Would you still love me if my story was a worm?
The cat is out of the bag! I was part of a group of 13 author to write with Google’s Wordcraft (one of the latest generation of large language models). Wordcraft is one of the largest language models, trained with a massive amount of text into predicting the words that will follow or perform a series of tasks the author instructs it to do using prompts with specific parameters.
The story I wrote is called “Worm-Mothers” and it’s as weird as it sounds. 😆 It’s about a world with strange creatures, demanding strange sacrifices.
Each author created a diverse array of stories and engaged with Wordcraft in their own unique ways. For me, the process was more or less having fun and playing around with the tool while I familiarized myself with its interface and what it could do. It turns out in my particular case Wordcraft was used in two ways: 1) For brainstorming purposes, giving me the initial seed of an idea. The weirder the better. 2) For worldbuilding and honing details on the nature of the creatures in my story.
For other writers it was a poetry generator or a research tool, for others is was a problem-solving companion. It was, of course, not perfect as it needs the user to guide it and select the fitting parts to incorporate in a story, but it was still a fascinating experiment with a machine learning writer.
The “Worm-Mothers” got a shout-out at CNET which is pretty cool! I hope you will read and enjoy that little story.
You can read the stories and learn more about Wordcraft here.