AI in Agent Submissions

At the risk of sounding like a broken record, I have some more AI thoughts and news this week, as the topic continues to be hotly debated. I’ve joined the Editorial Freelancers Association’s AI taskforce, so you can expect more frequent discussions on this topic.

AI in agent submissions

First, an article in The Bookseller (paywalled, but you can register to read a few articles for free), describes how literary agents have noticed a shift to widespread use of AI in submissions.

I can see the allure. Many writers struggle with their query letter and synopsis. I would expect even writers like myself, that is, those who don’t use AI tools at all in the novel drafting and revising process, might be tempted to ask ChatGPT to come up with a snappy summary of their novel.

Antony Topping, literary agent and managing director of Greene & Heaton, said:

“There’s an uncanny, flattened nature to a lot of the manuscripts; the submission letters are becoming quite formulaic. It’s quite a subtle thing to get across, because, superficially, a lot of these letters and submissions look quite plausible until you spend a little longer reading them, and then you realise that there’s something else going on,” he said.

“What we’re most concerned about is potentially good writers who don’t trust themselves, and they are trusting this concept of AI that they’re told repeatedly is so brilliant and transformative, when some of them are missing the opportunity to discover that they themselves might be a great writer.”

As I’ve said before, besides the flattening of voice, one big problem with using these tools for your own writing is that they obscure and cover up what knowledge about writing craft you do have. I really like the framing here around trusting yourself as a writer, because confidence is a super important piece of the puzzle when querying.

I’ve also noticed an uptick in requests on Reedsy (a platform that connects authors and editors) where the author has used (or I’ve strongly suspected they have used) generative AI tools for novel drafting and revision. Reedsy asks authors to disclose voluntarily whether and, if so, how they have used gen AI tools in their process, but it doesn’t require that authors make such a disclosure, and many authors clearly choose not to fill in that box.

As an editor, that creates extra work on my side, as I have to have those conversations upfront. I’ve found Reedsy quite frustrating in this aspect because the platform design deliberately lowers the barrier to contacting an editor: it’s relatively easy to send five requests on the platform to editors, whereas it takes a little more research and guts to contact an editor through their website. This lowering the barrier means a lot of authors who aren’t quite ready for editing often get in touch, and I’ve had to develop a triage process for that – which includes asking questions about AI assistance.

Finally, I’d also like to share a quote from a great essay by Ayana Zaire Cotton: Transmuting AI Shame Into Erotic Self-Respect: On Listening To Your Body, Not The Machine:

I want to talk about how we completely refrain from using AI inside the desire to increase our productivity, a tender conversation I don’t see us holding enough space for. I didn’t start using AI “for fun”, I started using AI because I was burnt out and I’m writing this because I’m sure I’m not alone. There are folks who are using ChatGPT to tell them what they do for work and then there are the folks who are using ChatGPT to get through the work day. Not because they can’t do their job without it, but because we exist in a work culture that worships speed over spirit and “breaking things” over the sacred work of tending, mending and maintenance. This is what came up in the conversations I had. Some of us are using AI to “create an advantage” over folks capitalism recasts from comrade to competitor.