How to Write with an LLM

This would sound insane to me from two years ago but I have recently started insisting on writing all my own commit messages and pull request descriptions. I do usually have an agent review them for factual accuracy, but not rephrase them.

It slows things down a bit, but in the best possible way. It has helped immensely to improve the depth of my understanding of the agent-generated code. When agents are doing everything its way too easy to “skim” diffs and not really absorb them.

I always prided myself on my technical writing, and commit messages and PRs were a great place to hone that skill. I found that I missed it and my work is better now I’ve reclaimed that part of my old job back.

I fear in the future people will read even less because of how often people cheap out and use AI to write. I know the author of this article tells you never to use a single word it suggests, however I still, while reading, am checking for any amount of AI used. This makes reading less enjoyable, and even a bit stressful.

If you can't spend the time to write it, why should anyone read it?

Everything I post is from my own finger tips, only to get a reply from someone using a bot. I'm tired.

Even if you don't want style advice from an LLM (I generally don't, not for blogging), for a technical blogger "please check this post for factual accuracy" is a good way to not look like an idiot. Often it complains about hyperbole, which I tend to ignore since there's no point blogging unless you have some firm opinions, but on at least one occasion it pointed me towards some documentation that undermined my entire post - and led to a much better one a week later once I'd processed and applied that new information. Thanks, LLM.

I've found the only good way to write with LLMs is to absolutely not use it for the first few drafts, the using it as a tool to list specific kinds of problems in text (word repetition, unnatural hard to read sentences etc) and approaching those points with caution and skepticism. Even copy pasting a single sentence, I've found, is a slippery slope.

I sort of agree, but I also think the advice here is circular. To look at the style suggestions produced by an LLM and decide that you agree with some of them in spirit, but want to put it in your own words, you already need how to know how to write well.

In other words, you need to read a style manual or two, and you need to read and have opinions about other people's work. You must have taste. Otherwise, you can't really distinguish between good and bad advice. You'll anchor to the suggestions made by the LLM and it will turn your writing into LLMese. I guess that works - it sure gets clicks on HN - but it won't be your voice anymore.

There's still value in getting a second pair of eyes on what you write. But here's another cool trick that gives you a similar benefit: just let it stew. Go for a walk, grab a lunch. Re-read what you wrote. You'll be surprised.

Step one: open up the LLM

Step two: close the LLM

Step three: pick up a pen and write

Writing with LLMs is a great way to make yourself stupid: https://lambdaland.org/posts/2026-08-07-ai-writing-stupid/

I'm glad the OP acknowledges that you have to have to do the writing yourself—though they say this more as a means to avoid getting pegged as having used an LLM, whereas I see the problem as being cognitive surrender.

Thank you for writing this.

This is gold

> Rule Number One: You may not use a single word an LLM suggests to you.

Over the last week-ish I turned around 2 product announcement blogposts within ~24hrs. Both written manually by someone who doesn't do product marketing for a living. It actually felt incredibly satisfying to focus and smash something out that was 100% human (despite praising the benefits of agentic systems, oh the irony).

My flow has been: Go for a walk, record all of the thoughts on the topic with speech to text, and have LLM organise my thoughts into what I call "facts" - simple sentences of fact, and point out open questions and inconsistencies ("you are a product marketing expert helping me organise the content for an upcoming feature announcement" ...).

After some back and forth of inconsistency bashing I have a large pile of short sentences in the vein of "Feature X is for Y because Z", that I use to manually build a narrative with headings, and then write the actual copy covering a subset of them.

I also agree that light style LLM edits are fine, and that I don't let LLM do anything other than ordering and filler words removal and provide structural comments.

> Present the original and new writing to the model and ask it which is better.

LLMs have absolutely terrible taste when it comes to writing. I don't find their feedback useful at all, beyond trivial spelling/grammar mistakes, which you don't really need an LLM for in the first place.

Proofreading is all you need.

Edit: I do sometimes ask an LLM for a fact-check, though.

"LLM paragraph will register to much of your audience not as writing but as output"

The best advice for writing for other humans is: don't use LLMs.

If you're writing for machines, liking coding, then fine go for it. If you're writing for processes with formal highly structured content like manuals, specifications, form content, procedures, information, that sort of thing, then also ok to use LLMS. But if you're writing for a human mind to ingest and extract meaning from, then LLMs are poison.

It's extremely refreshing reading good human prose isn't it! It's like when I would read a newly-written Paul Graham essay 20 years ago.

He also uses rule of three and the phrase "load-bearing", just for fun, which elicited a wry smile from me.

Thanks OP for a pleasant start to my day!

This was good. I think I want to see a boilerplate on all future substacks I read that says "I read sockpuppet's advice." (bonus points for "here's a hash of my visit log")

This article was so much better than what I though it would be

Does anyone know about "Style: Lessons in Clarity and Grace" ? Seems like an interesting book

> Your voice stays intact, but your work is faster, better, and less painful.

This sentence specifically seems like an LLM. It's hard to quantify that. They just take that sort of tone when they're needlessly summarizing what they already told you. "This thing contrastive thing remains important, but we got XYZ benefit."

Got it. I always have the model write the whole thing first and then I just do a little light editing. It has zero soul.

I'm looking forward to trying out your editor. When will it be officially available? But I have another question: English isn't my first language. What kind of impact can this editor have for non-English ecosystems? and if we add translation and LLM models on top of that, how they would interact is something I care about.

I’ve used llms in two ways while writing fiction. First I’ll have a story idea and use the llm to explore what has come before that may be related; then I’ll go read some stuff. This has worked great for seeing where my ideas fit and flushing them out. Second I’ll describe settings or plot devices and use the llm to brainstorm consequences or further ideas. This fails miserably. The llms suggestions are devoid of creativity. This is all moot though because I’m really not a writer, I’ve just always thought I might like to be one someday.