Blog

Latest from the blog

Editing an AI-written product announcement before it goes out

AI drafts a confident, well-structured announcement. The problem is that announcements require the one thing AI can't supply: what actually happened.

What AI gets wrong when writing about a technical product

AI writes confidently about how technical products work. The confidence is real. The accuracy isn't always.

Editing AI-written onboarding emails that users actually ignore

Onboarding emails written by AI tend to explain the product instead of prompting the next action. That's why users don't open the second one.

What to look for when reviewing AI-generated marketing copy

The problems in AI marketing copy aren't random. They cluster in the same places every time.

Turning an AI-written case study into something a prospect will believe

AI case studies read like press releases. They describe outcomes without evidence, success without conflict, and results that could belong to anyone. Here's what to fix.

Making AI-generated social media posts sound like you wrote them

AI posts are grammatically fine and completely unmemorable. What's missing is the thing only you know — the detail that makes it specific.

How to review a ChatGPT-written blog post before publishing (checklist)

AI drafts look fine on the surface. The errors are invisible unless you know what right looks like.

Why editing feels more natural out loud

Most writing tools are built for drafting. Editing is different — and voice fits it better than typing does.

Adding your own voice to an AI-written blog post

AI drafts are confident and generic by default. The fix isn't rewriting — it's knowing exactly where your voice is missing.

Building Redraft: editing as a precision tool

Why I built a voice-first editor for the specific problem of correcting text you didn't write.

What to fix in an AI-generated product description that sounds like every other product

The problem with AI product descriptions isn't inaccuracy. It's that they describe a category, not a product.

Reviewing an AI-drafted investor update before sending

AI investor updates hedge everything and sound like corporate filings. The only fix is putting back the actual story of what happened.

What an AI-generated FAQ gets wrong — and how to fix it

AI FAQs answer questions nobody asked. The fix is putting back the questions your actual users are confused about.

Editing an AI-written cold email so it actually gets replies

AI cold emails sound like AI cold emails. The fix isn't a better prompt — it's putting yourself back in.