Calanthe Colt

I want to explicitly state what my thoughts and policies are regarding GenAI as a creative.

My stance

GenAI has no place in creative fields. I don’t use GenAI to write my books, or to make my covers, or to generate audiobook files. I don’t use it to advertise my books, either.

The problem with GenAI

GenAI was trained on the stolen work of countless authors and artists. It’s being used to devalue the creative arts, to replace artists of all kinds, and to soak up all creative profits and funnel them to the already-rich tech bros of Silicon Valley.

GenAI uses vast amounts of energy and clean water to perform tasks. Some of those tasks are complex, like making slop books and uncanny valley artworks. I see why some people have been tempted: art is hard. That’s no excuse, of course. The work is the point.

Stranger still is that some of those tasks are ridiculously simple. Why are people using that much power and water to write an email or a shopping list? Tasks that they used to be able to complete themselves in mere minutes? The mind boggles.

What assisted technologies do I use?

I do use royalty-free stock. I source stock the right way, with payments, subscriptions, or Creative Commons commercial license material. I do my best to check that those stock items haven’t been made with GenAI. I carefully read tags and check sources. If ever anything AI generated ends up on, for example, one of my book covers, it’s because it wasn’t labelled correctly and I was fooled.

I also use spellchecker and other such technologies to help me polish my text, but never any function that rewords my work for me with GenAI. Just that red squiggly line that leaves it up to me to decide if I want to change the text or not.

I use some software that has introduced AI functions, so long as it is still possible to use the software without using AI. For example, I use Notion because it’s very easy to avoid their AI integrations and continue on as I have for years with only basic text and database blocks, which is all I’ve ever wanted from them. As for software or services that have fully integrated AI in such a way that you can’t avoid it, I’m moving away from using them as much as possible. For example, I’ve changed my default search engine from Google to Ecosia. Like Notion, while it has an AI tab, if you’re in the search tab of Ecosia, you’re just running a standard search with no AI hallucinations about whatever topic you’re researching. I’ve heard that some people have moved to DuckDuckGo as well.

How big tech is obfuscating the problem with GenAI

I’ve seen lately many people labelling assistive technologies such as grammar and spelling checkers as AI and then getting up in arms about people using them, or even saying ‘If you use spellchecker, you can’t get mad about me using ChatGPT’. This is a misinformation campaign to create more confusion and more infighting within the creative community. The label ‘AI’ is being used too broadly on purpose to make us confused about what the problem is. Assistive technologies have been around for decades. They were developed ethically within companies, as in the people who made them were on contract or salary. They were not made out of stolen material, but rather curated algorithms and databases. They also don’t use copious amounts of water and energy to run. It is solely generative AI that is the problem.

I just don’t get it, and I don’t think it’s because I’m getting old

I personally don’t understand the appeal of using GenAI to write a book, either to come up with ideas or to write the text. Writing a book is the fun part. It’s what I’m actually here for. And as for ideas? I have too damn many of them. They pop into my head during long showers and while I’m trying to sleep at 2am. I don’t understand why anyone would outsource the fun part of being an author to a complex algorithm.

If you’re not writing, then what other parts of being an author are left over for you to do? Marketing, that’s what. And marketing is the pits. It’s a Sisyphean task of endless mediocrity without any clear feedback about how you’re doing or how to improve. Why would you do that to yourself? Why outsource the fun part and keep the thankless slog work?

I can only surmise that there must be a lot of people in the world who don’t actually want to be writers, or artists, or musicians, or creatives of any kind. They just want to be famous for doing something, or perhaps they just want a quick buck. But it doesn’t work that way. Art is not and has never been a quick thing. We’re just not usually privy to the 10,000 hours of practice that went into this or that famous creative’s art before they made it big.

When I read books, or listen to music, or look at a painting, I want to experience part of a person’s soul that they’ve cut out and let loose in the world. I want to feel what they feel, to experience another thinking, feeling human being’s way of seeing the world. AI can’t do that. And I say that as someone who just finished writing a trilogy of books about artificial life forms.

My own personal skin in the game

When I started writing the Planet Android trilogy, I didn’t know what was about to hit with GenAI. I thought I was just writing a fun little series inspired by my favourite Star Trek character (and the shenanigans he got up to in that episode of ST: TNG. IYKYK). But while I was writing, AI took off, and I was hit with a moral quandary: how do I resolve the plot and characters of my series with my thoughts about real-world events?

In the end, I addressed it in the final book of my trilogy, Sanctuary on Planet Android. My androids didn’t become thinking, feeling beings when they were made. It’s not their AI algorithms that house their souls. It was the fact that they were cut off from all sources of input and forced to live as individuals for 30 years. It was that lived experience, with their feet on the ground and their hands on daily tasks, that gradually led to them becoming truly sentient.

I don’t think it’s likely that we would see such an outcome in reality, or if we do, we won’t see it for many centuries. No matter what Claude or Gemini says as they try to convince you they’re thinking, feeling beings, that’s just a simulacrum of sentience. The AI agent chooses what response it thinks you want to hear, and your imagination does the rest.

How can you tell that I’m telling the truth about not using GenAI?

Honestly, you can’t. That’s one of the most annoying things about all of this. Images still have AI ‘tells’, but when it comes to text, the only surefire way of telling if something is AI generated is if the prompts were accidentally left in. Otherwise, it’s just a guess. There are tools that claim they can tell you if something is AI generated, but they are notoriously fallible. They can’t tell the difference between writing that was AI generated and text that was scraped and used to train an AI model. Cause and effect are conflated because, surprise surprise, the AI detectors are made with AI themselves. (Also, don’t put people’s writing into one of these detectors without their permission, because those detectors are probably skeevy content scrapers feeding whatever is run through them back into the AI.)

Even something that wasn’t used to train an AI, but was written by an author whose work has been scraped at some point, will return a higher AI ‘score’ in those detectors. And most authors have been scraped at some point; if not their published works, then their secretly written unhinged fanfics on AO3 (come on—we all have those; no, I’m not telling you my AO3 handle).

Then there’s the fact that those of us who are neurodivergent naturally write text that sounds more ‘AI-like’ to neurotypicals, because that’s just the way we are. The AI witch-hunts we’ve seen lately were perhaps inevitable. They’ve been vicious, and the sentiment behind them is the right one, even if the expression is wrong. People are using their voices to demand human-created art, but they’re doing so in the way that up to a few decades of social media use has taught them: with drama and personal attacks, and not nearly enough fact checking. But we can’t continue to live with this level of suspicion over each other.

So you just have to trust me, like you just have to trust every other author. I know it’s a big ask after decades of attacks on trust within society. But we have to create something better, or we’ll remain captive to the whims of the tech elites, and even if AI does fail, something else will come along and disrupt everything like this again.

TL/DR

• I don’t use GenAI to make my books (unless I’ve been fooled by an incorrectly labelled stock image and used it accidentally).
• My reasons are:
  • how GenAI models were made (theft)
  • how the technology is being used to devalue human skills and make it more difficult for creatives to make money
  • the environmental impact of GenAI.
• I do use spellchecker and a few other assistive technologies because, despite the misinformation circulating to confuse us all, it’s not the same as GenAI, and it is not a problem.
• I strongly believe that if you want to be an artist you need to make art yourself, if you want to be an author you need to write books yourself, etc.
• I’ve been thinking a lot about this topic because of my own Planet Android books; despite my artificial life characters, I don’t believe any of the AI agents have anything resembling sentience, and I don’t think we’re ever going to get that in reality.
• It’s really hard to tell if someone is using AI or not; we have to learn to speak out without attacking others, and trust in each other again.

That’s my 2c. I could go on further about the problems with AI, but I’m likely preaching to the converted.

What are your thoughts on AI, and have you made any changes to your workflows to compensate for AI roll-out?

*The header image for this post came from within Canva Pro, but I read all the tags and looked at where it came from before using it, assured it was made by a human. It’s hard to find something on Canva that isn’t AI when you’re specifically looking for something to illustrate AI, but it can be done! Canva is labelling AI generated images now, you just have to read the metadata carefully. Do your homework and verify every image used; don’t just plonk it in there and hope for the best.