With a sufficiently effective language model it would be fairly easy to bridge t... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		Veedrac on July 27, 2020 \| parent \| context \| favorite \| on: Are we in an AI Overhang? With a sufficiently effective language model it would be fairly easy to bridge this model, by letting the text direct humans on the other side. `Hypothesis: <AI writes this> Results: <human observations> <repeat>`

endgame on July 27, 2020 [–]

That seems like a _very bad_ habit to get humans into.

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