Agree that it’s the ‘crafting’ part that is what matters, and I don’t think we can say a writer/DM is going to explicitly be thinking about all the details of the story at each turn. From the examples.. well as a side effect of doing AI research is that you can’t help but read the text of the story and see that the “thoughts” about it are just picking details in a way that even SOTA ML systems have a problem with. They don’t read as actual notes about the process. Perhaps there needs to be a request for samples, with a 30k word limit (so no one invests too much time in something that might not be used), and a focus on capturing the process of writing a story as the plot unfolds.
Agree that it’s the ‘crafting’ part that is what matters, and I don’t think we can say a writer/DM is going to explicitly be thinking about all the details of the story at each turn. From the examples.. well as a side effect of doing AI research is that you can’t help but read the text of the story and see that the “thoughts” about it are just picking details in a way that even SOTA ML systems have a problem with. They don’t read as actual notes about the process. Perhaps there needs to be a request for samples, with a 30k word limit (so no one invests too much time in something that might not be used), and a focus on capturing the process of writing a story as the plot unfolds.