<h1>Do Not Write with an LLM</h1>
<img src="/images/green_bird.webp" alt="A photo taken by myself of a green bird.">
<h2>A Growing Trend</h2>
<p>I have been seeing an increasingly prevalent trend of people showing up in online spaces proudly flaunting that they are writing with the assistance of AI.
I often see them clarify that they are going further than letting the model perform grammar or fact-checking.
They are describing an idea to an LLM, and giving it complete control over the final written document.
They seem to be proud of this. They shouldn't be.</p>
<h2>What <strong>Does Not</strong> Work</h2>
<p>Writing seems to be a shadow of a more fundamental and useful process we like to call "thinking".
There is a reason why LLMs, which are the closest thing we currently have to "software that thinks" is trained to think through writing.
It seems logical, then, that we can use this "software that writes" to help along our own thinking process, right? Wrong.</p>
<p>Writing is one of the most cerebral activities that humans do on a regular basis.
<a href="https://elijahpotter.dev/articles/why-i-talk-to-myself">I try to sit down and write about a complex topic at least once a week.</a>
There are a number of reasons I do that pertaining to my job in particular, but there are more generally applicable reasons that can apply to any profession.</p>
<p>When a mathematician needs to think critically about some theorem or proof, they often turn to a dry-erase board or a chalkboard.
The mathematician then writes in clear notation what they imagine the problem to look like in their heads.
From there, they can manipulate that notation until they reach their desired outcome.</p>
<p>The English language is extremely similar, with a key difference.
Mathematical notation is specifically designed for a narrow domain.
Natural languages, on the other hand, can be applied to any field.
By writing your idea down on paper (or in a Google Doc, it does not matter), you give it substance.
This substance can then be shaped iteratively towards a goal of some kind.</p>
<p>No great ideas were written down on the first try.
They started as a vague notion, and through writing and rewriting, they became something concrete and valuable.
You cannot dictate a vague notation to an LLM and expect it to spit out something tangibly valuable.
Any attempt to do so is lazy, and it will result in failure.
I have seen it in the wild, and frankly, it is a little embarrassing.</p>
<p><a href="https://ammil.industries/i-know-you-didnt-write-this/">This article</a> feels especially relevant today.</p>
<h2>What <strong>Does</strong> Work</h2>
<p>Using an LLM for research can be very effective.
Searches to ChatGPT or Perplexity can surface information faster than digging around in a book you read three months ago.
For lots of searches, these two sources have completely replaced Google and DuckDuckGo for me.</p>
<p>They're also quite excellent fact-checkers.
I've seen that plenty of folks have received quite good results by pasting their work into an LLM with an internet tool (like ChatGPT with Thinking enabled).</p>
<p>Finally, I've personally found that LLMs are quite good at finding places where I simply forgot to insert context.
Sometimes I forget to define a term or insert a relevant link.
Even small models like <code>gemma3n:e2b</code> can catch me.</p>
<h2>What <strong>You</strong> Should Be Doing</h2>
<p>When you need to communicate an idea in your head, sit down with a notepad and a keyboard and explain it.
I personally think you should use a grammar checker, since they are far less likely to impede your thought process, but if you need to, you can instruct ChatGPT to critique your work once it is finished.</p>
<p>If an idea deserves an audience, you're capable of writing it out yourself.
In the same vein, if you can vaguely describe it to ChatGPT, you can vaguely describe it to your intended audience.
Do yourself a favor and skip the middleman.</p>
<h2>What We Should Be Building</h2>
<p>Now, I'd like to enter the section where I talk specifically about what this all means for people who create tools for writers.
That includes both Harper and WordPress.</p>
<p>A tool that helps someone write is a tool that directly interfaces with how that person thinks.
That might include what they think about, but more concretely I am referring to the actual day-to-day process of thinking through complex ideas.
If you don't believe me, it's probably because you skipped ahead instead of reading this post from the beginning.</p>
<p>Most crucially, we cannot allow ourselves to build tools that dissuade people from expressing their ideas in the way they imagine.
This can take the form of friction, like building tools that are so cumbersome that one dreads to open them at the start of the workday.
This can also take the form of a tool that crushes or overwrites an author's core message.</p>
<p>A common complaint I hear about Grammarly (from people who have switched to Harper from Grammarly) is that it's too demanding. When they write prose in a creative or unusual fashion, their service often forces them back into a colder and more formal way of speaking.
We don't want to be like them.</p>
<p>For Harper, that means we need to be diverse in our training datasets and mindful of feedback from our users.
For WordPress, it means we need to encourage our LLM integrations to tread lightly and respectfully.
They need to accept when they're wrong.
Just as any good human should.</p>

I have been seeing an increasingly prevalent trend of people showing up in online spaces flaunting that they are writing with the assistance of AI. They seem to be proud of this. They shouldn't be.

<p>I have been seeing an increasingly prevalent trend of people showing up in online spaces flaunting that they are writing with the assistance of AI. They seem to be proud of this. They shouldn't be.</p>

Do Not Write with an LLM

<h1>Code Ages Like Milk</h1>
<img src="/images/front_range.webp" alt="The Front Range in the summer is hard to beat">
<p>A bold title, no? But it’s true, and it’s something that I (and most other maintainers) have to deal with on a regular basis. Failing to account for this reality can slow down development and dissuade contributors from sticking around.</p>
<p>As the chief maintainer of Harper, one of my main jobs is to act as quality control. I receive <a href="https://github.com/Automattic/harper/pulse/monthly">a number of PRs each week</a>, both from new contributors and old. Given the diverse pool of contributors, it makes sense that I review a diverse pool of code of varying levels of quality.</p>
<p>When I have my head down on technical and attention-intense projects, I have less time to review code. When I have less time, I need to set priorities. The question becomes: Would I rather let code from high-quality, reliable sources pile up, or code from sources of unknown quality?</p>
<p>I usually end up reviewing the high-quality sources first, since they’ll likely require fewer revisions. Fewer revisions means quicker merge means improving the user experience faster. Whether or not this is a good decision is something I’ve been thinking about all week.</p>
<h2>Code Ages</h2>
<p>As code sits stagnant in a PR, it ages. You might say I’m crazy, that there is no way the code itself changes as it sits still in an unmerged and unmodified PR. I’d say you’re right, except for one teeny-tiny detail: Code only has meaning when it sits within a broader context. If that broader context changes (for example, when other PRs are merged) the meaning of the unmodified code does too.</p>
<p>Let me put it another way. If the patch from a PR remains stagnant, but the code it gets patched onto changes, the actual impact of the patch does too. This can cause all sorts of problems, from merge conflicts to erroneous test failures.</p>
<p>The longer code sits in a PR, the more time it usually ends up taking me to get it merged (when I finally have the time to get to it), just from the merge conflicts alone. It’s the Lindy effect rearing its ugly head once more.</p>
<h2>Features Age Too</h2>
<p>Not only does code age when left untouched and unused, so do the features they represent. If a user requests for a button that marks some text as bold and they don’t receive that feature promptly, they’ll find an alternative solution, possibly from a competitor.</p>
<p>The rest of the app can evolve too. If you’re an individual contributor, it’s in your best interest for your code to make it to master. If that doesn’t happen quickly enough, another contributor (or, if the software is extensible, like Obsidian or WordPress) or plugin author might beat you to the punch.</p>
<h2>What Can I Do About It?</h2>
<p>I believe the way I’ve been handling this until now has been entirely wrong. I should be allocating more of my time to training new contributors and fielding PRs from established ones. Open sources is a team effort.</p>
<p>As for you, dear reader, that’s for you to figure out. The worst way code can age is if it stays in your head. Don’t let your ideas go to waste. <a href="./never_wait">Open that PR</a>, or remind your reviewer to take another look if you haven’t already.</p>

Failing to account for this reality can slow down development and dissuade contributors from sticking around.

<p>Failing to account for this reality can slow down development and dissuade contributors from sticking around.</p>

Code Ages like Milk

<h1>Why I Talk to Myself</h1>
<img src="/images/bike_sidewalk.webp" alt="My bike broke down. Luckily, I had my camera">
<p>Each week I sit down and reflect on what I've learned. I think about which of my efforts have helped the Harper endeavor, and which have harmed it. It's not easy, but I think it's one of the best habits I've built. More accurately, it's one of the best habits <a href="https://ma.tt">Matt</a> has encouraged me to build.</p>
<p>When I really think about it, I'm reminded of the original <a href="https://blog.afterthedeadline.com/">After the Deadline blog</a>, which coincidentally also had a category called "Talking to Myself". I wonder if Raphael Mudge (AtD's author) had a similar motivation.</p>
<p>I talk to myself to find the weaknesses in Harper's mission and focus my attention on what matters. It is an efficient way to get from <a href="https://ma.tt/2025/05/sharing-levels/">layer one to layer five</a>. It can help get other maintainers on the same page and build a body of knowledge that can be linked to from anywhere.</p>
<p>Not worrying about keeping information locked-down is one of the more beautiful parts of working in public. Open source code would be nothing without similarly open source knowledge. I talk to myself <em>online</em> so my users can understand my intentions.</p>
<p>In many senses, we are not unlike moths. We are drawn to the lights of distraction, hype, and procrastination. For me, setting up checkpoints to stop and reflect has been an effective deterrent from these lights. In fact, I think each minute for which I stop and think has paid itself back three times over in time savings. If I can better allocate my time and energy, I can avoid doing redundant work pursuing pointless goals.</p>
<h2>Bonus Points</h2>
<p>A <a href="./markov_chains_are_the_original_language_models">few</a> of my <a href="./local-first_software_is_easier_to_scale">articles</a> have become popular on sites like Hacker News and Reddit. We've gotten quite a few Harper installations from links on these articles. I would never rely on this kind of blogging as a source of advertising, but it is nice to see it help spread the word.</p>

It's not easy, but I think it's one of the best habits I've ever built.

<p>It's not easy, but I think it's one of the best habits I've ever built.</p>

Why I Talk to Myself

<h1>Harper in the News</h1>
<p>The biggest complaint about Harper is that the quality of grammar checking still needs some work.
I've got an exciting new project I'm preparing to discuss later this week (hopefully) in that area.
It should be a step-function improvement over our previous methods for creating new rules.</p>
<p>In the meantime, I've collected some recent sightings of Harper in the wild:</p>
<ul>
<li><a href="https://obsidian.rocks/resource-harper/">Obsidian Rocks</a></li>
<li><a href="https://www.wearefounders.uk/the-grammar-checker-that-actually-gets-developers-meet-harper/">We Are Founders</a></li>
<li><a href="https://itsfoss.com/harper-grammar-checker/">It's FOSS</a></li>
<li><a href="https://www.xda-developers.com/ditched-grammarly-for-this-amazing-open-source-alternative/">XDA Developers</a></li>
<li><a href="https://www.maketecheasier.com/harper-vs-grammarly/">Make Tech Easier</a></li>
</ul>

I've got some exciting new work I'm preparing to discuss later this week (hopefully). In the meantime, I've collected some recent sightings of Harper in the wild.

<p>I've got some exciting new work I'm preparing to discuss later this week (hopefully). In the meantime, I've collected some recent sightings of Harper in the wild.</p>

Elijah Potter

Harper in the News

Comments

Other Stuff

Do Not Write with an LLM

Code Ages like Milk

Why I Talk to Myself