CS professor: if you never learned to write well yourself, how can you judge LLM-assisted papers?

CSProfKGD · x · 2026-10-28

A CS professor recalls his PhD supervisor's advice: read the conference version of a paper before the journal version, since journal versions often become overly formal and lose the original's intuition.

He draws a parallel to today's flood of LLM-generated writing — little intuition, excessive formality, no regard for succinctness, with some papers and blogs going on for pages without saying much.

He isn't against LLMs for writing, but worries the next generation may never develop the skills and judgment to steer them effectively: 'How can you recognize good scientific writing if you've never learned to produce it yourself?'

Original post →

CS professor: if you never learned to write well yourself, how can you judge LLM-assisted papers?

More from AGI Musings