Hacker News Reader

Comment by seanmcdirmid | Hacker News Reader

There is a lot of curation and editing that go into AI generated content. Like if you just use an AI to write from a prompt, you’ll just get slop (vibe writing), but if you plan it, review the bullet points, delete and add points that make the narrative stronger, have it sketch some writing out after you are happy with what it is covering, then rewrite that because LLMs have absolutely no sense for word rhythm.

View on HN

This is logical, but it’s also a lemon market. You don’t know the author, editor and publisher did anything of the sort, unless you’re the sort of person who doesn’t need the book. But if you do see an LLMism, it’s a strong sign of cost-cutting.

If done right you couldn’t tell either way, and you probably wouldn’t care. The point of writing is to get a message out to the reader as efficiently as possible, don’t waste their time. The writer ensures that, AI is just a tool at that point. The problem with vibe writing with AI is that AI has no sense about wasting the reader’s time.

We should just be judging writing by its quality. There were always bad writers who wrote like LLMs or worse. It takes a lot of practice to write well.

Nothing has really changed: bad writers using AI will still be bad writers. Good writers using AI, with some effort put into learning how to use it, can benefit a lot from it in terms of (at least) writing good content faster.

Nice, now add slackers and non-writers to the equation.

Nothing changes, bad writing is just more uniform.

Look, you're must not saying anything. "If done right ... you probably wouldn’t care". Yes, correct. It's done bad, so I don't like it, and the visible slack is a sign of slack. Would you like to push back in some way for some reason? What's the call to action here?

My pushback is that AI can be used for writing, and the results will depend on the person using it. It’s like a type writer doesn’t make anyone a great writer either, but it improve their legibility at least.

LLMisms are almost universally results of using a narrow set of mainstream offerings from Anthropic and Openai. Even Muse Spark has style much less "sloppy" than Claude et al, let alone Chinese LLMs. I mean yes, they have their own tics, but if you don know them you won't even notice.

But I do know them, and so do you, and they’re signs of cost cutting partly for the reason you mention. They’re the careless, basic “no-brainer” choice of models, plus a lack of editing, or a very light “edit” in which obvious, even famous tics are left in. (Many real editors would clean up writers’ tics.) So it’s possible care was put into it, but tics visibility is evidence against it and not evidence for it. If your text does pass and educated readers don’t see it, it’s a consequence of care and intent, which is what we’re really asking for.

Reply on news.ycombinator.com