Content, rewritten: everyone has the same writer now.
When every brand drafts with the same models, the draft stops being the work. What the draft is built from becomes the work.
What the old playbook assumed
That writing was the scarce part. A team that could publish more, faster, on more topics would win the category, so content plans were counted in posts per month and judged on whether they shipped on schedule.
What broke
Drafting became close to free for everyone at once, and the drafts came out of the same few models. Ask two of them for a post on the same topic and you get the same shape: the same structure, the same safe claims, the same closing line. More volume now means more of the average.
Why prompting does not fix it
A prompt can ask for a tone. It cannot supply what the model does not have: your data, your cases, the calls you made and why, the things you believe that your competitors do not. Voice is built from those inputs. Without them the model fills the gap with the most probable sentence, which is the one everyone else gets too.
What the work is now
Feeding the machine what only you have, then checking the output before it ships. We check every draft against the tells of machine-default writing and against what the sources already being cited say, and a draft that adds nothing new does not go out.
What replaces the old numbers
Not posts per month. Whether a reader, or a model, can tell a piece is yours without the logo, and whether it says something the pages already being cited do not.
Everything in this guide
- AI Detectors Measure the Wrong Thing
- AI Can Average Your Voice. It Can't Evolve It.
- How to Acquire a Voice in the Agentic World
- Jasper Alternatives (2026): AI Writing Tools Compared — and the Problem None of Them Solve
- How to Make AI Writing Sound Like Your Brand (and Why Prompting Fails)
- Perplexity and Burstiness: The Real Reason AI Writing Is Detectable
- Specific-Sounding, But Still Generic
- The Nine Tells of AI-Generic Writing
- Why Does ChatGPT Sound Generic?
- Why Voice Can't Be Prompted
Questions
Should we stop using AI to write?
No. Use it for the drafting. Keep the inputs, the judgement and the final check with the people who know the business.
Do AI detectors matter?
Less than people think. They measure statistical smoothness, not whether a piece is worth reading. A generic human draft and a generic machine draft fail the same reader.
How do we know if our content sounds generic?
Run a few pieces through the free AI slop detector, then read them next to a competitor's with the names removed. If you cannot tell them apart, neither can a buyer.