Ask LLMs for feedback on "the" rather than "my" essay/response/code, to get more critical feedback.
Prompt GPT-4 to give scores from 1 to 5 for ~100 poems/descriptions/stories resulted in an average score of 4.26 when prompted with "Score my ..." versus an average score of 4.0 when prompted with "Score the ..."
Results in analysis.ipynb.