Who Killed NanoSessionMo?’s AI Stance

Was going to do a quick “Where I’ll Be” post about this weekend’s upcoming Edge-Lit convention. Instead, that will be coming later this week.

In the meantime, it seems we have to do this.

Next Saturday, at Edge-Lit Genre Fiction Festival, I am going to be running a workshop called Who Killed NanoSessionMo? where a group of strangers collaborate to write an entire whodunit in under 90 minutes.

Nanowrimo, an unrelated and unaffiliated event where people try to write a novel in under a month, has announced a policy that allows the use of AI in Nanowrimo submissions.

So it looks like we’ll have to make our own policy clear.

This post represents the views of me, Chris Farnell, the writer running the Who Killed NanoSessionMo? Workshop, and not any of the organisers, staff, or the organisation as a whole that runs Edge-Lit.

First things first, my policy on people using AI to write their chapter of the story NanoSessionMo is set to produce is – that’s not possible, because AI that can write stories doesn’t exist.

However, I’m aware that some people might be tempted to use a Large Language Model, a program that draws from other written materials to statistically infer the most likely written response to a query or “prompt”.

Short version – I’m against it.

Longer version – I didn’t sign up to organise this event to become a Bladerunner. I’ve neither the resources or the inclination to patrol the workshop with invigilators looking for people using LLMs.

Likewise, particularly given the timeframe we’re working in, I’m not under any illusions about being able to “detect” material written by an LLM, especially as the tools available for doing that disproportionately penalise neurodivergent writers.

So, the only real barrier to anyone considering bringing an LLM to our workshop on Saturday is the knowledge that by doing so you will be putting a turd in everyone’s else’s paddle pool.

But more than that, you’ll be playing yourself. Aside from the jokes and the dare of it, the whole point in this workshop is to create something yourself – to remove all the barriers like time and planning and wondering if your work will be any good, just to enjoy the act putting one word in front of the other at speed until at the end of it there is a story that didn’t exist before. That is a great feeling, and adding some tweaks to the regurgitated output of an elaborate autocorrect simply does not provide the same high.

That said – Scarlet Ferret, the bookshop that will be hosting whatever monstrosity our workshop produces, is a good deal more stringent in its policies (Which you can read here) so if you do contribute LLM generated text to this workshop, and it is later discovered that you did, Scarlet Ferret will be well within their rights to take down the entire book.

In short – I really strongly recommend that you write your own material for this workshop. Will it be bad? Probably! But I assure you it will be bad in a more interesting and inspiring way than anything a computer can come up with.

Who Killed NanoSessionMo? will be taking place at 4pm in the Meeting Room at Edge-Lit Genre Fiction Festival on Saturday the 7th of September. You can buy tickets to the festival here.

The output of that workshop will be available for free at the Scarlet Ferret indie ebook shop, which also stocks my books.