2
1 Comment

Is AI writing actually making writers worse?

I’ve been testing most major LLMs for writing this year: ChatGPT, Gemini, Claude and Perplexity.

AI is incredible for one thing:
it completely removes the blank page problem.

You open a document, write a prompt, and suddenly ideas appear.

For brainstorming, it’s amazing.

But I’ve started noticing something strange happening online.

The overall quality of writing on the internet seems to be dropping fast.

A lot of content now feels:

generic

repetitive

structurally weak

written without a clear audience in mind

Personally, I spend a lot of time building personas before writing anything.

Understanding:

who the reader is

what problem they have

what level of knowledge they have

Without that, the writing becomes vague.

And AI often amplifies that problem if you use it without a clear direction.

Ironically, something else happens when you actually try to write carefully.

When I structure an argument properly and take time to write clearly, people sometimes assume:

“This must be AI.”

And honestly… it drives me a bit crazy.

Good writing used to signal thinking.
Now it sometimes gets mistaken for automation.

For me the best workflow is hybrid:

1️⃣ define the idea
2️⃣ define the audience (personas)
3️⃣ structure the argument
4️⃣ then use AI to expand or refine

AI is great for starting writing.

But structuring ideas and understanding the reader still requires human thinking.

Curious how other founders here use it.

Do you rely on AI mostly for ideas, structure, or full writing? 🤔

on March 8, 2026
  1. 1

    @Donloub ooooh so painfully on point . I'd say this. It depends, I voted because most use it exactly like you said. Prompting it and use it to generate a piece of content. Even having user personas won't help much...because other people also use user personas and yet content sounds generic.