How to spot AI science slop

airhangerf15

Reads kinda dumb, weird and painful. > Note: Some people write the text, then format it with Claude. In my opinion this isn’t AI slop, whereas AI-generated text that’s manually formatted (also very common) would be AI slop. I hate it when people do this. I had a manager that fed all our stories through Opus 4.x a year ago and they could easily turn from well defined human-readable requirements to a massive mess of every possible error condition and HTTP code. When you write, you are communicating one human to another. You are organizing your thoughts. The fact that all of us CAN read and write is only a few centuries old. 1,000 years ago, literacy rates were very low (I've never found a good number on this. Most papers claim it was 2% ~ 5%. I think it was probably closer to 10%, and more than just merchants who needed to track inventor. But a lot of those wax tables and old paper has probably all disintegrated). I have a feeling the author AI-assisted this. I'd rather read something real, grammar mistakes and all. I realize it may matter less and less as more people start adopting the speaking patterns of their favorite models. "That's a great idea..." and "You're absolutely right to push back on that..." Generative text use to be really painful for me to read. As I use it more in code, I'm getting more use to it, and I'm not sure if I entirely like it, even when it's helpful.

urig

TLDR - random person shares their gut feeling about how to spot slop. Not useful.

cyanydeez

Did openai or anthropic publish it? There, O(1) algorithm confirmed.