--- title: LLM Voice | daily.dev description: Martin Fowler warns writers against letting LLMs draft their prose, arguing it creates an 'uncanny valley' reaction he calls LLM-Voice. Even human editing... canonical: https://daily.dev/posts/llm-voice-v7q2iq2v5 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: LLM Voice | daily.dev og:description: Martin Fowler warns writers against letting LLMs draft their prose, arguing it creates an 'uncanny valley' reaction he calls LLM-Voice. Even human editing... og:url: https://daily.dev/posts/llm-voice-v7q2iq2v5 og:image: https://api.daily.dev/og/posts/v7Q2iQ2v5.png og:image:alt: LLM Voice og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # LLM Voice **[Martin Fowler](https://daily.dev/sources/mafo)** · 6 min read · 1 upvotes · 0 comments ## Summary Martin Fowler warns writers against letting LLMs draft their prose, arguing it creates an 'uncanny valley' reaction he calls LLM-Voice. Even human editing rarely removes it fully, though he knows one case where a professional editor at a major periodical cleaned up an LLM-drafted piece successfully. He argues delegating writing to LLMs also delegates thinking, since writing forces authors to resolve their own confusions. He makes exceptions for cases like Jim Highsmith, who uses an LLM to write despite Parkinson's disease, and for non-native English speakers using translation tools. Fowler states he used no AI while writing this post and requires contributors to his site to declare AI use in acknowledgments. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is LLM-Voice and why does it bother readers? LLM-Voice is the uncanny, slightly off quality readers sense in prose drafted by a large language model, even after human editing. Martin Fowler describes it as creating an 'uncanny valley' reaction that makes him lose interest in reading on. He argues it's very hard to expel once an LLM has written the first draft, even with heavy human revision. _Writers weighing AI drafting tools can track this debate as it evolves on daily.dev._ ### Should I let an LLM write the first draft of my article and then edit it myself? Editing an LLM draft rarely removes its telltale voice, so it's risky if you want prose that reads as genuinely human. One known exception required a professional editor at a major periodical to fully clean up an LLM-written first draft. Using an LLM only to build an outline or structure may be acceptable, but filling in the actual prose should be done by the author. _Anyone deciding how much AI help to use when writing can follow this discussion on daily.dev._ ## Community take How the wider developer community reacted, aggregated from 1 discussion and 101 comments across hackernews (as of 2026-10-29). **TL;DR:** Discussion largely agrees that LLM-generated prose has a distinct, draining 'voice' that signals low effort and erodes trust, though many push back on blanket shaming, arguing AI use is fine for personal writing, accessibility, or low-stakes communication versus material meant to persuade or connect with readers. **Sentiment:** 10% positive · 35% mixed · 55% skeptical **The case for** - AI writing can level the playing field for people with disabilities, low English fluency, or limited time to write. - Using AI purely for personal or low-stakes writing (notes, drafts, bureaucratic docs) is seen as harmless. - Some argue AI can be used as a critique/questioning tool during drafting without taking over the final prose, preserving the author's voice. **The pushback** - LLM prose is described as exhausting to read, low in information density despite being verbose, and lacking the 'surprise' of genuine human communication. - Sharing LLM-written content with others without disclosure is compared to serving reheated frozen food as if freshly cooked — a breach of trust. - Writing is viewed as inseparable from thinking, so offloading writing to an LLM also offloads the reasoning and resolution of ideas. - Some worry good-faith human writers are now unfairly suspected of being AI-generated, or feel pressure to avoid natural phrasing lest it 'sound like LLM.' - A few suspect commercial incentives or least-effort motives behind strong pro-AI stances. **By community** - hackernews (mixed): Most participants are critical of LLM-voice as exhausting, dishonest, or empty, but a vocal minority defends AI-assisted writing for accessibility, low-stakes use, or as a thinking aid rather than a replacement for writing. **Hottest debate:** Whether using AI to write for others is inherently disrespectful/dishonest (like serving reheated food as fresh) or simply a legitimate tool depending on context and disclosure. **Open questions** - Will the signal that distinguishes human from LLM writing eventually disappear as models improve, or will human writing evolve to stay distinguishable? - What neurological or psychological mechanism makes LLM prose specifically feel 'exhausting' or 'wrong' to read? - Can frontier labs meaningfully fix LLM writing style by training on curated 'better' prose, or will any broad training produce the same bland average? **Highlights** > I suspect it’s partially an average / regression to mean. If you built a model only based on long form journalism and The New Yorker voice, including both the template structure of the article and the grammar and syntax, the model would replicate the patterns it is built upon. You could do the same for high vocabulary rap. When you stuff all of English or all of human thought into it with somewhat equal weight, including very technical writing the sum is a lack of style and elegance and personality. The second problem is that writing often doesn’t happen linearly. You write bookends and fill in, edit and revise. You sort of know where you’re going and can then make decisions how to get there. Language models may benefit from being redesigned to perform passes. Select a template of some sort, fill in the mad lib, then run a detail pass, then run a persuasion / emotion pass, then run a style pass. Current model writing very much sounds like I would expect autocomplete to. > — [basch on hackernews · 2 comments](https://news.ycombinator.com/item?id=50293767) > > I just had one of my posts removed from the front page here which was disheartening because it feels as though what was judged was prose rather than the technical insight in to a niche problem, which seems counterintuitive considering i've always thought of Hacker News being an inclusive place for the technical and engineering skills rather than their level of the english language. The problem with LLM voice is that it sends the signal that it's a likely waste of effort to try to figure out if there's any "technical insight" there, because more often than not, there isn't. Maybe it's not fair that you're paying the price for others using LLMs to abuse their readers, but it's the reality that you have to deal with. It's also not fair to expect readers to wade through all that abusive LLM prose in order to find your (assumed) non-abusive post. Once LLM writing becomes indistinguishable from human writing, it will be really sad. There will just be a collapse of human engagement, and you might as well just talk to your machine because then at least you know what you're getting and aren't being deceived. > — [10111001 on hackernews · 1 comments](https://news.ycombinator.com/item?id=50292986) > One of my favourite rules: perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away. This is invaluable to writing. So much text from an LLM feels superfluous, like I could've said the same wall of text with just a few sentences. Writing is tightly coupled with thinking: refining thoughts, arguments, ideas, discarding them, mulling over them. LLM text always raises this nagging question in my mind: am I reading what you think, or what your agent parroted back and you accepted? > — [cmsefton on hackernews](https://news.ycombinator.com/item?id=50293908) > Has anyone done any work on what it is that makes LLM Voice hard? Something I've noticed, and have seen other people comment in passing is that reading LLM output is "exhausting". It's often attributed to trying to detect the machine, or filter out BS or word density or some other explanation, but it feels like something more than just that. Like - there is something about LLM output that just doesn't "fit" our minds correctly, even though words are in the right places. Surely someone with a neuro/psych background has picked this up and developed an evidence-backed explanation. > — [thadt on hackernews · 1 comments](https://news.ycombinator.com/item?id=50294048) > I’m fine with my neighbour eating TV dinners, but if they invite me for a meal I expect to not be served a TV dinner, or to at a minimum be informed beforehand that that’s what going to be served. Same thing with LLM-assisted writing. It’s one thing to do it for yourself, quite another to share it with other folks without warning. > — [latexr on hackernews](https://news.ycombinator.com/item?id=50293362) **Source threads** - [hackernews](https://news.ycombinator.com/item?id=50290177) · 135 points · 101 comments ## Similar posts on daily.dev - [Don't Let AI Write For You](https://daily.dev/posts/don-t-let-ai-write-for-you-vbrxr72gu) · Hacker News · 5 upvotes · 1 comments - [I don't like LLMs](https://daily.dev/posts/i-don-t-like-llms-cy8kqjrl7) · Martin Fowler · 2 upvotes · 0 comments - [The social contract of writing](https://daily.dev/posts/the-social-contract-of-writing-7ouz8bqa6) · Lobsters · 7 upvotes · 2 comments - [The rebellion against robot drivel](https://daily.dev/posts/the-rebellion-against-robot-drivel-zgx8rnjer) · InfoWorld · 1 upvotes · 0 comments - [It's Not Just True, It's False](https://daily.dev/posts/it-s-not-just-true-it-s-false-o6dkgy8f8) · Ibrahim Diallo · 6 upvotes · 4 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ethical-ai](https://daily.dev/tags/ethical-ai) [View this post on daily.dev](https://daily.dev/posts/llm-voice-v7q2iq2v5) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"LLM Voice","url":"https://daily.dev/posts/llm-voice-v7q2iq2v5","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/llm-voice-v7q2iq2v5"},"datePublished":"2026-10-29T14:32:56.030Z","dateModified":"2026-10-29T19:25:23.599Z","description":"Martin Fowler warns writers against letting LLMs draft their prose, arguing it creates an 'uncanny valley' reaction he calls LLM-Voice. Even human editing...","image":"https://media.daily.dev/image/upload/s--qPvKM23u--/f_auto/v1722860399/public/Placeholder%2009","thumbnailUrl":"https://media.daily.dev/image/upload/s--qPvKM23u--/f_auto/v1722860399/public/Placeholder%2009","isAccessibleForFree":true,"articleSection":"Martin Fowler","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Martin Fowler","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/mafo","url":"https://daily.dev/sources/mafo"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/llm-voice-v7q2iq2v5","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ethical-ai","timeRequired":"PT6M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Martin Fowler","item":"https://daily.dev/sources/mafo"},{"@type":"ListItem","position":3,"name":"LLM Voice"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/llm-voice-v7q2iq2v5#faq","mainEntity":[{"@type":"Question","name":"What is LLM-Voice and why does it bother readers?","acceptedAnswer":{"@type":"Answer","text":"LLM-Voice is the uncanny, slightly off quality readers sense in prose drafted by a large language model, even after human editing. Martin Fowler describes it as creating an 'uncanny valley' reaction that makes him lose interest in reading on. He argues it's very hard to expel once an LLM has written the first draft, even with heavy human revision. Writers weighing AI drafting tools can track this debate as it evolves on daily.dev."}},{"@type":"Question","name":"Should I let an LLM write the first draft of my article and then edit it myself?","acceptedAnswer":{"@type":"Answer","text":"Editing an LLM draft rarely removes its telltale voice, so it's risky if you want prose that reads as genuinely human. One known exception required a professional editor at a major periodical to fully clean up an LLM-written first draft. Using an LLM only to build an outline or structure may be acceptable, but filling in the actual prose should be done by the author. Anyone deciding how much AI help to use when writing can follow this discussion on daily.dev."}}]} ```