Von Sprachnotizen zu veröffentlichten Posts: ein Claude-Workflow

Frequently asked

Why start with voice instead of typing?

You think faster out loud. A voice note captures the whole argument while it’s hot and complete; the blank-page version usually loses the spine somewhere around the third paragraph. Voice-first removes the friction that kills most drafts.

How do I get Claude to keep my voice?

Feed it two examples of past posts you love, name the moves you want it to keep (sentence rhythm, specific examples, the one opinion), and tell it explicitly what not to do. Without examples, you get blander output; with them, you get yours-but-tidier.

What goes wrong if I publish the Claude output directly?

It’ll be fluent and dull. The model sands edges; voice lives in the edges. Your final pass — putting back the specific names, the rhythmic variation, the one strong opinion — is where the post actually becomes yours instead of generic AI tidy.

How long does this take end-to-end?

Ten minutes voice, thirty minutes drafting and refining, twenty minutes editing and publishing. About an hour for a sharp post, several times what typing-from-scratch usually takes. The compounding benefit is shipping more often without burnout.