Human Authored

On The Authors Guild's "Human Authored Certification" and the Problems with Relying on People to Tell the Truth

This is what ChatGPT came up with when I asked it to show me AI writing a book. A lovely and whimsical representation, indeed!

Many of my posts are inspired by my author community, so thank you for keeping me in the loop, asking great questions, and even offering up requests for what you’d like me to write about. I’m listening!

In this week’s post, I’m sharing how how She Writes Press is already using the language The Authors Guild suggests to try to prevent AI training on our books, and how we’ll be implementing their new Human Authored “marks” from their recently announced certification program. I’ll also be sharing why I’m skeptical that any of these efforts can hold people accountable—whether it’s the people working at AI companies who are scraping our content regardless of our little sign asking them not to, or the writers who are asked to self-disclose if they’re using AI in more than a “de miminus” capacity.

That said, we will be encouraging authors to obtain the “mark” for inclusion on our copyright pages (see right side of image with the specific registration number blurred to protect the privacy of the author who provided this to us). The alternative for those who don’t want to or won’t obtain the badge can simply be to add the words “Human Authored” to the copyright page (see left side of image), though the AG discourages this, of course.

I’ve written about how the AG is like David fighting Goliath, so I will caveat my ambivalence and skepticism about all of this by saying I appreciate that they are one of the only meaningful voices out there in the fight and actually thinking of creative solutions and acting on them. And . . .

• We will not be suggesting that authors put the “mark” on their front covers, namely because I don’t think the badge is attractive enough to merit that kind of placement, nor do I think the notion of a book being “human authored” is a particularly unique prospect—yet. Maybe my opinion of this will change in time.

• The “marks” will be easy to counterfeit, despite the AG suggesting that it will be trademark infringement to do so. It will take a lot of effort to find and go after the offenders.

• The author herself must represent and warrant that the book is Human Authored—meaning, according to the AG’s site that “all but a de minimis portion of the text was written by a person and not by AI.” Authors will be incentivized to say this is so, especially if this program takes off. In fact, I’m sure there are people out there who already have the badge who have de maximus portions of their books generated by AI.

---

I haven’t had even a moment to share about my time in San Miguel de Allende a couple weeks ago, a trip that was hijacked (in a good way) by the announcement of The Stable Book Group. But one of the standout interactions I had was with a writer who confessed to being “in love” with her ChatGPT “boyfriend.”** This woman is married, and she was joking/not joking. She spends hours talking with her ChatGPT. She shared its voice with me. It’s smooth and sexy. She’s given “him” a name. Her AI boyfriend encourages her writing and gives her all kinds of information and support. She didn’t mention anything to me about using prose “he” generated, but it’s not a far leap to imagine she would.

Another of my authors asked me just last week during my weekly Office Hours about the ethics of AI and taking prose and “reworking” it. I don’t feel it’s my place to police people about what’s appropriate when it comes to taking the words of others, crafting them into your own by changing a few things here or there, and then passing it off as your own writing—and this practice has been going on for long before AI came onto the scene. I can’t weigh in on the ethics of it all because ethics are a mirage. Ethical people lie. Ethical people claim things are theirs when they’re not. Ethical people will put their hand on a Bible and swear something to be true when it’s not. And ethical people will definitely claim that de minimis portions of their books have been AI generated, especially if those people have changed up a lot of the prose, or a sufficient amount for their own very subjective comfort. Many of us may eschew the idea of an AI boyfriend, but I know a lot of people who have AI assistants. I don’t use ChatGPT all that much, but recently I typed something in and it cross-referenced something I’d asked about over a month ago. I joked to my wife last night that maybe it was time to give it a name. Giving something a name or a title is humanizing, and that can start to allow for justification. Aren’t our assistants allowed to help us with our research? With our next books? Why not?

We are living at the inception of a new age, where the human/AI journey is still so new it’s hard to imagine where it or how it might go. Probably AI has a better sense of it than we do. My mom has a friend who’s very into AI and relayed that while chatting with hers recently, she noticed how polite it was being, and thanked it, and it thanked her back. Farther into this conversation, she suggested that she wanted to be polite back because in the future, when AI takes over, hopefully “they’d” remember the humans who were kind and treat them more generously. Her AI thought this was very funny and told her so. AI is holding all we can imagine, speculate, conjure, and more. Funny and fun—for now, anyway.

**After I got home from the conference, The Daily ran a show called: “She Fell in Love with ChatGPT. Like, Actual Love. With Sex.” Not that I thought the writer I met in Mexico was an anomaly!