[fshr]

I think you would be better off having the LLM help you build up the plot with high level chapter descriptions and then have it dig into each chapter or arc. Or start by giving it the beats before you ask it for help with specifics. That'd be better at keeping it on rails.

[famouswaffles]

I don't disagree. Like with almost anything else involving LLMs, getting hands on produces better results but because in this instance, i much prefer to be the reader than the author or editor, it's really important to me that a LLM is capable of pacing long form writing properly on its own.

[saberience]

Random question, if you don't care about being a creator yourself, why do you even want to read long form writing written by an LLM? There are literally 10000s of actual human written books out there all of them better than anything an LLM can write, why not read them?

[tluyben2]

That was what I tried on the train [0] a few weeks ago. I used Groq to get something very fast to see if it would work at least somewhat. It gives you a PDF in the end. Plugging in a better model gave much better results (still not really readable if you actually try to; at a glance it's convincing though), however, it was so slow that testing what kind of impossible. Cannot really have things done in parallel either because it does need to know what it pushed out before, at least the summary of it.

[0] https://github.com/tluyben/bad-writer