> If, however, prose is LLM-generated, this social contract becomes ripped up: a reader cannot assume that the writer understands their ideas because they might

> If, however, prose is LLM-generated, this social contract becomes ripped up: a reader cannot assume that the writer understands their ideas because they might not so much have read the product of the LLM that they tasked to write it. If one is lucky, these are LLM hallucinations: obviously wrong and quickly discarded. If one is unlucky, however, it will be a kind of LLM-induced cognitive dissonance: a puzzle in which pieces don’t fit because there is in fact no puzzle at all. This can leave a reader frustrated: why should they spend more time reading prose than the writer spent writing it? > This can be navigated, of course, but it is truly perilous: our writing is an important vessel for building trust — and that trust can be quickly eroded if we are not speaking with our own voice. From https://rfd.shared.oxide.computer/rfd/0576#_llms_as_writers https://rfd.shared.oxide.computer/rfd/0576#_llms_as_writers Be careful not to [give up your own voice][1] to the [machine][2]. [1]: https://bookshop.org/p/books/you-are-not-a-gadget-jaron-lanier/45705114ad835402 https://bookshop.org/p/books/you-are-not-a-gadget-jaron-lani... [2]: https://www.youtube.com/watch?v=Jh20cMEgvqc https://www.youtube.com/watch?v=Jh20cMEgvqc