Hey you,

Yes, you, who are thinking about not hiring a technical writer this year or, worse, erased one or more technical writing positions last year because of AI. You, who are buying into the promise of docs entirely authored by LLMs without expert oversight or guidance. You, who unloaded the weight of docs on your devs’ shoulders, as if it was a trivial chore.

You are making a big mistake. But you can still undo the damage.

It’s been a complicated year, 2025. When even Andrej Karpathy, one of OpenAI’s founders, admits, in a fit of Oppenheimerian guilt, to feeling lost, you know that no one holds the key to the future. You flail and dance around these new totems made of words, which are neither intelligent nor conscious, pretending they can replace humans while, in fact, they’re little more than glorified tools.

You might think that the plausible taste of AI prose is all you need to give your products a voice. You paste code into a field and something that resembles docs comes out after a few minutes. Like a student eager to turn homework in, you might be tempted to content yourself with docs theatre, thinking that it’ll earn you a good grade. It won’t, because docs aren’t just artifacts.

You keep using that word. I do not think it means what you think it means

—The Princess Bride

When you say “docs”, you’re careful to focus on the output, omitting the process. Perhaps you don’t know how docs are produced. You’ve forgotten, or perhaps never knew, that docs are product truth; that without them, software becomes unusable, because software is never done, is never obvious, and is never simple. Producing those docs requires tech writers.

Tech writers go to great lengths to get the information they need. They write so that your audience can understand. They hunger for clarity and meaning and impact. They power through weeks full of deadlines, chasing product news, because without their reporting, most products wouldn’t thrive; some wouldn’t even exist. Their docs aren’t a byproduct: they tie the product together.

An LLM can’t do all that, because it can’t feel the pain of your users. It can’t put itself into their shoes. It lacks the kind of empathy that’s behind great help content. It does not, in fact, have any empathy at all, because it cannot care. You need folks who will care, because content is a hairy beast that can only be tamed by agents made of flesh and capable of emotions: humans.

AI generated docs are broken

You can’t generate docs on autopilot. Let me tell you why.

First, AI-generated docs are not intelligent. They not only make up things in subtle ways: They lack vision. Even if you fed them millions of tokens, they couldn’t develop a docs strategy, decide what not to document, or structure content for reuse. And they fail to capture the tension, the caveats, the edge cases, the feeling of unfinishedness that only someone who cares can feel. Without that grounding, docs are hollow.

Second, liability doesn’t vanish just because AI wrote it. When docs cause harm through wrong instructions, someone will be held responsible. It won’t be the model. You can’t depose an LLM. You can’t fire it. You can’t point at it in court when a customer’s data evaporates because your GenAI runbook told them to run the wrong command. That someone will be you, or someone who reports to you.

Third, even your favorite AI must RTFM. All your Claude Skills, Cursor rules, all the semantic tagging that makes RAG work, is technical writing under a new name: context curation. You fired or didn’t hire the people who create high-quality context and then wondered why your AI tools produce slop. You can’t augment what isn’t there. The writers you let go were the supply chain for the intelligence you’re now betting on.

The solution is to augment your technical writers

It’s not all bad news: Marvelous things can happen if you provide your writers with AI tools and training while you protect the quality of your content through an AI policy. I’ve described the ideal end state in My day as an augmented technical writer in 2030, a vision of the future where writers orchestrate, edit, and publish docs together with AI agents. This is already happening before our eyes.

Productivity gains are real when you understand that augmentation is better than replacing humans, a reality even AWS’ CEO, Matt Garman, acknowledged. Read how I’m using AI as a technical writer. I’m not alone: Follow Tom Johnson, CT Smith, and Sarah Deaton, and discover how tech writers are building tools through AI to better apply it to docs.

Develop an AI strategy for docs together with tech writers, and give them time and resources to experiment with AI. Tech writers are resourceful by nature: they’ve spent careers doing more with less, optimizing workflows, finding clever solutions to impossible quests. Give them the tools and a bit of runway, and they’ll figure out how to make AI work for the docs, not instead of them.

So here’s my request for you: Reconsider

Reconsider the positions you did not open. Or the writers you let go. Reconsider the assumption that AI has solved a problem that, at its core, is deeply human and requires not only concatenating words, but also chasing subject-matter experts and understanding the subtleties of product motions, among many other things.

Technical writers aren’t a luxury. They are the people who translate what you’ve built into something others can use. Without them, you’re shipping a product that can’t speak for itself, or that lies. Your product needs to speak. AI can generate noise effectively and infinitely, but only a technical writer can create the signal.

Don’t choose the noise. Get them back. Get them onboard.

Acknowledgments

Thanks to Tiffany Hrabusa, Casey Smith, and Anna Urbiztondo for their reviews of early drafts and for their encouragement. Thanks to my partner, Valentina, for helping me improve this piece and for suggesting to wait a bit before hitting Publish. And a heartfelt thank you to the tech writing community and its wonderful human beings.