Why is AI slop so frustrating to read?

AI writing is frustrating, the argument goes, because it often makes readers work harder to find the point: sentences grow more complicated, wording turns awkward, and important ideas get buried. The thread says these habits conflict with established English-writing principles and reports that a benchmark found AI-generated writing violated those principles much more often than human writing. Claude Code replies were the worst offenders in the comparison, though the thread provides no benchmark rates or detailed testing method in the material presented.

It highlights three recurring problems. First, sentences can hide their main character and action, putting something other than the central subject and verb in the foreground. Second, models string nouns together into unnatural phrases, with examples such as “a label’s awareness entry” and “vision-and-transcript pass.” Third, AI writing often fails to rank its points, producing either sentences that say little or ones crowded with distracting ideas. These flaws matter beyond style: as AI increasingly generates material people encounter online and in work, research, and technical communication, its writing needs to convey ideas clearly to human readers.