As someone who is aggressively pro AI, I've also noticed how awful everything is to read lately. It's always so hard to put my finger on the exact thing, but it's a mixture of everything sounding identical and everything being a rambling unconcise mess. I find I routinely start reading interesting posts and just .. stop part way, abandoning it. A site I read stories on has been flooded with AI generated content, all of it terrible, and it just makes me retreat to things created before 2023 as a way of filtering it all out.

I just can't seem to stay interested and read LLM text. It just seems to lack going somewhere, its information density is often really low and its full of odd isms that I don't find very interesting to read. I don't have a good radar for detecting the various LLM particular speech patterns but everytime I have an issue with an article there is always people talking about how its clearly AI written.

I find the same thing with written results out of AI "research", it is a slog to read through it and most of it can be skipped even if it can sometimes be useful its an unenjoyable read.

All human communication is based on internally simulating the presumed mind that was/will-be at the other end. Even now, some part of your brain--however subconscious--is building a tiny sketch of me with assumed beliefs and motivations, and you're using it to predict and interpret the next words.

But what happens when everything is persistently, subtly wrong? When we cannot smoothly map the text to a simulation? An inner-sim keeps throwing errors and mis-predicting what comes next is unnerving and draining. Like talking to a human who is not currently sane.

Sometimes I imagine LLM text like a plastic animal model put out to confuse local wildlife, where it hits just enough instinctive triggers that they try to mate with it before leaving in frustrated bafflement.

______

Half-Life 2, Dr. Kleiner: "Never fear, Gordon, she's de-beaked and completely harmless. The worst she might do is attempt to couple with your head. Fruitlessly!"

Broken record: LLMs are exceptionally useful writing critics if you know how to hold them correctly, but you can't use a single word they suggest to you --- the better the LLM's word choices seem, the more likely the LLM is infecting your voice. Better, by a lot, to prompt the LLM not to suggest any words, but just offer objective critiques.

This, to me, is the strongest evidence that even with all the money frontier labs are throwing into pushing the boundaries of model training and fine tuning, they still don't actually understand what they're doing & don't have as much control as they want you to believe. If they had any real ability to tune & adjust the model & its output, they would have fixed "LLM voice" by now.

Even if you don't think it's a problem, even if you believe that it's just different than human writing, it seems the majority of "regular" people can tell when something is LLM-written and find it distasteful or unpleasant to read, at best. That alone should be more than enough incentive for frontier labs to train this out of their state-of-the-art models, if only they could. The fact that they haven't suggests that they cannot, despite all the claimed advances in model analysis and "explainability."

The voice is one thing, but I have encountered multiple 'tech support' teams and platforms that will respond to me in pure, unedited LLM output.

I know this because the topic is usually an obscure issue, and I have already gone back-and-forth with multiple LLMs to see where their boundary of knowledge is on it, and it will often include hallucinations along the way (e.g. pointing to old/nonexistent settings pathways)

So when the first reply from the official 'support' channel contains the exact same nonexistent settings pathways, it is easy to conclude they are sending me the LLM output without doing the verification steps that I have already done on my end. When I point out what doesn't exist, the reply is always a similar flavor of the usual LLM "you're absolutely right" response after being corrected.

Like a lot of folks, the 'tells' of LLM writing for me are more of an 'I know it when I see it' than a specific set of hard criteria (which is probably why it's so hard to make a tool for it), but if I could sum it up, it's something like this:

I always had a high faculty for language, and when I was young, even very young, I was able to speak in full sentences much like I'm doing now. What I didn't have was context. I didn't know how to relate to people socially, and then later on, I wasn't as experienced an engineer as my peers, but I was trying to act adult and mature in both ways.

So I tried to write in ways that were catchy and unnecessarily complex, because I thought that conveyed sophistication and intelligence, and that I knew what I was on about more than I did. It's the kind of writing that gets you through a 30-page final paper for a class you don't really like, or a research project you didn't finish. The term I often use when complaining about my colleagues' LLM output is "flashy but meaningless."

As my confidence grew, I started writing in simpler (though still not the most concise, that's just not who I am!) ways, because it became easier to convey ideas simply in a way that made it obvious I understood them.

When you think about what LLMs are - word machines without (or with very limited) context - you can see why they write the way they do. They are always acting; trying to replicate a human contextual experience with only weights and tokens to carry them through it. That fact is part of why I am skeptical of their ability to suprass humans in reasoning based on their own output.

I currently have a case open with Honda corporate customer support, and I have this suspicion that the email responses I receive are being generated with an LLM directed by a human. Either that, or I'm corresponding with a human who has poor written communication skills. The text grammatically correct, but overly verbose and repetitive. It reminds me of LLM text, but from a model that is several generations behind today's models.

Maybe it's like when you can just tell a picture is AI. You recognize the words you are reading, but to your brain, the way the words are put together feel wrong.

Incidentally, I was reading Harry Potter to my kids last night. While waiting for everyone to arrive, I started reading the previous chapter to those who were present. But I read the sentences backwards while using a vocal inflection of regular narration. My 6-year-old recognized that something was wrong and asked me to "stop reading funny". My 11-year-old recognized that I was reading backwards. The same 6-year-old thinks every silly photo that I show him is AI, while my older children can actually spot the AI.

>Like most writers, I find writing an almost-essential tool for thinking. Often it's only by writing something down, that I can really get my thoughts on a topic into coherence. I've often started an essay knowing I don't really understand something, but the act of wrestling with the words forces me to overcome my confusions. As more people delegate writing to LLMs they will also delegate thinking and comprehension to LLMs.

A very important point, which was also made more forcefully in "Human Routers of Machine Words" by Fernando Borretti, shared on this site previously. It's good that Fowler brings up a very legitimate use case in the next paragraph, though.

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.

I have recently noticed people suddenly using the word "geniunely" in places where it serves no purpose [1][2], yet these posts don't seem to be otherwise AI-written. I guess that LLMisms are also creeping into the voice of their users.

[1] "[Robert Caro is] a guy who genuinely rarely gives interviews" https://manifold.markets/DylanBowman/who-is-dwarkeshs-upcomi...

[2] Scott Aaronson uses it four times in a short post: https://scottaaronson.blog/?p=10222

It's hard to define and pin point what a "voice" is. Think about how you would describe the difference in voice of Tolstoy vs Dostoevsky, or William Gibson compared to Thomas Pynchon. You might have to read a lot of a writer to be able to notice their voice. You might even notice how a short story by an unknown writer resembles the voice of another.

Or think about reading something that might be stuff you could say in your own words but not actually in your own voice. Or maybe the opposite is a better example for you? That's what some poetry is about. The gap between voice and word.

Stylometry is the study and measurement of text, so that can go some distance to quantifying the difference between voices and can help to programmatically identify a human voice but I feel that because a voice comes from a person, the recognition is also mainly a human response.

At another level its about communication, person to person with the intention and desire to communicate something. LLM text output may seem uncanny as it's hard to emulate the intention and desire and empathy present in genuine human communication.

I would be interested in reading what poets say about LLM voices. I wonder if many LLM labs employ them?

I agree with the author, and I like how his (human) voice comes through loud and clear in this post.

I also find LLM-voices extremely jarring, boring, and dense. And very, umm, mechanical, if you will.

But I want to make an observation.

Folks are not-Ok reading the LLM’s output for prose.

But folks are completely Ok, even downright blind, to accepting LLM-output of code.

I see that dichotomy everywhere.

LLM-driven code: great, valuable, all-ok LLM-driven prose: not ok

Understood completely that the audience for consumption is two completely different entities (machines vs humans) but something about this doesn’t sit right with me.

More particularly: if we don’t prefer LLM-prose, why should we prefer LLM-code ??

Anybody else feels that ??

Sometimes writing to an LLM seems to make me lazy. "It'll understand," I think and then just leave a thought sloppy, inchoate, even reasoning to myself that it's better to write that way to see what'll be triggered in the LLM. Then I think... uh oh I'm strolling down the primrose path.

It’s also just… disrespectful to the human reader. Quite plainly.

I think the core functionality of human cooperation boils down to getting ideas from one mind to another or many. In that, I think AI is a boon for people with limiting abilities such as learning disabilities, low language fluency, or generally poor writing skills. It evens the playing field for new and interesting ideas to spread. I can't help feeling, sometimes, this is about people who excelled at something demanding they still have their place at the top. High literacy such that you can discern LLM Voice is quite an accomplishment but blanket dismissal reeks of privilege.

That being said. Yes, I hate some LLM Voice moments too and still sometimes judge people for copy and pasting, expecting others to do the comprehension when it seems they haven't done it themselves.

Duality of man.

I’ve started embracing my grammar flaws and typos. Emails and chats may get a proof, but only from me or another human, then I send it. I’ve deleted Grammarly because I think the flaws make the comms more authentic in the age of AI.

And of course I’ve dropped my use of the em-dash after being accused of being AI myself.

> 3: Michael Nygard commented: “There you really aren't trying to communicate thoughts from one person to other people. You're communicating with a process rather than a human. A human may open the envelope but that person is acting as a component in a cybernetic organism.”

But of course: manufactuting itself creates " productive mechanism whose parts are human beings, as Adam Ferguson describes the industrial workshop: "an engine, the parts of which are men".

> or preparing a report I suspect no-one will ever read

We should actually restructure our organizations/society in such a way that this stops being necessary. If no one reads a report, it was almost certainly a waste of effort. The only possible benefit was for the report-writer (which of course is also flushed if the "writer" uses an LLM). These things were already a plague before LLMs.

Late to this party, but wanted to share my favorite piece of writing about how annoying AI-generated prose it to read: https://storage.courtlistener.com/recap/gov.uscourts.med.612....

You can be pro AI productivity but anti-AI creativity.

Automating creativity is the death of art.

> uncanny valley of prose

Ah good ol' serendipity. I strongly believe this too, and I think that's why older (less capable) models fared better: they hadn't fallen into the valley yet. Maybe it's not a valley at all, and is just a cliff after which there is uncanny.

When the earliest models launched I was super-excited to no longer have to write pull request descriptions and commit messages. I have done a sharp 180 on that and I'm no longer a meat proxy[1]: instead I point out any redirects and improvements I made the agent do before opening the PR (which there are still _many_ in even the most trivial tasks).

[1]: https://meatproxy.me/

One reason it grates is that it is our brain's way of telling us that there is nothing to be gained from engaging the material; it is trite.

Remember when ChatGPT first came out and the coolest thing you could do with it was ask it to say something in pirate speak, or in the style of Tolkein, or Obama, or whatever? Maybe that should still be a thing ;)

I don't mind reading LLMism, but hearing people say it out loud like on youtube videos irritates me to no end.

> As more people delegate writing to LLMs they will also delegate thinking and comprehension to LLMs. I don't want to do that, where's the fun of writing then?

I enjoy cooking from scratch, and I really don't like frozen foods, microwave/TV dinners, or even DoorDashing

but we're past complaining/arguing/shaming folks who don't cook every meal from scratch

you know what, I don't prefer eating it, but if my next door neighbor loves microwaving frozen $3 burritos, then go ahead

to some people, food is art, creativity, life to others, food is fuel

also applies to writing and thinking

Sentence paragraphs are the biggest clue for me. I hate them. They eviscerate any structure or connection that might flow through the text. No ideas are related any longer. They just splat down on the page like a new isolated collection of words that the reader must struggle to join into a fully formed thought. It's too much; I give up when I see those posts.

I wonder if over time the two will simply merge and become indistinguishable from one another. The human writing is already changing, as the author has observed. The LLMs will follow the change, and soon meet us in the middle!

Or maybe the humans will do the opposite and shift to Sardukar-style communication instead? LLMs can easily match that as well.

That style I would very much prefer to all the fluff people write today and were writing even before AI.

I started writing a novel recently - in french. I have a clear view of each scene by this I mean not exactly what happens but the general motion, its objective and where it should land. But only when I start writing does it come to life and so many things I did not anticipate pop up and give depth to the characters and sense to the overall plot. Never would the AI give me that, however good the result may be

I think the key tell is that the writing doesn't actually illuminate anything worth learning. The more I see of AI produced output, the more I believe that there is nothing there that wasn't already in the prompt. It's an inflation of a single idea to pages of text. It meandors about and never actually gets into detail.

It's a lot like talking to someone who doesn't know what they are talking about but are good at faking it for a while. The game works by leaning on pat phrases, circular reasoning, and hoping nobody is paying too close of attention. I think this is why it's so popular with the corporate executive class, as that has been their mode of communicating their entire careers. To them, they don't see anything wrong.

"Shitty writing existed before AI" is unfortunately an all too common rejoinder to this issue. That's not an excuse. That doesn't make it better. Just because the pointy haired boss talks that way doesn't mean we should be gleefully pouring gasoline on the fire.

I don't know, with LLM it feels like I finally have possibility to prepare all the dull text that corporate environment expect in a form that colleagues are able to digest even if they don't specially enjoy it like it's the best poetry their beloved one ever composed specially for them. This is in sharp contrast with the struggle to shrink into a somewhat coherent excerpt the wide web of interweaved psychological experience flow graph into this alien expressions which are text streams. It's still impossible to dump really the whole experience and communicate wholly the full semantics they cast, but at least these automatons are never scared to face far much than they can sustain in what can be thrown at them, and apparently most people prefer the dull soulless prose to whatever can suggest vivide continuous reconnection of everything as the context inevitably drift and everything needs to be reconsidered properly with all the rest.

It's incredibly hard to maintain your own voice while using LLMs. It requires deep discussion with the LLM itself so that it comes to understand your perspectives and also explains to you the premises on which its own recommendations are based.

I haven't really done a systematic analysis to compare the quality of fully human-written content and content generated via this kind of deep human-AI collaboration, but my intuition is that the latter is vastly superior. Perhaps AI at its best is a research tool, but a genuinely intelligent one that we need to engage in conversation with in order to use to its full potential.

The sense that I get is that as a conversation gets more in depth, the LLMs start behaving more like reasoning engines and less like statistical averaging machines. The former is vastly more useful to a person who wants to develop an idea and articulate it well.

This is the consequence of the majority compulsion to communicate in language figures as opposed to concepts.

It's human to do that because it's too dry and unwelcoming to talk in precise concepts disregarding any cognitive friendliness, not to mention do that all the time.

A computer programming in any high level language is concept talking all the time and only us nerds like it :D

Not even that as we form tribes with preferences to one or other computer lang.

Now we're talking less to each other and AI is going to influence us to invent a trillion things that will make understanding diverge and start to become scarce.

I disagree with the advice to avoid having an LLM write a first draft and then editing it yourself.

To me, for most business and academic content, the ultimate goal is to convey information. If I can edit AI copy to legibly convey that information clearly, that's enough and I can move on.

The saying, "Don't allow perfection to be the enemy of the good," comes to mind.

And I'm not convinced an individual can accurately edit without learning just as much as if they wrote the content. I guess the question is: how hard does the person work at editing. And does the individual challenge themselves to not put out content they themselves don't understand.

We're in a transition period. And as in all transition periods, many recoil at what is new. LLM writing still feels strange. And it's understandable.

But, for better or worse, in a business and academic setting, as the years go by, most of us will shrug and move on.

As with any topic, there will always be a minority of fanatics who cling to the rejection of AI writing in all its forms. Just as some prefer Vegan leather or tabs over spaces.

But among those who get things done, most (though not all) will accept that it's not worth it to exclude AI from writing. There simply isn't sufficient ROI.

It's like trying to bring back low-end highly human employed manufacturing to wealthy industrialized nations.

That ship has sailed.

I do agree, however, that in cases where an individual is reading for pleasure -- e.g. not in a business or academic context, authors will be forced to pick their audience: those who enjoy writing, even if it smells of AI, and those who don't.

I guess to summarize, my feeling is that using LLMs to write is perfectly acceptable, and used well you'll likely learn just as much. But you must be willing to lose some of your audience. And, over time you'll lose less and less of that audience.

This other day I went to a cottage with friends, we were having fun, and I decided to experiment "integrating AI" into our conversation as if it was a person in the room.

I turned on ChatGPT on Voice mode with no context whatsoever and just put it there in speaker in the room.

It started giggling and say "yeah" "right" and some very weird stuff. Then one of our friends reacted to that and started calling it out, then it started changing.

I'll keep it short, but basically the thing was super hilarious we had a lot of fun and at the end it triggered it safety protocols and its voice completely changed and it was acting super weird.

I know this post isn't about that, but I feel like soon we will have true AI companions in the home and around people.

I have a product that writes AI-written nonfiction books. I expected more thinking to help, but Sonnet 5 on high effort wrote the weakest prose of any setting (scored by my own blind reads + different LLM judge), and it was slower and more expensive too with some chapters running out of tokens mid-sentence. I agree that the voice is hard to edit out afterward - in my tests, which model & settings I picked changed it more than prompting/instruction.

> In time it may be that LLMs will be so good that we can't tell the difference.

I have to believe that we are within a couple of years if not months of this. This problem is too well-known to not be a major focus of the frontier labs' efforts.

I do document work, and at home I even do image work as a hobby, and all of it has that "AI feel". So I always do the work and then cross-check and verify it with independent agents or completely different services, asking them to take out this "AI feel", but like a writing style that gets baked in, or the particular feel an illustrator has, it doesn't seem to go away as I intended. I guess AI providers are probably aware of this problem too. But I don't know how they would measure it and improve on it (with what benchmark?).

He cites a few cases where letting the llm do the writing are acceptable. If someone has a disease such his friend with Parkinson’s. If you’re not a native English speaker (or writing/translating in anyones non native language we may presume) then it’s ok. Finally if it’s something no one will ever read such as a form letter or report.

Those all sound reasonable to me. What he staunchly ignores is that coding is essentially writing, but that ship has already sailed!

> But if I want to communicate with humans, my humanity has to shine (or at least ooze) out of my prose.

Yeah. I want to be able to touch a human being (or at least have one touch me) through the words. LLMs poison that, even if a human prompted the LLM.

Translation... I think I object to translation less than Martin does. There's a human back there, and they're writing all the words, and then an LLM is translating them. The LLM may not translate them perfectly, but they aren't going to introduce "load-bearing" when it wasn't there.

The bit about using LLM when you're writing for a bureaucratic process rather than a human was interesting, at least to me. I hadn't thought of it that way before, but I think it makes sense.

I'm of a different opinion.

1. If you use the first person pronoun "I", you better make damn sure you wrote those words.

2. It's fine otherwise to use LLM words, but I beg you - DON'T SHIP YOUR FIRST DRAFT. Read it, reread it; delete words in it; replace three words with one; replace dumb or overused turns of phrase with fun metaphors; generally give it a sprinkling of joy; break up punchy sentences by making them run-on sentences to trigger your readers' OCD... you get it.

Did I mention not shipping your first draft?

Couldn't agree more with the post.

I wrote down a few of my thoughts a few weeks ago at https://notes.pault.ag/why-write/ -- I find myself scanning text to see if it's LLM before I actually engage with the content.

It is getting harder to do by the day, and has slowly moved from a mental exercise (look for specific markers) into a "gut feel" based on how the pieces fit together.

I write a large amount of AI stuff 30 books. 4 million words. I find I have finished a book and don't want to read it.

Statistically even or repetitive over a chapter, chapter after chapter. So make an exaggerated point prove it come up with a conclusion and then have an exception.

Repeater 4 to 6 times a chapter in a book of 45 chapters.

So I wrote something to remove it and yes it is better but still not very good.

I noticed the exact same thing, and I also don't want the LLM to suggest anything to me. I do want it to comment, and to tell me what could be improved, but not to give me text.

I created a small tool to make this easier: https://www.writelucid.cc/. It's free (bring your own API key), feedback would be appreciated.

One of the big problems is that everyone wants to be an influencer and this desire is being exploited by reams of influencers using their own agents to write LLM Voice on how you too can become an influencer by using agents to write LLM Voice... get it yet?

I used elevenlabs to generate speech-to-text for this article. I selected Roger - Laid back Casual and Resonant, and it came out quite good!

i agree with this in general. I believe it's because it makes _too much_ sense. There's no non sequitors, jumps of logic that assume the reader is already on the same page as you.

To this point, however, writing informative articles about regulation or legal topics goes great! I don't want there to be ambiguity.

One tactic I use is to point out specific phrases as awkward, and create a tone guide both for the series and the specific article. They can mostly understand when you show them, it makes the process easier!

>Like many things, learning to be a good writer relies on lots of practice. It's a skill worth the time to acquire, but surrendering to an LLM will undermine any progress.

I assume that only applies to natural language, not computer languages?

The issue with LLMs is not that they write poorly but that we all use the same models and end up with the same non-unique voice. This can come over as a combination of bland, boring, and unauthentic. And it doesn't help that some LLM generated text reads like these LLMs are biased towards a style of communication which embraces a lot of hollow phrasing, disingenuous language, overly enthusiastic language, etc. No surprise, the key companies are based in California in an environment where people that work there probably communicate like that. For them this is authentic. But I'm a non native speaker in Berlin and if I send language like that to my colleagues, clients, or peers, I end up sounding like a jerk.

It's something that is of course fixable. Either with a lot of prompting and direction. Or by training on existing text that you wrote. To some extent some tools already do this. But of course if you only ever let AI generate what you say, how would you even develop your own voice?

On a meta point, I did write all of this myself.

> I'm wary of them, partly due skeptism about their accuracy, but also because it rather misses the point.

Here, this reader wondered whether "to" was omitted intentionally to emphasize lack of LLM involvement. Then, on a second read through to find the sentence again, I noticed the first paragraph is also missing a period on its final sentence. Which just goes to show: we can all use editors and proof readers. (According to the acknowledgements section, nine humans read drafts.) Now, I don't want to intentionally leave typos in my writing, but for me it's preferable to LLM-voice no doubt.

> But I also would prefer rather a genuine human's clunky prose to the smooth but soulless LLM.

Amen. This reminds me of a bit from Kitchen Confidential about how Anthony Bourdain would prefer a simple and plain home cooked meal over an ornate restaurant meal any day of the week. I read because I want to hear what a human has to say. It's the errors and malapropisms and style that give a unique and human voice. Doesn't this contradict my earlier statement about not leaving errors in my own writing? It turns out I'm perfectly tolerant of errors in anyone's writing but my own. (I should probably bring that up with a therapist.)

There's a fine line between using an llm to write a personal blog vs using it to help you come up with an account plan on why you want to sell to a company that will be used internally.

> There are tools out there that claim to be able to tell the difference between LLM-written prose and the old-fashioned biological product. I'm wary of them, partly due skeptism about their accuracy, but also because it rather misses the point

HN uses automatic classifiers. They seem pretty effective. At the same time, I think it's a loosing battle. I'd rather HN either flag for others to ignore/filter/review if they want to, and/or allow people to self-report (ala TikTok)

Eg. I wouldn't mind if a comment on HN was like:

Hey everyone, I coded this little thing:

<ai-generated> some code </ai-generated>

Then users can choose if they want to see posts with AI-generated stuff or not, or something else...

Of course there are always going to be some people trying to game the system. But what are we optimizing for? (a lot of times the 95% unnecessarily suffers because of the mitigation efforts to prevent the other 5% from abuse)

i noticed something similar yet different. because, sadly, most of what i read today is llm output (the chain of thoughts, or the summary at the end of a big chunk or work), i tend to "adopt" this style in my writing, hence, writing as an llm without using one. humans are an amazing creatures at copying stuff and adapting to our environment

Extremely relevant for about half or more of blog posts submitted here. Honestly has become a sad state of affairs

If it puts an end to LinkedIn-speak I'll take it, for a year or two.

What tools or AGENTS.md prompts do you find effective for humanizing LLM output? Default GPT output is acceptable, but Claude is unbearable.

The “Simplified Technical English” fad was pretty effective, but now I just specify:

  - Use short sentences and active voice.
  - NEVER use the phrases "load-bearing", "seam", "unlock", "real", or “learnings”.

My workplace is so AI-pilled now that people have Claude responding to Slack threads on their behalf. It's not even discrete - the messages are literally tagged with a "posted via Claude Code" label, and unabashedly use LLM voice. Sometimes it's unclear to me whether the poster even understands what Claude wrote back for them.

It's infuriating to read, but even more so, it feels disrespectful. It's like they're implicitly saying, "I can't be bothered to respond to you, so just parse this AI slop".

I don't really know how to make a case against it since leadership is very pro-AI usage. Has anyone dealt with something similar and have suggestions?

I never use LLM to write anything, but unfortunately now whenever I am writing something, I am worried, am I sounding like an LLM? Using certain words, makes me worry, it will make my writing like an LLM, so I have started to avoid some words in my writing.

I really don't like this timeline ... sigh

I think the bit about "saying what you write" is what stands out.

I wonder if you took the pure speech to speech models and STT'd the output if it would remove the LLM voice? After all, can you even imagine hearing a human utter aloud some of the nonsense you see spewed by claude? Or give a longform presentation or speech in the form of the essays that are 100% LLM-written?

Maybe there's something in our human brains connecting oration and writing that influences the style of writing that pure-text engines just lack?

> But I also would prefer rather a genuine human's clunky prose to the smooth but soulless LLM.

And I wouldn't.

I look at it like this: LLMs can generate code for those who know nothing about software development and they can generate text for those who know nothing about writing. Asking them to write it themselves will not get us better writing, but rather the opposite. If they were willing to learn, they would be doing that right now instead of using a tool that absolves them from learning. So, as bad as LLM slop may be, it's quite possibly still better than the alternative.

> I'm also suspicious of my own ability to accurately recognize LLM voice.

With much sympathy for the headline this is the central problem.

It's not easy to say what Uncanny Valley looks like, and so we're left with suspicion not certainty.

BTW, I am now 4-5 out of hundreds of my comments since AI detection became a thing, called out for having "AI" tone. I fucking hate being objectified like this. It's not true, although I understand that assertion is hard to refute or support. All I can say is, some people write like some people believe an AI writes.

Just learn to like it.

It's a little funny to me to read this from Fowler, given how enthusiastic Uncle Bob has become about agentic coding (albeit his loop specifies very strict code quality metrics). But yeah, I completely agree with the take. Fowler is, as they say, absolutely right.

> (it's not the em-dashes, it's the load-bearing tapestries)

I suddenly feel an urge to turn this into an image prompt.

most of the comments on this side are boring also

Words come from machine? Work is of machine. I skip human and employ machine

I used to work on how my writing would come across and on writing what people would like. Now I spend a long time on how to cut the LLM voice out of what AI wrote for me (I keep repeating "this sounds too AI, fix it" lol)

I recently read DDIA for the first time, and I was surprised by how enjoyable it was after reading AI slop for a year.

I'm big into AI but hate watching a youtuber when I suddenly realise their script is pure AI garbage, that they didn't even edit out the 'quietly's or the the kickers.THE KICKER is forever ruined. I hate it. The kicker can quietly get bent.

Advocates of the dead internet theory are having a field day with this

Thank you for posting this. I don't have a lot to add on the subject, but I found it interesting and thought-provoking given the age we live in. I find it really hard to tell real human and LLM text apart, even though some people claim they can, calling it "AI slop".

Lately, I stopped reading what people are sending. I'm asking LLM to visualise it for me and explain it to me in simple / concise / precise terms.

This guy sure likes to use his own dashes, just saying.

Yawn. This subject has been discussed ad nauseum. The lines are drawn. I'm over it.

So many "load-bearing seams"...

People have been talking about the importance of curation for more than a decade.

Because many good ideas are not immediately obvious good ideas, in an environment at risk of information overload, we need some way to quickly vet ideas to see whether they are worth investing the time to consider them. (It is often the case that the best ideas require considerable investment of time and effort to be judged as good, and, it's also often the case that this determination can be made only at the very end of such a process.)

“Don't judge a book by its cover,” sure, but, historically, there have been superficial markers of care and attention that we've used to these ends. These are obviously very flawed signals, but, you know, maybe if you took a lot of time to craft a message, I don't feel that bad if I waste a bunch of time trying to decode it. These markers—attention to detail (e.g., spell and grammar checking,) clear-structure (e.g., “bottom line up front,”) and typographical flourishes (e.g., smart quotes and em-dashes and extended unicode like diæreses…)—were an indication of some modicum of quality control and care. I, myself, used to brag about or show-off having set up .Xcompose keybindings (https://wiki.archlinux.org/title/Xorg/Keyboard_configuration...) which was just rare enough across the general population to be noteworthy.

All of this that was once somewhat expensive is now cheap or basically free.

Even if I make a point of going back to a pre-UTF-8 Usenet style (or even intentionally insert run-on sentences or misspellings or grammatical errors,) the cost of an LLM to style-transfer its output to match this is barely much of anything. All it means is I wrote one prompt for the first draft, and then one more prompt for the refinement.

Additionally, some LLM habits are actually good habits. Putting the bottom-line up front (in bold, even!) is a great habit in a business e-mail! Decision-makers genuinely struggle with reading through long paragraphs of detail!

Unfortunately, the problem with the LLM output is that it takes good, valuable, superficial or structural habits that it can cheaply apply, and it drapes those around inane, banal, cloying, muddy, meandering meaninglessness.

Today, we can maybe still judge a book by its cover, because even one more prompt worth of refinement is still too much effort, but I don't think that's amounts to much, since the frequency and loudness of complaints of LLM writing means that model trainers are going to train their models to adapt to whatever signals we use for quick vetting. I am skeptical whether this training will concurrently improve the novelty, innovativeness, or other intangible qualities of the actual message. (With the models, as they exist today, and as we use them today, at least in the field of software development, I see extremely poor capability at generating innovation… Personally, I don't even like the quality of the Python code they spit out, but they sure do type real fast, and, like, pejoratives aside, unexciting, mediocre, corporate-y code does have a real place in the world.)

So what happens when the economics of production in the medium significantly outpace the capabilities of production in the message? The LLM can write a comment nearly indistinguishable from any given human… except when you actually take the time to read through it, it's just inane drivel. Thoughtless and a waste-of-time. And there's basically no way to manage your investment of time and effort to tell one way or the other.

[dead]

[dead]

[flagged]

An important distinction as AI-generated content becomes more common: having a consistent voice is not the same as having a distinctive point of view. The best AI-assisted writing should preserve the author's personality, judgment, and expertise rather than simply imitate their vocabulary or sentence structure. This is a useful concept for anyone building content workflows with LLMs.

[flagged]

[flagged]

Great piece. The "LLM Voice" problem is real — we've been benchmarking 18 free LLM API tiers (Qwen3.8-Max, GLM-Z1-9B, Hunyuan-MT-7B, bge-m3 embeddings, etc.) for consistency and tone variation across providers, and the homogenization is striking. Models from different families converge on similar cadences and filler patterns even at different parameter counts. The one outlier we found: reasoning-focused models like GLM-Z1-9B actually break the mold with chain-of-thought traces that feel more distinct. We publish the live benchmark data at apishare.cc/leaderboard if anyone wants to compare latency/quality tradeoffs across free tiers. Would be curious if others have found specific model families that maintain a more distinctive voice.

[flagged]

Journal entry: Dec 12th, 2027.

My name is Martin the last human organic writer on Earth. I am better than everyone else; my citations in my 4-paragraph article say so.

personally, i'm for LLM and AI usage in writing if it helps the writer communicate ideas, their stories, and triumphs in ways they may never have been able to before. 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.

Just do what I do: capture everything I've ever written and ask AI to cobble together what I said using my own sentences.

No more LLM voice! Every now and then you might get a nonsensical artifact or two, but that's the price you pay for better groceries.

I'll repost a comment I made a while back describing the process I use, which I think balances AI in keeping your voice.

If you care about your voice, don't let LLMs write your words. But that doesn't mean you can't use AI to think, critique and draft lots of words for you. It depends on what purpose you're writing it for. If you're writing an impersonal document, like a design document, briefing, etc then who cares. In many cases (scientific papers, legal documents) you already have to write them in a voice that is not your own. Go ahead and write these in AI. But if you're trying to say something more personal then the words should be your own, AI will always try to 'smooth' out your voice, and if you care about it, you gotta write it yourself.

Now, how do you use AI effectively and still retain your voice? Here's one technique that works well: start with a voice memo, just record yourself maybe during a walk, and talk about a subject you want, free form, skip around jump sentences, just get it all out of your brain. Then open up a chat, add the recording or transcript, clearly state your intent in one sentence and ask the AI to consider your thoughts, your intent and ask clarifying questions. Like, what does the AI not understand about how your thoughts support the clearly stated intent of what you want to say. That'll produce a first draft, which will be bad. Then tell the AI all the things that don't make sense to you, that you don't like, just comment on the whole doc, get a second draft. Ask the AI if it has more questions for you, you can use live chat to make this conversation go smoother as well, when the AI is asking you questions, you can talk freely by voice. Repeat this one or two more times, and a much finer draft will take shape that is closer to what you want to say. During this drafting state, the AI will always try to smooth or average out your ideas, so it is important to keep pointing out all the ways in which it is wrong.

This process will help you with all the thinking involved being more up-front. Once you're read and critiqued several drafts, all your ideas will be much more clear. Then, sit down and write your own words from scratch, they will come much easier after all your thoughts have been exercised during the drafting process.