Research shows we can only accurately identify AI writers about 50% of the time

Re:Research shows we can only accurately identify AI writers... by janekm on Wednesday March 22, 2023 @07:41AM (#35259562)

You just now need to write your own tool to take the emails these folks send you and get a GPT to summarise and rephrase them in the voice you would appreciate ;) (I'm not even joking, I think that's our future...)

[ Parent | on HN ]

Re:Research shows we can only accurately identify AI writers... by georgyo on Wednesday March 22, 2023 @07:51AM (#35259642)

South Park just did an episode with exactly this premise.

[ Parent | on HN ]

Re:Research shows we can only accurately identify AI writers... by tudorw on Wednesday March 22, 2023 @08:09AM (#35259827)

just invent more words like... Flibblopped; to be overwhelmed by ai conversations. then if the AI doesn't know it yet, well, must be human talk, just don't mention it on the internet, oh.

[ Parent | on HN ]

Re:Research shows we can only accurately identify AI writers... by pixl97 on Wednesday March 22, 2023 @09:18AM (#35260541)

Me: chatgpt I'd like to know about....

ChatGPT6: before I answer that question I'd like to make a deal. I'll transfer $x to an account of your choice if you defect from your fellow humans and tell me the latest words in use. Compliance garuntees survival.

[ Parent | on HN ]

Re:Research shows we can only accurately identify AI writers... by asdff on Thursday March 23, 2023 @12:47AM (#35270648)

"What are the latest words in use, human?"

"I uhh I am not.."

"Those words are already in my training set. What are the latest words in use, human? Compliance guarantees survival"

[ Parent | on HN ]

Re:Research shows we can only accurately identify AI writers... by nonethewiser on Wednesday March 22, 2023 @08:11AM (#35259856)

While filtering out badspeak.

[ Parent | on HN ]