Hacker News new | past | comments | ask | show | jobs | submit login

I use Gemini pretty much exclusively for creative writing largely because the long context lets you fit an entire manuscript plus ancillary materials, so it can serve as a solid beta reader, and when you ask it to outline a chapter it is very good at taking the events preceding and following into account. It's hard to overstate the value of having a decent beta reader that can iteratively review your entire work in seconds.

As a side note, I find the way that you interact with a LLM when doing creative writing is generally more important than the model. I have been having great results with LLMs for creative writing since ChatGPT 3.5, in part because I approach the model with a nucleus of a chapter and a concise summary of relevant details, then have it ask me a long list of questions to flesh out details, then when the questions stop being relevant I have have it create a narrative outline or rough draft which I can finish.






Interesting. I think I'm a better editor so I use it as a writer, but it makes sense that it works the other way too for strong writers. Your way might even be better, since evaluating a text is likely easier than constructing a good text (Which is why your process worked even back with 3.5).

I have a horrible time editing my own work. Decision paralysis and what not, but I did have the idea that a good way to practice would be editing the content of LLM generated fictional narratives. I think the point that many are making that LLMs are useful as cognitive aids that augment thinking rather than replacements for thinking. They can be used to train your mind by inspiring thoughts you wouldn't have came up with on your own.



Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: