The UX of Good Reading

blog.michalprzadka.com

3 points·michalwarda··4 comments

Open articleView on HN

> Here is my new rule: If I read anything that feels like slop, I’m unfollowing you.

> …

> It’s not just about excessive emojis or em dashes. Here:

> …

> • Manufactured segue questions connecting sentences that don’t need connection

> …

> • Pompous, repetitive patterns like “This isn’t just X—it’s Y”

Then, later in TFA, the author (an LLM itself, perhaps?) commits those very sins:

> Why are so many people afraid of sounding personal and opinionated? Is it because being opinionated feels risky? Because bland feels “professional”? Because having a point of view means you might be wrong?

> The solution isn’t more polish—it’s more humanity.

Meh, looks more like parroting. Also for word "polish": https://news.ycombinator.com/item?id=42864854 - I do not want AI to "polish" me

did you read the post? because it looks like you did not

hi there, I am the author of this post.

> commits those very sins:

you are wrong, let me explain:

1. There is nothing wrong with rhetorical questions per se - I just feel that LLMs often put them between sentences that do not require connection. I would even argue that any LLM would vote against putting the questions I wrote like this because they feel out of place. But I wanted them there.

2. "The solution isn't more polish—it's more humanity" isn't the pattern I'm referring to. The slop pattern is using "This isn't just X—it's Y" for emphasis reinforcement, like "This isn't just wrong—it's baffling." Mine is a simple contrast statement.

Thanks for the comment though - I'm glad a fellow human read it carefully enough to have an opinion on it! :)

Reply on news.ycombinator.com