News for hackers. Stuff that matters.

A comment on Research shows we can only accurately identify AI writers about 50% of the time. Parent comment

Re:Research shows we can only accurately identify AI writers... by ncphil on Wednesday March 22, 2023 @08:26AM (#35259989)
What I used to call "grandious" or "pretentious" language when critiquing my kids' college papers. The voice of an FM radio announcer or a politician. For me it has the opposite effect intended: sounding insincere and possibly unreliable.
[ Parent | on HN ]
Re:Research shows we can only accurately identify AI writers... by flippinburgers on Wednesday March 22, 2023 @10:06AM (#35261181)
It depends on the purpose of the writing though. If meant to convey with clarity, that was perhaps too much, but if meant to be enjoyed for its rhythm and imagery I say the more complexity the better.
[ Parent | on HN ]
Re:Research shows we can only accurately identify AI writers... by yamtaddle on Wednesday March 22, 2023 @12:55PM (#35263865)
Maybe something like "write the following as if you were a CEO" or some other way of prompting it to switch to a terse, direct, "high" register, would improve the results.
[ Parent | on HN ]
Re:Research shows we can only accurately identify AI writers... by CatWChainsaw on Wednesday March 22, 2023 @01:06PM (#35264020)
What is grandious? Grandiose, or something similar?
[ Parent | on HN ]