Only 61% of the papers themselves were mostly human written.
Reviews
Papers
pull down to refresh
71 sats \ 1 reply \ @south_korea_ln 6h
How can one even fully generate a paper using AI. Those must be so bland and/or lack much novelty. Or i must suck at prompting.
Currently, i write the first draft myself, and then parse the paragraphs one by one to improve the flow. But still needs heavy personal editing of those improvements to keep my intended nuances.
reply
50 sats \ 0 replies \ @gmd 49m
most human written scientific papers are already so bland and hard to read. it takes some work for most people to go from more narrative drive human writing to scientific style writing.
reply
150 sats \ 2 replies \ @SimpleStacker 12h
I strongly suspect someone used AI to review my paper recently.
Very annoying.
reply
0 sats \ 0 replies \ @south_korea_ln 6h
Yeah same experience. Lots of useful feedback one the one hand, but some very weird feedback on the other hand that an actual expert would never ask. Paradoxically, it made it easier to write the rebuttal, but i would have expected better from Springer referees.
reply
0 sats \ 0 replies \ @optimism 8h
As in did they C&P the output?
reply