Report: How Developers React to AI-Scented Blog Posts

SubscribeSign in

Report: How developers react to AI-scented blog posts<br>98% of readers prefer imperfect-yet-authentic human writing; most will stop reading, block you, and downvote you if they suspect AI<br>Jun 16, 2026

Share

“Wow, looks like Claude just wrote another blog post…can’t wait to read it!”<br>– No developer, ever

The Writing with LLMs report explored how developers – many of whom never published before – are now using LLMs to write technical blog posts. The most commonly cited reasons were to get past the blank page barrier, publish more often, and save time. That (potentially) helps the writer. But what about the reader?<br>While working on that writing report, I wondered what specific actions readers take when they suspect articles are LLM-authored or LLM-assisted. Do they hold their nose and continue if they sense underlying human insights? Block the domain and try to sabotage the article’s success on the aggregators? Something in between? And are they any more forgiving when it’s a non-native English speaker who uses an LLM to “polish” their prose?1<br>So, out went another survey. I shared this one across X, Bluesky, and LinkedIn. Lots of people feel quite passionately about this topic; I ended up with 668 responses (versus 181 for the writing survey). Note that while I targeted the survey to developers and readers of tech blogs, the survey was anonymous – so there’s no demographics guarantee.<br>TL;DR Readers really despise and distrust articles that seem shaped by AI. They not only stop reading, but also block and punish the author. Non-native English speakers get little mercy. Readers overwhelmingly prefer imperfect-yet-authentic writing over LLMified prose.

Get Write that blog! updates in your inbox

Subscribe

Readers care a lot…and will punish you to the best of their ability

The first question was “How much does it matter to you if a tech blog seems AI-assisted/AI-authored?” A lot, actually. Here’s what people reported:

85% of respondents rated their concern a 5 out of 5, and 11% rated it a 4.<br>How does that concern translate to action? The next question asked, “How do you respond if you think a tech blog is AI-assisted/AI-authored?” Respondents could select multiple responses. The results:

So if readers think your article is AI-assisted or AI-authored, most will immediately leave (78%), avoid you forever (71%), and try to downvote you if they can (57%). 17% try to finish but lose interest, and around 15% will only continue if the underlying insights seem authentic.<br>And that brings up an interesting point. There are a lot of different ways that a tech blog might end up feeling AI-assisted/AI-authored:<br>A non-native English speaker writes a complete draft, then hands it to an LLM to “polish my English”

A novice writer isn’t sure how to get started, so they collect their notes, voice recordings, PRs, etc. and ask the LLM to “make this into an engineering blog post”

An aspiring influencer is trying to get attention, so they ask the LLM to write an article on some trendy topic and publish the response

All might end up with the same LLM writing tropes, even though the amount of thought behind them is certainly not equal. But readers generally don’t differentiate in their reactions. They distrust the article, instinctively.<br>Does English mastery matter?

That leads to the second part of the survey, which looked at whether the author’s English mastery mattered.<br>Would readers be more forgiving if they knew the author was a non-native English speaker writing their own thoughts but using AI for translation or language assistance?

No, not much mercy here – only 23% would respond differently. Still, if you are using AI/LLMs for this purpose, you might want to disclose that prominently in your article. It could keep a fraction of readers from instantly dismissing it as pure AI slop.<br>The final question asked if readers would rather read the author’s authentic words – even with some awkward phrasing, grammar gaffes, and other nits. Or, do they prefer an LLM-polished version?

98% preferred the author’s own writing, as is. So writers, please – don’t let LLMs blandify your text. People would rather hear your authentic voice than suffer through an eerily soulless rewrite.<br>Could you create a Claude Skill to run a light editing round, revising awkward sentences in a way that approximates your actual voice with textbook grammar? With some work, maybe.2 Would a human editing your article be better or worse? Totally depends on the editor. And the success of both ultimately depends on how thoroughly you review every suggested change.<br>My unsolicited advice on the matter: focus on clarity over “perfection.” If you share interesting insights and your target reader can understand what you’re trying to communicate, some rough edges shouldn’t be an issue. A natural human voice is way more readable than perfectly predictable LLM prose. And – unless you’re opposed to...