Medium’s New Tragic AI Policy

Navigating the Robo-Rumble

Esther Jordan
3 min readApr 15, 2024
truman show like landscape, blue retro chair, serene flower landscape, wizard of oz like landscape, dillusion
Photo by Fiona on Unsplash

Alright, folks, let’s talk about Medium’s latest bombshell: they’re cracking down on anything that smells like AI. Yup, if your piece looks a bit too robot-y, they’re yanking the monetization. But here’s the kicker — most AI detectors are just catching predictive text, not actual AI.

Let’s unpack this. AI language models? They’re like our digital minions, trained by engineers to talk and act like us. Without us humans guiding them, AI would be as lost as a cat in a dog park. AI is based off of human language models. Meaning, a bunch of engineers asked AI to study all kinds of digital content (stories, blogs, historical pieces, etc) and said, “Write like this.”

So, even if Medium runs an AI detector, all it’s picking up on is AI writing like humans. Your writing could be flagged as AI if it sounds too “professional” or formal. You can pretty much only get around AI detectors by writing in millineial or informal tone. If you write an entire original piece in formal/professional tone all by yourself (no AI assistance), it may get demonetized if the AI detector and Medium deem necessary.

Only writing in millennial and informal tone is, however, super annoying for most readers.

--

--

Esther Jordan

Quantumplation. Legal writer. SEO marketer, copy writer & editor of 15 years. High fashion model. Former children and adult special-needs counselor. Risk taker.