Claudefishing: WTF Just Happened to Substack
Substack Can Now Tell Your Readers a Robot Wrote This
Four days ago, the CEO of Substack sat down and wrote something that sounded less like a product update and more like a manifesto.
He invented a new word to describe what he was declaring war on. And he shipped a tool that puts a lie detector in the hands of every reader you have.
The word he invented was Claudefishing.
And by the time you finish this issue, you’ll understand exactly what it is, who built the machine behind it, how that machine works, whether it can actually hurt you - and how to make sure your real, human writing never gets caught in the net.
Let’s take it in order. No panic. Just the accurate version.
By the end of this issue, you’ll know:
What the Substack CEO actually said in his announcement post
What Claudefishing means, in plain English
Who Pangram is and why Substack picked them
How the detection technology actually works under the hood
Whether Substack is going to take action against your AI text
How to keep genuinely human writing from getting flagged as AI
And at the end, a gift: The Human Signal Checklist - the pre-publish pass that keeps real human writing from reading like a machine. Notion-ready. Waiting for you after the walkthrough.
