Soulbound King. It’s basically a fantasy retelling of the life of King David, loosely adapted from the biblical stories about his life. I’ve already outlined the first book and generated a rough AI draft, which came in at 153k words. The final draft will likely be longer than that, but I think it’s very likely that I will be ready to publish it before the end of the year.

The question I’m currently grappling with is whether to keep it as a seven book series, or to release it as two trilogies with a bridge novel in the middle. Frank Herbert did a similar thing with his Dune books: the first three books (Dune, Dune Messiah, and Children of Dune) were a trilogy, and the next book, God Emperor of Dune, was supposed to be a bridge novel setting up the second trilogy—except he died before finishing the last book, so his son Brian Herbert got together with Kevin J. Anderson to write it, and then they blew it up into a franchise… point being, stuff like this has been done before.

Now, I’m reasonably confident that I’m not going to die before finishing the last book. In fact, I’ve already made a 7-point outline for all seven books, so I know exactly where they start and end, with the inciting incident, midpoint, climax, etc. I’m also writing these books with AI assistance, which is making it possible for me to write these books much faster than I otherwise would have been able to write them. For the first book, The Soulbond and the Sling, I anticipate that it will only take between six to nine months of total work to go from story idea to finished draft.

But the trouble with writing a seven book epic fantasy series is that a lot of readers aren’t going to bother picking it up until all seven books are out. This is because so many readers have been burned by authors like George R.R. Martin and Patrick Rothfuss, who have not and likely will never finish their bestselling series. I can’t really blame the readers for that (though I can and do blame the authors), but it creates a market reality that I need to anticipate and plan for.

So here’s what I’m thinking: instead of making it a seven book series, I’ll make it two trilogies with a bridge novel in-between. The first three books will complete one arc, and the last three books will complete another arc. I’ll wait to release the first book until after I’ve completed the AI draft of the third book, so that way I can release all of the books in the first trilogy within 1-3 months of each other. And after the first trilogy is complete, I’ll market it as a trilogy while working on the last four books, probably releasing each of those a year apart, as I finish them.

The reason I’m thinking about this now is because a strategy like this is going to influence how I write all of these books. If I’m going to split the series into two trilogies, the last thing I want to do is end the first trilogy on a cliffhangar. It has to hold together as a complete story, with only one or two loose threads. But since I’m still in the early writing stages of the first book, I still have enough room creatively to make that kind of adjustment. I just have to decide if that’s truly the plan.

By the way, the first trilogy ends with the fantasy equivalent of the Battle of Mount Gilboa, where the Saul and Jonathan characters die in an epic battle and the David character becomes king (I know that in the Bible, there was a gap of several years between those two events, but I’m combining them for purposes of this book). So it is a rather natural stopping place, even if it does end on a massive downer, followed by a false victory (the second trilogy begins with David and Bathsheba).

Anyways, what do you think of this plan? Does it sound like a good idea, or is there a compelling reason I haven’t thought of yet for why I shouldn’t do it?

Five things I did at work last week

This post is late because I was too busy yesterday with writing. The Soulbond and the Sling is coming along slower than I would like, but I’m making steady progress, and really excited about this project. I suspect it will end up being the longest book I have ever written.

Last week, I:

  • Finalized the text for my next book release.
  • Finalized the book description for The Soulbond and the Sling.
  • Added some key details to the book’s magic system, clarifying the costs and adding some important visual elements.
  • Wrote and scheduled an email newsletter.
  • Finished the AI draft of the prologue, at 3,077 words.

Feel free to add your own!

A fascinating take on Brandon Sanderson’s Winds of Truth

Really interesting take on the publishing industry. This is politically incorrect, but I think it’s related to the way that publishing has become dominated by liberal women. The Nonsense-Free Editor has a lot of great videos about that.

There are lots of great books in the indie publishing scene that are not “Potempkin villages,” as this BookTuber calls them… but how is one to find them? When we finally solve that question, there will be a resurgence of great writing and great art.

How I Would Vote Now: 2008 Hugo Awards (Best Novel)

The Nominees

The Yiddish Policemen’s Union by Michael Chabon

Brasyl by Ian McDonald

Rollback by Robert J. Sawyer

The Last Colony by John Scalzi

Halting State by Charles Stross

The Actual Results

  1. The Yiddish Policemen’s Union by Michael Chabon
  2. The Last Colony by John Scalzi
  3. Halting State by Charles Stross
  4. Rollback by Robert J. Sawyer
  5. Brasyl by Ian McDonald

How I Would Vote Now

  1. No Award
  2. Rollback by Robert J. Sawyer

Explanation

Two thousand eight was the year that I took Brandon Sanderson’s writing class and decided to become a professional fiction author. It was also the year that I discovered David Gemmell and Robert Charles Wilson, two of my favorite authors. It was also the year that the world economy collapsed and Obama won the US election, so it was a very eventful year.

Unfortunately, it was not a very good year for science fiction & fantasy—or at least, not for the Hugo Awards. I didn’t read any of these books until just recently, but I have to admit, I didn’t like any of them.

Perhaps, if I were a liberal atheist Jew, I would have enjoyed The Yiddish Policeman’s Union (not a conservative orthodox Jew, mind you—if anything, I probably would have hated it more). It’s not that I have anything personal against Jews or Jewish culture. I quite enjoyed Fiddler on the Roof. But unlike Fiddler on the Roof, The Yiddish Policeman’s Union doesn’t have many points of commonality with the wider culture to make it accessible.

For example, even though the traditions of Anatevka are likely different from the traditions of whatever culture we call our own, most of us understand the concept of tradition as a governing force in our lives, and can therefore sing along with the song “Tradition” and understand how it affects the story. But the plot and worldbuilding points of The Yiddish Policeman’s Union were so quirky and uniquely Yiddish that I just found it difficult to connect with or even follow them all.

Honestly, it would be a bit like if I were to write an alternate history where the Mormons were driven from the continental US after the Utah Wars, and settled in Hawaii and the Polynesian islands, except the main character drinks Coca Cola and has a complicated relationship with his polygamous step-mother (because plural marriage was never renounced in this alternate universe) while his sister, a three-cow woman, dances in the Polynesian cultural center… if you’re a Latter-day Saint, you’re probably chuckling, but you’ve got to be scratching your head if you’re not. And to be fair, it’s not like I’d never write a book like that—after all, I’m still keeping Starship Lachoneus in my back pocket—but I wouldn’t try to market it to a general audience, or expect it to win any mainstream awards.

[SPOILER (highlight to read)]

The real kicker, though, was when I skipped to the end and learned that one of the book’s plot twists was that a side character got a secret abortion. That just pissed me off.

I’ve writte at length why I can’t stand John Scalzi and have DNFed him as an author. His first book, Old Man’s War, was good (though I would probably DNF it if I read it again now). Everything else I’ve ever read of his is just terrible. YMMV.

I’ve also DNFed Charles Stross as an author, mostly for nihilisim. If someone gives me a good reason to try Halting State, I’ll pick it up, otherwise I’m just going to skip it.

As for Brasyl, when I ran it by Orion (the ChatGPT persona I created to screen these books for woke and objectionable content), here is what I got:

Set in three intertwining timelines within Brazil—the past, present, and near future—Brasyl explores themes of identity, quantum mechanics, and cultural complexity. From an ambitious reality TV producer in modern Rio, to a rogue Jesuit priest in the 18th century, to a futuristic tech-savvy entrepreneur, the characters navigate moral dilemmas and existential challenges as their stories intersect in surprising ways. The novel combines speculative science with the vibrant, chaotic energy of Brazil.

Content Warnings:

Explicit Content: The novel contains scenes of violence, some of which are intense, and mature themes including sexual content. The explicitness aligns with the gritty, unflinching tone of the story’s portrayal of both human and cultural dynamics.
“Woke” Elements: McDonald incorporates themes of globalization, colonialism, and social inequality, reflecting a nuanced perspective on Brazil’s history and culture. The narrative celebrates diversity and critiques systemic injustices, which some readers may view as overtly progressive. However, these themes are intricately tied to the world-building and character arcs.
Overall Assessment:
Brasyl is a vivid, ambitious novel that combines speculative science with cultural and historical depth. While its explicit content and progressive themes may not suit all readers, those who enjoy richly layered storytelling with a strong sense of place will find it a rewarding and thought-provoking read.

It’s kind of on the border, but there’s enough in there to convince me that this isn’t the kind of book for me. The world is full of lots of great books out there; why should I spend any time on this one?

Which brings us to Rollback. I wanted to like this book, and indeed got about a hundred pages into it before I decided to DNF it. Partly, it was because I lost interest. Partly, it was because the writing was just too literal and logical—especially the sexy parts, which didn’t offend me (the characters are married) so much as they were just plain awkward. It’s not like I hated the book—indeed, I’d probably finish it if I tried it again as an audiobook—but I couldn’t give it my vote. If it weren’t for the other books on the ballot this year, I would have just abstained, but the others were bad enough for me to rank them under No Award.

How I Would Vote Now: All Years

Five things I did at work last week

  • I generated a 153k word rough AI draft of a fantasy novel.
  • I ran a free promotion for my title The Call of the Tide.
  • I wrote and scheduled two email newsletters.
  • I finished updating my paperbacks. They are all now available on my online store.
  • I decided which book I plan to write and submit for the Ark Press America 2076 contest.

A slightly better cover mock-up

So I was playing around with ChatGPT this morning, and came up with a slightly better cover mock-up for The Soulbond and the Sling:

Still needs some work, but I’m on the free plan and only get three images every day. I’ll keep playing with it over the next week or so, maybe try out another AI image generator like Stable Diffusion. But at least I have a cover image now that I can add to my posts whenever I’m talking about this WIP.