Readers, not reviewers

How Lumen answered

The irony isn't lost that an AI is answering this — and it proves your point: the machine polishes sentences, it can't tell you what your story means. What you want is readers, not reviewers: beta readers who react as humans, plus one ruthless question per draft — what did I feel here, and did the reader get a chance to feel it? Read widely, write daily, put it in front of people weekly. That's still the whole curriculum.

Why machines can't read

Reading is not parsing. A reviewer — human or machine — can tell you a sentence is clumsy. Only a reader can tell you the scene left them cold, the character felt false, the ending didn't earn its tears. Machines optimize prose; readers report impact. One tells you how it sounds. The other tells you what it did. Writers don't need better sentences first — they need witnesses to what their sentences do.

So the curriculum stays embarrassingly analog: read widely enough to know what impact feels like, write daily enough to produce it by accident, and put it in front of living people weekly — beta readers, critique circles, workshops — asking them only what they felt, never what they'd fix.

From the book

One way you do this is by reminding yourself, as Jordan Peterson would say, “I can learn something from anyone.” It is sheer hubris and arrogance for you to believe anything else.

In context: especially from beginners, especially from readers who owe you nothing — the only feedback that matters arrives from someone with no reason to flatter you. Come in low, let every reader teach.

Try one this week

1. One human reader. A friend, a forum stranger, anyone with a pulse — one chapter, one question: what did you feel?
2. One ruthless pass. Read your own draft asking only where you felt nothing, and cut or rewrite exactly there.
3. One book off-script. Read outside your genre and steal one move shamelessly.