After an open and assessed conversation with our members, advisors and other interested parties, the Alliance of Independent Authors has developed some practical and ethical guidelines to AI for authors, answering questions like: How can AI help me become a better writer, a better publisher? And what are the ethical boundaries for authors? With thanks to Joanna Penn, ALLi's AI & Enterprise advisor, Laurence O'Bryan of BooksGoSocial, Amit Gupta of Sudowrite, JD Lasica of Authors AI, Holly Payne of Booxby for their contributions to this post offering practical and ethical guidelines on AI for authors.
When an author hears the words “artificial intelligence” it usually evokes one of two responses–excitement at the prospect of powerful new technologies for bringing out more books and better books. Or fear fed by movies like I, Robot, the Matrix and the Terminator that the machines are going to displace humans.
The job of the independent author is to write great words, publish great books, and make them available to the readers who will value them most. Today, AI (artificial and augmented intelligence) and machine learning systems are making certain parts of that job easier than ever before.
AI is the single most important development facing humanity in the first half of the 21st century and already our most powerful technology. Authors need to understand the implications of this, so we can surf the waves of change, rather than drowning in their deluge.
Understanding something new and unfamiliar always begins with the questions we ask.
Many questions raised by concerned authors center particularly on the ever-improving ability of AI to generate creative writing–see this previous post for more details of how AI tools based on Natural Language Processing (NLP) and Natural Language Generation (NLG) are generating poetry, plays, films, and even full novels–though the quality for longer works is not there yet, things are moving fast and it's only a matter of time before AI can handle longform text well.
Given that AI can already generate or translate the draft of a book faster than you can read this blog post, should we all pack up our word processors and go home? Short answer: No.
Using an NLP-based tool to aid your writing productivity, for example, is not about clicking a button and outputting a fully formed book. Yes, you can generate text based on parameters that you select, but you still have to drive the tools, you still have to curate and edit the output.
As a comparison: you might have the fastest supercar sitting outside your house, but it won't go anywhere or do anything unless you get in and tell it where you want to go. With AI, it's the same. You provide the destination and the direction.
As AI writing and publishing tools expand in their capability, you'll always have something that a machine, a tool, a robot, an automaton, a computer, by definition, does not have: agency. You get to decide how you harness this power and where you draw your ethical lines.
AI for Authors: The Questions to Ask
In framing the questions we want to ask about AI, we do best when we think of AI as a range of tools that can help us in our work. As Kevin Kelly says in The Inevitable:
“This is not a race against the machines. If we race against them, we lose. This is a race with the machines. You’ll be paid in the future based on how well you work with robots … It is inevitable. Let the robots take our jobs, and let them help us dream up new work that matters.”
The most important questions for a working indie author to ask about AI right now include:
- Which of my writing and publishing challenges can be met by AI solutions?
- Which AI tools do I want to implement in my writing and publishing?
- Where do I draw my creative, commercial and ethical boundaries with AI?
We have attempted to answer these and other key questions below. This is early days and all of this is a work in progress. Please keep those questions (and answers) coming!
To learn more, and to keep up with this rapidly changing arena, we highly recommend that you follow our Enterprise and AI Advisor, Joanna Penn, who keeps authors up to speed in the AI section of her website and podcast and whose book Artificial Intelligence, Blockchain, and Virtual Worlds is a terrific introduction to the topic for authors.
AI for Authors: Practical Guidelines
In the ALLi glossary, we define artificial intelligence as “technology that replicates human behaviors and abilities conventionally seen as ‘intelligent' and augmented intelligence as “applications or tools that combine human and machine intelligence.” Both are referred to as AI.
As machines become increasingly capable, the definition of what requires “intelligence” changes. As Larry Tesler, the computer scientist, famously said, in a quip that has become known as Tesler's theorem: “AI is whatever hasn't been done yet.” As soon as AI successfully solves a problem, the problem is no longer a part of AI, a phenomenon known as the AI effect.
You are already using AI processes in your author-publishing businesses that have come, though use, to feel like everyday tools e.g. Google search or Amazon algorithms. These are examples of what's known as “narrow AI”, offering specific expertise in a particular area e.g. Google's Search AI is better at search than you, using an encyclopedia or library. Amazon's AI recommendation engine is better at recommending the right books to millions of readers than the most knowledgeable book reviewer can ever be.
Some recently introduced tools based on narrow AI of interest to authors include advertising algorithms, reader analysis and engagement, search and sales engines, translation software, and speech-to-text and AI narration for audiobooks.
With complex activities like writing book-length text, or AI narration for audiobooks, we are talking specifically about NLP and NLG tools, as referrenced earlier–AI that creates words based on training algorithms on big datasets. Such tools are already being used in the generation of journalism on major newspapers, and making inroads in creative works like novels and nonfiction books, screenplays and poetry.
Producing AI-generated text just got easier with the arrival of GPT-3 (Generative Pre-trained Transformer, Version 3) in June 2021. GPT-3 is currently in beta but it won’t be long before it is open to all.
OpenAI, the nonprofit organization that created GPT-3, wants “to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.” In 2019, Microsoft invested a billion dollars, which has allowed OpenAI to create one of the top five supercomputers in the world, processing more than 23,000 teraflops per second.
This investment has proved transformative. “GPT-3 hints at a world in which machines can generate language,” says The New Yorker in a recent article, The Computers Are Getting Better at Writing. “The consequences are vertiginous… whatever field you are in, if it uses language, it is about to be transformed. The changes that are coming are fundamental to every method of speaking and writing that presently exists.”
There are a lot of new tools being built on GPT-3, new ones popping up every day and this is only the beginning of such AI tools. In recent weeks, a Chinese language model 10x bigger than GPT-3 was launched, as well as an open-source language model, EleutherAi, all of which will accelerate tool development in language processing.
As the use of AI systems in the production of creative works grows ever more common, more and more authors want to use these tools to generate drafts.
AI & Authors: Practical Guidelines FAQ
Will AI make writers defunct?
How might an author use writing AI?
We already have so many books and competition in publishing. What about market saturation?
The market is already saturated. There are already many millions of books available, and many other forms of entertainment, for example, gaming, TV, and film. But we still create, and readers still find our books. AI tools can actually help discoverability, by reading the emotional resonance of a book, then using ever-improving recommendation algorithms to help us find more of the right readers. (See Booxby Case study below)
Also, readers are not stupid and they read for human connection. They know how to find what they love to read and make choices to support creators. As AI tools improve, authors have to double down on being human. How? By finding the value in what you offer and doing it more deliberately: more often, more intently, more freely.
The publishing future, like the past, belongs to those who can personally communicate and engage readers and stand out with a unique and identifiable voice—not those pushing a punishing, cookie-cutter, productivity model that cannot outpace the machines, or that passes off machine-generated work as their own (see ethical considerations below).
How can we make the most out of AI as creative business owners?
- Which of my writing and publishing challenges can be met by AI solutions?
- Which AI tools do I want to implement in my writing and publishing?
AI & Authors: Ethical Guidelines
The third of those core questions was:
- Where do I draw my creative, commercial and ethical boundaries with AI?
Ethical questions are always complex and as well as giving rise to commercial and creative application challenges, have societal and environmental implications too. Societal hazards that have been raised in relation to AI include loss of trust, deception, infringements of privacy and confidentiality, and loss of employment.
In publishing, there are also ethical considerations around copyright, which we have covered in this post about ALLi's stance on copyright implications of AI.
A new generation of ethical standards is emerging slowly, as we come to better understand the environmental and societal impact of AI, but it's early days. Such codes as do exist are as yet in development, with little public information. We need ethical guidelines for AI technologies in the creative industries and the publishing industry is not exactly to the forefront on such matters.
Here in the UK, British Standard BS 8611, guidelines prepared by the British Standard Institution “for the identification of potential ethical harm arising from the growing number of robots and autonomous systems being used in everyday life”, defines ethical harm as affecting “psychological, societal and environmental well-being” as well as economic rights.
It also recognises that we must balance physical and emotional hazards against expected benefits. And it highlights the need to involve all the stakeholders, including the general public, in the debate about the development of AI and robotics, as well as any ethical codes arising.
In publishing and the other creative industries, there are many challenges, including but not limited to: fair sharing of benefits; assignment of responsibilities; exploitation of workers; the environmental cost of high energy needs and copyright, not to mention the more complex and less certain implications of AI, such as those regarding human relationships.
Those creating AI tools and conducting research in the creative industries, and beyond, need to engage the public, consider public concerns, work with experts from other disciplines, correct misinformation, and provide explicit instructions.
What Do Ethical Authors Need to Think About?
In this uncertain environment, ALLi is aiming to empower authors to think for themselves and make their own ethical decisions in using AI. From an ethical perspective as a self-publishing author you need 1) to be able to undertake an ethical risk assessment of an AI that you're considering for use; 2) identify any potential ethical harm; and 3) mitigate any ethical risks you might find.
Based on its list of key considerations, and responses to our call for comments, ALLi recommends the following principles to be upheld by ethical authors in relation to AI:
- Humans remain responsible agents and must recognise their own responsibilities in using an AI tool
- It must be possible, and should be easy, for all users to find out who is responsible for any AI tool
- AI should never be designed to be deceptive
- User privacy should be built into the design
- Users should not be forced to use an AI if they choose not to
- Users should be aware that behind every apparently human-free machine and process are real people.
- No worker should be exploited in the creation or use of AI systems
- AI should expand not diminish human relationship
As part of our ethical self-publishing campaign, the Alliance of Independent Authors runs an ethical author program. This program provides a code of conduct for authors to follow, in relation to author ethics. The code is voluntary and open to all, and is additional to the required Code of Standards signed up to by ALLi members. Any author, regardless of publication method or membership, can sign up to this code, once they agree to the guiding principles.
We have added a new clause, to the effect:
Use of Tools and AI
I edit and curate the output of any tool I use to ensure the text is not discriminatory, libellous, an infringement of copyright or otherwise illegal or illicit. I recognise that it is my job to ensure I am legally compliant, not the AI tool or service I use. I declare use of AI and other tools, where appropriate.
You can find ALLi's Ethical Author information here.
What are examples of using AI unethically as an author?
Cutting and pasting generated text, willy nilly. Not checking the output. It is your job as an ethical author to edit and curate the words generated by a tool you've prompted, to ensure the text is not, for example, derogatory or offensive. AI tools can be used for idea, character and story generation as well as text generation. If you copy and paste text generated by an AI, as well as adapting and editing, run the final work through a plagiarism checker to ensure you have not unwittingly infringed someone's copyright.
Should you declare your use of AI tools?
It depends on the context. For example, 1 The Road by Ross Goodwin declares the use of AI, as does Pharmako-AI by K Allado-Mcdowell. The co-creation is embraced as part of the point of using the tool so it is more appropriate to declare it. Orna Ross will declare her use of Sudowrite in her books, as she uses all its functions. By contrast, Joanna Penn is currently using Sudowrite as an author might use an online thesaurus, and does not see the need for declaration in her current usage, but will do so if specifically using generated text in a finished work.
What effect does AI have on the Environment?
AI for Authors: Being Human
Finally, keep in mind that there is one thing that AI, and indeed other authors, can never do — be you. Specialize in giving your writing your particular stamp. In your writing, communicate from your deepest experience, using your own voice, telling your own truths. In your publishing, focus on local, imperfect, real connections with other human beings–your readers and the authors in your niche whose work you most admire.
Think about incorporating audio and video, where you cannot help but reveal yourself, into your marketing, as a supplement to your books. But however you do it, be personal, be honest, be authentically you. Personality, values, personal branding: these become ever more key. As does true human connection with your readers.
Amit Gupta, who created Sudowrite with his business partner so that authors might collaborate with AI to more effectively and playfully realize their creative vision, insists: In the end, I’m still the writer. I bring my experiences, my history, and my taste to the subject matter. And I need to make the choices and do the work that imbues my piece with pathos and meaning.
AI for Authors: Case Studies
Orna Ross and Sudowrite for writing
Sudowrite is a GPT-3 enabled AI writing tool and Orna Ross is a novelist and poet and director of the Alliance of Independent Authors
When my friend and ALLI's AI and Enterprise Advisor Joanna Penn first introduced me to Sudowrite earlier this year, I was sceptical at first. ALLi has contributed to the UK government’s IPO (Intellectual Property Office) consultation on Artificial Intelligence (AI) and Intellectual Property (IP) (see consultation outcome here) and I have written and spoken widely about the challenges AI presents to framers of copyright legislation.
But also, I love to write. I could well imagine the glee of some people at the idea of pressing a button, slapping a cover on the text that spewed out, and selling it on online retailers. No doubt, there are already such books for sale. For me, playing with sentences and stories is what I most love to do. Why on earth would I want a machine to do it for me?
But Joanna, who was also a contributor to the IPO consultation, and is as concerned as I am about copyright and ethics, has had the experience of dragging me by the ear into tech adventures before, and was insistent: “You’re going to love it. And there’s nothing to worry about, copyright wise, if it’s used ethically.” She was right.
As I go along, writing to my outline or veering off in unexpected directions, it throws up suggestion after suggestion for plot twists, character types, and sensory descriptions—-sights, sounds, smells—-as well as poetic lines and images.
This allows me to progress faster and more freely. From “reading” my work, the AI then suggests new characters, story directions or phrases that can enhance what I'm portraying.
For me, it’s like having another writer in the room, but one who is tireless and never runs out of ideas.
Would it work for everyone? My guess is no. If you’re a speed reader, happy to skim text and light only on the bit that's most useful, knowing just how to use it while ignoring the rest, it will give your creativity and productivity a boost. If you get confused by options, put off by things that don’t make sense (the AI goes a bit crazy sometimes), or take a long time to sift through text, it might well have the opposite effect. You'd have to try it to know.
For me, it's working. I look forward more to my writing sessions, knowing it's there. I'm also looking forward to the improvement, upgrades and new features Amit and James plan to add. And to seeing what other ways AI can help me write more books and reach more readers.
JD Lasica and Marlowe for Manuscript Critique
Marlowe is an AI virtual developmental editor / critique partner and J.D. Lasica is a thriller author, entrepreneur and chief experience officer of BingeBooks, Authors A.I's book discovery site.
Story archetype and plot structure
Primary emotions color wheel
Offer to ALLi members
If you’d like to explore this new frontier yourself, we’ve set up two special coupon codes for ALLi members: To run a free Marlowe Pro report, enter ALLiFREE at checkout. To run more than one report, use ALLi20 to get 20% off. There’s also a free Marlowe Basic version.
Holly Payne and Booxby for Book Discovery
Booxby is an AI driven platform that optimizes the marketing of story content to help creators reach their widest audience and Holly Payne is a novelist and Booxby's founder and CEO
I wanted to know if there was a more efficient and intelligent way to answer the three most critical questions that guide any marketing campaign; who would love my content? Why would they love it? And how big is the market? With so much existing content, answering these questions has become a big data problem, but hopefully, and as we are about to discuss, one that AI is uniquely suited to solve.
I never thought in a million years, I’d be using artificial intelligence to solve this discovery problem, but the age of machine learning is upon us. It’s here and it’s changing the way we will acquire, market, and discover books forever. Some call this the fourth wave of disruption. I like to see it as the fourth wave of opportunity.
… The publishing world is full of lore about what sells and what gets read but precious little of the lore is informed by data and analysis. Publishers have been and using reader analytics for many years, as we know. Companies like JellyBooks and Inkitt engage readers, and then they analyze the book experience by embedding tracking software into digital advanced reader copies, ARCs. The tracking software is activated by roughly 300 focus group readers who sign up to get free books in exchange for providing information about the reading experience. But here’s the thing, while the insights gained are really helpful, the challenges with this process are timescale and sampling bias.
And then on the consumer side, metadata and collaborative filtering have been the key techniques for solving book discovery. Unfortunately, neither offer personalized recommendations. For example, customers who bought Y also bought X, that analysis doesn’t understand why anyone chose X or Y. Of the avid readers we interviewed, none of them were satisfied with Amazon’s book recommendations and most distrusted or ignored them. And now we understand how those recommendations are actually made, a lot of those banner spaces, that’s paid for by publishers, so those books don’t actually have a true mapping. There’s no mathematical correspondence between them. It’s been paid in, essentially, advertising space.
And while helpful in many pursuits, this technique doesn’t generate recommendations specific to a reader’s (inaudible) taste and mood. That's what Booxby does. We uncover the literary DNA of your work to see where it fits in the larger publishing landscape.
AI for Authors: Resources
Blog Posts & Podcasts
Joanna Penn: List of AI writing tools (regularly updated). This is a list of growing AI tools that you can use to help you generate a variety of different types of writing from prose, inspiration and prompts for your own prose, poetry, marketing copy and more. See also: 9 Ways That Artificial Intelligence (AI) Will Disrupt Authors And The Publishing Industry and Joanna's interviews with authors using AI tools:
- Cowriting with Artificial Intelligence with Yudhanjaya Wijeratne
- The AI-Augmented Author. Writing With GPT-3 With Paul Bellow
Orna Ross & Joanna Penn: AskALLi Advanced Salon: Artificial Intelligence and the Indie Author Orna Ross interviews Joanna Penn about recent AI developments
Monica Dube: How Can Writers and Publishers Utilize Artificial Intelligence Blog post from PublishDrive
Holly Payne: AI and Book Discovery: SelfPubCon (Self Publishing Advice Conference) session where Holly Payne introduces AI as a book discovery and author comparison tool.
Argues that Universal Basic Income is only part of the solution to how we'll manage a world where AI and other tools do most of the work. We are probably going to need an entirely new economic system. You may not agree with everything in this book but you will certainly be given food for thought.
Davenport, Thomas and and Kirby, Julia. Only Humans Need Apply: Winners and Losers in the Age of Smart Machines
The authors reframe the conversation about automation, arguing that the future of increased productivity and business success isn't either human or machine. It's both. The key is augmentation, utilizing technology to help humans work better, smarter, and faster.
Tools methods and best practice guidelines for those developing AI products in a human-centered way. Originally launched in 2019, now updated with new insights, offering a set of methods, best practices, and examples for designing with AI.
Lee, Kai-Fu: AI Superpowers
Taiwan-born Kai-Fu Lee believes China will be the next tech-innovation superpower and in his new (and first) book, AI Superpowers: China, Silicon Valley, and the New World Order, he explains why.
How authors and publishers can embrace the opportunities and engage in conversations around AI in ways that positively direct the reinvention of our industry. Clear, practical and eye-opening. The must-read book for authors and publishers.
A rich and visionary exploration of whether AI will machines eventually outsmart humans and replace us altogether or help life on earth flourish as never before.