FavouriteShapes comments on Let's take this offline

[–] FavouriteShapes@sh.itjust.works 65 points 2 months ago (5 children)

This is a ChatGpT prompt followed by a "make it sound more human" prompt 3-5 times

permalink

fedilink

source

hideshow 5 child comments

[–] very_well_lost@lemmy.world 19 points 2 months ago (2 children)

Because here's the thing: it still reads like slop

permalink

fedilink

source

parent

hideshow 2 child comments

[–] RustySharp@programming.dev 15 points 2 months ago (1 child)

We've been writing slop long before the slopmachines were a thing. Where do people think corpospeak came from?

permalink

fedilink

source

parent

hideshow 1 child comment

[–] very_well_lost@lemmy.world 20 points 2 months ago*

Unironically LinkedIn has probably seen the smallest decline in quality since LLMs hit the scene. Hard to go any lower when you're already at rock bottom.

permalink

fedilink

source

parent

[–] Hungry111@lemmy.today 11 points 2 months ago (1 child)

Genuinely, how can someone recognize AI in a text like this?

permalink

fedilink

source

parent

hideshow 1 child comment

[–] slaacaa@lemmy.world 7 points 2 months ago

You get a gut feeling if you see enough slop.

Examples: It’s not X, it’s Y. 2 separated short lines near the end. These things are repeated in all LLM generated “viral” linkedin post

permalink

fedilink

source

parent