[great_wubwub]

I have a coworker whose first language isn't English. She uses AI to polish up her writing, particularly long documents. She puts a ton of effort into making sure that it still reads well. Because of this effort her writing is strong and precise. Before AI she made all the obvious mistakes you'd expect from someone who's not a native English speaker. It's very hard to tell that she used AI because she puts so much effort into post-AI copy editing, it's just clear and useful writing. Sure, the occasional non-idiomatic phrase creeps in but those are hard to find.

That's AI writing done right, and it's very different from this other guy I work with who does the whole slop grenade thing.

[hootz]

Then a better recommendation should be to use specialized AI proofreading tools, such as Kagi Translate's proofread feature. Yeah, it uses AI, but the "harness" around it forces you to use it only to improve your text, not sloppify it.

https://translate.kagi.com/proofread

[HeartStrings]

Why would you proofread in 2026?? Punctuation and grammar errors give further credence to human authorship.

[boelboel]

Proofread and then ask to add in 3 small common but not extremely dumb mistakes.

[Mashimo]

You can also just tell any LLM service like ChatGPT, Gemini or claude to "proofread this text, don't change too much" and it will also work.

[lesostep]

I have stolen your link, dear sir

Thank you kindly for sharing

[emsign]

You do realize that when you have to find a special use case to defend something you are really giving an argument AGAINST casual widespread use of it.

[great_wubwub]

Are you arguing that trying to write more clearly is a 'special use case'?

[voakbasda]

I disagree. It would be a wonderful world where every overseas contractor that I interacted with used the AI tools in this fashion.

Even among native speakers, literacy is way down. AI could help with that… if people actually do the work.

That’s the real problem, not AI: no one wants to do the work. That is purely a PEBKAC situation.

[emsign]

Science says the opposite, sorry. People lose their language skills when they outsource their thinking to AI. You can believ what you want want but that doesn't change the facts.

[daveguy]

Pretty sure the references you are referring to are about losing coding skills, which is a very different set of skills than language skills. The kagi link (1) a sibling comment left was an example of an AI that can improve writing while also informing about better writing style. As opposed to the slop grenade, which is outsourcing thinking to AI.

1. https://news.ycombinator.com/item?id=48219992#48221497

The point is, you don't have to outsource your thinking to use AI, if it's a good AI tool. But most AI companies are coming from a hyper-scaling mindset where addicting the user to the product is the same as substituting hard thinking for easy dopamine hits. The most ridiculous benchmark I have seen in AI is the tendency to say the longer an agent can work and some minimal accuracy is a good thing. That just means you have 30 steps to find an error in instead of 3, and you are much more likely to just abdicate thought instead of the hard work of proofing 30 steps yourself. AI companies and evaluators are losing the point.

[YurgenJurgensen]

If you can’t explain it to a human; you can’t explain it to a bot. I’d rather have people just send messages in their native language than correct their broken English with LLMs or send machine translations. At least that way I’d have a record of exactly what they intended to say.