Hacker News Reader

Comment by wdutch | Hacker News Reader

I think you're right. Since LLMs went mainstream, I've seen a lot of my colleagues' presentations and thought "was this written by ChatGPT?" but I've come to wonder if it's just given me the frame of mind to identify low-effort slop that lacks any original insight but uses all the right sorts of words and phrases, regardless of if it was authored by a human or not.

View on HN

My hope is that equivocating waffle will look so much like ChatGPT that humans will have to write clearly and precisely to differentiate ourselves and we can put this horrible era of essay writing style behind us.

Though I’m starting to think that AI might improve faster than us so there might only be a diminishing margin of opportunity left to do this.

We’ll have AI tools to take our bullet-points and expand them into prose (crappy now, eventually beautiful prose). Then we’ll use AI tools to summarize that prose into bullet points.

Eventually we’ll realize we can just send the bullet points and generate the prose on the receiving side. This will be great because most of the time the AI’s will be able to say “let’s just be a nop.”

Reminds me of the SMBC comic https://www.smbc-comics.com/?id=3576 that explores the idea further.

I am less optimistic than the comic.

It's both. Especially the out of context tinfoil rage response. They always existed. But now it's so common to see some totally benign article about pizza, and the top comment is "don't let them tell you not to remember in 1995 when US implanted radios in Syrian babies".

The algorithm is being trained solely for engagement. It is horrifying.

This sounds completely unrelated. If someone is really leaving a comment like that, it has nothing to do with the article and everything to do with the way weirdos engage with the internet.

The Slop is the wordy vapid garbage that maximizes SEO.

Parent is right though: AI slop in an article is to maximize SEO. AI slop in a comment is a weird jumble of implausible claims to maximize engagement. I see both on a daily basis now.

I've had similar false positive experiences where I swore some content had to be some form of LLM generated content, until I discovered the source was just poorly done or even just text from some older writing that "sounded" wonky but was more of a product of its time (like a 1940s newsreel).

This. Bad AI output is indistinguishable from bad human output. It's literally the same exact shit.

With AI, there can be more content, produced faster and probably more cheaply, that is tailored to individual users.

Reply on news.ycombinator.com