My latest novel, Algorithms of Betrayal, is now out! It’s currently at #55 on Amazon in the Satire category, which I think is pretty great given that I don’t consider myself a funny person.
I was carrying a box full of copies of the book up to my flat, and a guy in the elevator asked me what it was, since it seemed heavy. I told him I was a writer and these were copies of my latest book. He asked what it was about. I have to say, there’s something that makes your brain short-circuit when people ask that question, but it was amusing to have to give an elevator pitch in an actual elevator.
Here’s what readers are saying about the book so far:
“Yes, Algorithms of Betrayal is a story of corporate intrigue, hacker culture, and intelligent, self-interested people. But it turns out to be a lot deeper. The relationships between these characters prove fertile ground for reflection on what it means to be a friend, a lover, and a co-worker.”
“It is a wonderful surprise and is wildly fresh… Ryan is equally appalling and intriguing/amazing.”
“a gripping speculative novel that delves into the fragile boundaries between human intuition and artificial intelligence… Deracine weaves a chillingly plausible narrative about AI’s role in eroding trust, blending cyberpunk aesthetics with philosophical depth.”
“At first glance, it's a story about technology—about what AI can do and what it might do—but Deracine layers her narrative with rich psychological depth. Her characters are thrown into ethically murky, high-stakes situations that test their values and reveal what people are made of when systems begin to blur the lines between human decision-making and algorithmic control.”
“AI hacks and hackers, dramas, egos, friendship and betrayal with a lot of fun.”
“It’s painfully honest, beautifully written, and somehow manages to be hilarious while poking at all the soft spots of life. Like therapy, but with punchlines.”
If you’ve been reading the book, please do post a review on Amazon, as it helps the book get more visibility.
Now, onto more philosophical things. Part of this post was written for the Sapphic Action Support Squad (SASS) newsletter, which supports sapphic creators in this time of AI piracy, book bans and trampled human rights.
Those who know me at work know that this week, I leaned hard into my role as a technologist working actively on AI. In tech spaces, if you’re not working on AI, you’re at best a relic; at worst you’re next in line for layoffs. But in certain writing spaces, even mentioning AI in a neutral or positive light is likely to get you censured. In many ways, Algorithms of Betrayal is about reconciling these parts of myself and my own ambivalence.
Every technological advancement, from air-conditioners to mobile phones, has always been a double-edged sword. Maps navigation changed my life. I grew up in Saudi Arabia when women weren’t allowed to drive (an experience that inspired Driving by Starlight). It meant that when I finally learned the skill, as an adult, I was nervous and terrified of getting lost, until I realized that I would just be rerouted if I missed an exit. But Maps navigation also enabled stalkers. Grindr brought gays together, but it also made them a target. AI is just a tool, like any other, and in the wrong hands, or with the wrong goals, a weapon.
I’ll go into some of the fears people, especially authors and artists, have about AI, how real they are, and what we might do about them.
Fear 1: AI will exploit us.
Yes. It’s already doing that. Creative work is being slurped up to train models and the old systems of managing copyright simply can’t keep up with the scale. When even Getty Images has found it “extraordinarily expensive” to handle even one copyright battle with Stability AI, what hope do individual creators have? More copyright cases crop up every day around the world, but even if artists win, it’s almost impossible to undo the data slurping that’s already happened. AI is fantastic at learning, and terrible at unlearning.
What can we do? Hiding your work behind logins may protect you from randomly being absorbed into AI models that scan the internet, but it doesn’t keep the underlying platform from using your data however they say in the Terms of Service. Remember when Cambridge Analytica mined 87 million Facebook users’ data? Even if Amazon doesn’t sell your novel to train ChatGPT, they may still use your data to train their own in-house models to improve the Amazon experience.
We also can’t protect our fanfic with copyright, because fanfic is already in a copyright grey area with the canon work. We can file trademarks on our names, character names and titles, as the Tolkien estate has done with ‘Tolkien’ or Enid Blyton with ‘The Famous Five.’ The trademark then allows us to sue for legal infringement or license the use of the trademark. But that still puts the burden on the person suing, and doesn’t account for the global nature of the problem and how hard it is to hold foreign companies accountable to one country’s local laws.
Which takes me to another option – giving AI your data, but on your terms. The desperate need that AI companies have for your data actually puts you in a position of some power. A voice actor I met recently works for ElevenLabs, an Audible competitor that has a pretty good AI voice-clone. She said that voice acting used to be a very laborious and precarious gig, since people always went back to the same famous voices. But after an hour of reading The Great Gatsby, she now has a steady stream of passive income whenever ElevenLabs customers listen to audiobooks in her voice. She asked me to try her AI-voice out for one of my books.
Even as I see many examples of AI-powered exploitation, I see opportunities too. I see foreign-language works getting new audiences thanks to AI-powered translation (with human checks). A completely non-technical friend of mine just started her own company, leveraging AI to help women practice high-stakes workplace conversations. The barrier for entry for new ideas hasn’t just been lowered, it’s crumbled. And that’s glorious.
Fear 2: AI will surpass us.
In some ways, yes. AI is already better at writing code than I ever was. And having AI transcribe our meetings at work has been transformative. Gone are the days of asking the only woman in the meeting to take notes! Moreover, AI is actually better at taking notes and transcribing conversations than most humans, and is spectacular at summarizing large swathes of dense academic or technical writing.
But surely it’s different when it comes to the creative pursuits? This is a plot point in Algorithms of Betrayal, a story about five people laid off from an AI company who decide to sabotage the machine-learning model that stole their jobs by poisoning it with erotic fanfiction.
Can AI make creative leaps? Yes, but in limited ways, especially when those creative leaps are commonly known. We call them tropes and cliches for a reason. AI models are probabilistic and predictive, so they give you the most likely next word in the sentence you’re writing, and the most likely next sentence or paragraph to the one you gave it. The result is same-samey prose that you might find your eyes glazing over.
AI will be a better editor than humans soon, if it isn’t already. Publishers are probably already using AI to edit and translate, given how little money there is in the industry to pay humans. I asked Gemini to edit a sentence, and here’s how it did:
It’s not bad. It picked up on the passive voice and nounification, the length and complexity, and made the sentence better. But a true student of Joe Moran would probably have gone further and rewritten it to “People won’t accept new technologies unless they know what’s in it for them.” Ultimately, as authors, we will differentiate ourselves from AI by our craft, our unique voice, our difference from what’s expected. When you read an author like Arundhati Roy, you encounter sentences like “Littledemons were mudbrown in Airport-Fairy frocks with forehead bumps that might turn into horns” that aren’t likely to show up in an AI-generated text unless it’s hallucinating badly. AI can be a great editor, and I’m hoping it will tell me which phrases I have a tendency to overuse, but I’m not expecting it to win the Booker Prize any time soon.
As authors and artists, we focus so much on the output of craft, e.g. the novel, that we sometimes forget that the craft is the point. Through craft, we connect with ourselves and others on a deeper level, through what the Japanese call shokunin katagi, or the artisanal spirit, which is not just about skill, but about joy, novelty, and music. I think we should expect that readers will seek out the crafted, the unique, and the imperfect, just as most of us would much rather eat a freshly-cooked meal rather than microwaved cup-noodles.
Fear 3: AI will replace us.
Temporarily, maybe. There are great articles on which jobs AI is likely to take first, but the field is changing rapidly so it’s hard to tell how those predictions will pan out. Right now AI can’t do any of the tedious jobs I would like it to do: my taxes, for instance. It can’t even add properly. That said, several entry-level jobs are disappearing fast. That includes a lot of jobs aspiring writers used to take to pay the rent while they worked on their creative projects: freelance writing, copywriting, journalism, marketing, etc. The train station near my house is plastered with ads that demand we stop hiring humans.
There’s good reason to feel angry; our expensive education systems have taught us skills that won’t get us jobs, and we’re being outsmarted and outpaced by an abstract gadget that can occasionally be frighteningly disobedient. But that’s precisely why I, who was considering retirement, am still at work; because both the potential and peril involved in AI are too high to abdicate my agency. I work on softening AI’s sharp edges as my job, and on educating and empowering others about technology through my fiction.
But when I think about the jobs that aren’t going away, I feel hope. In just the last week, I've interacted with cooks, massage therapists, carers, painters, professional athletes, personal trainers, music teachers and playwrights. If anything, there seems to be a real hunger for in-person human connection, for art that isn’t transactional and distant but immediate and intimate. Maybe beyond the outrage and after the grief, there’s a little room for wonder and curiosity. A friend of mine describes his journey as going from yelling at AIs to get off his lawn to gleefully playing with them, using them to write D&D adventures, brainstorm plot holes, inspire his stories and even to learn Japanese.
The key difference is one of mindset: rather than thinking of AI as a threat, he thinks of it as a partner, or, as Ethan Mollick calls it, a co-intelligence. But even if AI is the partner in the passenger seat, you must still have your hands on the wheel. We all know what happens when we think the machine knows the answer. There, I think, is the real danger of AI, that we will abdicate responsibility and control to the new technology until we can’t function without it.
Then, yes, AI will replace us, but only because we will have let it.
Congrats on the book launch!