Make what you mean legible.
A workspace for writers who want clear feedback without the slop. You write, ask “What do you think?”, and get notes on structure, clarity, and argument — without ghostwriting.
Not used to train AI models
Feedback, not ghostwriting
Control what’s saved & what the AI sees
Example notes.
A blog-argument snippet and the kind of feedback Stet gives.
Draft excerpt
AI makes smart people sound generic because it optimizes for fluency, not meaning. When you hand it a half-formed idea, it fills the gaps with confident glue--transitions, abstractions, plausible-sounding claims. That polish can feel like thinking, but it's often just moving the fog around.
Stet's notes
- Argument: "Fluency, not meaning" is strong--define "meaning" here (truth? warranted inference? intellectual honesty?) in one sentence.
- Clarity: "Confident glue" is vivid; show one concrete example of the kind of empty move you mean.
- Precision: "often" is doing work; either narrow the claim ("in first drafts") or add a quick example to earn it.
- Structure: Add a test the reader can run on their own paragraph to spot "fog-moving."
- Counterpoint: Acknowledge where AI does help (tightening, alternatives) so the claim doesn't read absolute.
- Close: End with a clear takeaway: what to ask for instead (notes on missing steps, not rewrites).