essay scoring

Instead of using GPT4 for Essay Architecture scoring, I should look into making my own weighted models. With around 200 examples, and a detailed breakdown of each, it will have a deeper and more intuitive understanding of the concepts.

Look into the automatic scoring systems they use for English classes. Why do they suck? What are their limitations?

I went into Susan Sontag’s “Against Interpretation” thinking we would disagree, but her take is more nuanced. She’s against interpreting an artist’s intentions to make it fit your own narrative; but she’s for analysis of the thing itself (it’s form and the sensation it evokes). By the end, she pleas for a “dictionary of forms,” which feels like the

Let me know if you want a 3,000 word report on one of your essays! Can break it down over my 27 patterns and score it out of 135 points. It will give you insight into your strengths and areas to work on. Looking for feedback on my feedback, and how to structure this information to be digestible and actionable.

The scoring system for Essay Architecture serves as: 1) a checklist in your own editing process, 2) as a way to analyze the overall quality of a draft, 3) as a way to track progress over time.

There’s an important subtlety in posting essay scores. First I have to quote and celebrate the article so a reader gets the jist. Then I’ll use 3 sections to break down idea, form, and voice. The score will be present (both in the thumbnail, and in a granular heat map). It’ll be something like Pitchfork, BUT, without snark or attitude, and all abou