What Lives Inside Second Draft Notes

AI gives me speed. It catches grammar slips and the repetitions I stop noticing after the third edit.

I’m not giving that up.

But something else started happening. The texts came back cleaner, without the small errors I was tired of making. I liked them. That was the uncomfortable part. They were still mine, but arranged too neatly. Like my house after professional cleaning: beautiful, unmistakably mine, and still waiting for my hand. The blanket put where I actually use it. The chair turned back toward the window. 

Something small returned to its place. That is where these notes begin.

Using less AI didn’t fix the shift.

It changed when I brought in something else. Not a better-engineered prompt, and not another round of settings. 

Human material, shaped over centuries. It would be almost foolish not to use it now.

Myths, figures, old stories, seasonal rituals. Not as decoration, but as human shorthand. One symbolic reference can save you a page of explanation. It tells the machine what kind of thought you mean before you have to turn it into instructions. The result gets there faster, with less flattening along the way.

Kitsune, who shifts form without losing intelligence. Ariadne, holding the thread when everything else dissolves. I keep coming back to Scheherazade, surviving by finding the next sentence.

A figure can change the tone in one move. Ask the same sentence to speak like a British king, then like SpongeBob, and you'll feel it immediately.

A season does something similar, only more quietly. A Christmas text is not written from the same place as a text imagined from a beach in July. The reader feels that too.

Then there are the patterns that appear under pressure.

They are not personality types, and not exactly flaws. More like small shifts in how we try to be received.

I built my first diagnostic product around them after years of working with language and the way people present themselves. I also know them from the inside: the moment when wanting to be understood turns into explaining too much, or when credibility becomes safer than clarity.

And then there is the sentence itself.

Just one line where something has shifted.

A sentence that explains too much may need one word removed. A sentence that hides may need a subject brought back.

Other times, the work is not to correct the sentence at all. It is to leave the slightly imperfect version because it says the thing the way you meant it.

Not fixing what no one has accused you of yet.

Sometimes the whole correction is smaller than the panic around it.

None of this is theoretical for me. 

The tools I'm describing are the ones I work with daily for speed and editing. 

And more recently, for something else: catching the moment when the text starts sounding correct and stops sounding like me. 

The difference isn't in which tools you use. It's in whether you notice what they're doing to the sentence.

That is what became Second Draft Notes.

A closed layer inside The Second Draft, with notes on:

Symbolic lenses — myths, figures, and cultural images used to see what plain explanation misses.
Seasonal notes — how time, ritual, and collective atmosphere change what a sentence can carry.
Distortion patterns — the small ways writing shifts under pressure.
Selfhood modes — authority, visibility, self-trust, and how they shape the voice before a sentence is written.
Sentence notes — close readings of the small decisions that change everything.

For people who use AI, but don’t want to disappear into it.
 

For people who care about voice, identity, and the strange gap between sounding better and sounding more like themselves.

Subscribe if you want to step inside that layer.