If using AI to write is nothing to be ashamed of, then you shouldn't feel the need to hide it. If it is something to be ashamed of, then you should stop doing it. If someone objects to you poisoning a well, the correct response is not to use a more subtle poison.
I sometimes like having my content editorialized. Some of the LLM writing tropes are ok to me—I'd delete them if I added this prompt to my instructions (but I wouldn't). But my editorial preferences—the sense of voice and tone I want the LLM to make—are rarely these tropes. Instead, I have a positive prompt of the angles I do enjoy.
However, what is cloying about these tropes for many is that they're becoming empty words. Instead of tack-sharp summaries or reductions to simple understanding, the model is spilling extra tokens for minimal value—I don't need to read "it's not X, it's Y" for the n-th time today. I'd really prefer tighter, more succinct reading that actually directly quotes sources (which modern models rarely do to avoid copyright traps).
This allows me to order them in order of the relevance to start getting my data and information faster.
Applying a constraints like in the published template will make it slightly less awful. It's going to be discarded anyway, but at least the experience is going to be better.
Not every LLM output is going to be published for you to consume. If hazard a guess most never sees the light of the day.
Improving prose to remove predictable patterns is the work of an editor. This process ensures the content is worth reading and respects the audience's time.
Comparing a software tool to "poisoning a well" turns a debate over style into a moral crisis that doesn’t fit the situation. If the information is accurate and the writing is clear, the water in the well is fine, regardless of the pump used to get it there. If the water tastes good, complaining about the plumbing is just a distraction.
The "why would I read what nobody bothered to write" argument only applies to people who ask a bot to hallucinate an opinion from scratch. It doesn't apply to authors using the tool to clarify their own ideas.
I find putting the former into my brain abhorrent to such an extent that I am willing to forego reading the few instances of the latter. I'd much rather have your raw research notes and observations.
> The "why would I read what nobody bothered to write" argument only applies to people who ask a bot to hallucinate an opinion from scratch. It doesn't apply to authors using the tool to clarify their own ideas.
You're wasting my time if you share LLM writing. If you're going to do it that way, share your notes and your prompt. Otherwise, you're being inconsiderate.
Speaking of poisoning wells, have you heard of this thing called Search Engine Optimization? Absolutely ruined the Internet.
For example it ignores the gazillion medium(-like) "articles" that are not much more than the output of a prompt. Here AI is not about style, is about content too. If you open such a post, maybe with the intent of learning anything, and you realize is AI slop, you might close it. Making it harder to recognize is poisoning the well in such cases.
I often hear this here: "if you don't bother writing, why should I bother reading?" In fact, save us some time and just share the prompt.
In a sense I think this is accurate, but not inevitable. I think there is a lack of creative thinking, but it has come from a world that doesn't value it and suppresses difference.
There is a brilliant line in Treehouse of Horrors IV where Principle Skinner says "Now I've gotten word that a child is using his imagination, and I've come to put a stop to it." Which is just the perfect comment on the modern education system.
Models trained on the lack of diversity will push one way, but I think it will also avenues for expression that didnb't exist before. The balance will come from how we react and support what we would like to have happen
A text by a human mind may be seen as a jagged crystal with rough edges and character. Maybe not perfectly written but it's special.
An LLM takes a million of crystals and trims the most likely tokens to be chosen into what would rather appear as a smooth pebble; the common core of all crystals. And everyone using the LLM will get very similar pebbles because to the LLM, regardless who is speaking to it, it will provide the same most likely next tokens. It's not that creativity is lacking in the input, but the LLM picks the most commonly chosen words by all humans in given contexts.
For that to sound imaginative and great as you go, it would have to not only exist in the data, but be a common dominating voice among humans. But if it was, it wouldn't be seen as creative because it would be the new normal.
So I'm not sure how there's a good way out of this. You could push LLM temperature high so that it becomes more "creative" by picking less popular tokens as it writes, but this instead tend to make it unpredictable and picking words it shouldn't have. I mean, we are still dealing with statistical models here rather than brains and it's a rough tool for that job.
I have always thought this is a rather misguided view as to what LLMs do and indeed what statistical models are. When people describe something as 'just statistics' I feel like they have a rather high-school-ish view of what statistics represents and are transferring this simplistic view to what is going on inside a LLM. Notably they do not find the most probable next word. They find the probability of every word that could come next. That is a far richer signal than most imagine.
And ultimately it's like saying that human brains are just chemical bonds changing and sometimes triggering electrical pulses that causes some more chemicals to change. Complex arrangements of simple mechanisms can produce human thought. Pointing at any simple internal mechanism of an entity without taking into account the structural complexity would force you to assume that both AI and Humans are incapable of creativity.
Transformers are essentially multi-layer perceptron with a mechanism attached to transfer information to where it is needed.
If we're being pedantic, they find a* probability for every token (which are sometimes words) that could come next.
What actually ends up being chosen depends on what the rest of the system does, but generally it will just choose the most probable token before continuing.
* Saying the probability would be giving a bit too much credit. And really calling it a probability at all when most systems would be choosing the same word every time is a bit of a misnomer as well. During inference the number generally is priority, not probability.
Most systems choosing the high probability thing is what probability is.
They're just relative scores. If you assume they add to one and select one based on that it's a probability.
It doesn't just have to be one problem.
1. Laundering your "ideas" through an LLM makes them less of your ideas, at best you get the classic two sentences of content embedded in two pages of padding.
2. LLMs removed a filter that help cut down on the amount of useless writing we'd have to wade through. The difficulty of expressing an idea acts as a filter to weed out many (but not all) ideas not worth expressing. That applies to both to people with ideas worth expressing and those without.
On the former, I've had the experience of having an idea, then witnessing it fall apart as I try to express it, as I think about it more deeply. LLMs let you avoid that.
That is an opinion somebody shared on X which has been mindlessly repeated over and over again in other places such as this site.
Why do you value those comments when all they are doing is parroting something they didn’t think themselves? It seems to undermine your point entirely. There is zero originality or effort in those comments. Why are you bothering to read them?
Copying and pasting somebody else’s opinion from one social media site to another is no more virtuous than what you are complaining about.
Sure, I still end up with a polished article, but a lot of it is not entirely my idea or something I would have written through the filter of my own experience. So in order to share my true take on a subject, I have to go through the struggle of writing and bouncing of ideas in my head, which almost always results in a better output.
If someone asks a model to "write a post about X," they are outsourcing the thinking, which results in the homogenized voice everyone is tired of.
If anyone who works on LLMs is reading, a question: When we've tried base models (no instruction tuning/RLHF, just text completion), they show far fewer stylistic anomalies like this. So it's not that the training data is weird. It's something in instruction-tuning that's doing it. Do you ask the human raters to evaluate style? Is there a rubric? Why is the instruction tuning pushing such a noticeable style shift?
[1] https://www.pnas.org/doi/10.1073/pnas.2422455122, preprint at https://arxiv.org/abs/2410.16107. Working on extending this to more recent models and other grammatical features now
Collapsed mode makes the models truncate entire token trajectories, repeat themselves, and indirectly it does something MUCH deeper, they converge on almost 1:1 input-to-output concept mapping (instead of one-to-many, like in base models). Same lack of variety can be seen in diffusion models, GANs, VAEs and any other model regardless of the type and receiving human preference.
Moreover, these patterns are generational. Old ones get replaced with new ones, and the list in the OP is going to be obsolete in a year. This is what already happened to previous models several times, from what I can tell. Supposedly this is because they scrape the web polluted by previous gen models.
IOW, won't code generated by the model have the same deficiencies with respect to lack of diversity?
Interestingly, because perplexity is the optimization objective, the pretrained models should reflect the least surprising outputs of all.
> Why is the instruction tuning pushing such a noticeable style shift?
Gwern Branwen has been covering this: https://gwern.net/doc/reinforcement-learning/preference-lear....
There's also just that weird thing where they're obsessed with emoji which I've always assumed is because they're the only logograms in english and therefore have a lot of weight per byte.
Isn't the instruction tuning done with huge amounts of synthetic data? I wonder if the lack of diversity comes from llm generated data used for instruction tuning.
Wonder how they can avoid the trop while not censoring themselves out.
It also struggles to maintain deep coherence. This is all probably related. It might be very hard or impossible to have deep coherence without human-like goals, memory, or sense of self.
I find AI very "human-esque", and its "self-reported" phenomenology is very entertaining to me, at least.
I also think AI writing might feel trashy also because most human writing is trashy.
> Add this file to your AI assistant's system prompt or context to help it avoid common AI writing patterns.
So if I put this into my LLM's conversation it is like I am instructing it to put this into its AI assistant's system prompt, so the AI assistant's AI assistant.
The alternative is to say:
"Here is a list of common AI tropes for you to avoid"
All tropes are described for me to understand what that AIs do wrong:
> Overuse of "quietly" and similar adverbs to convey subtle importance or understated power.
But this in fact instructs the assistant to start overusing the word 'quietly' rather than stop overusing it.
This is then counteracted a bit with the 'avoid the following...' but this means the file is full of contradictions.
Instead you'd need to say:
"Don't overuse 'quietly', use ... instead"
So while this is a great idea and list, I feel the execution is muddled by the explanation of what it is. I'd separate the presentation to us the user of assistants and the intended consumer, the actual assistants.
I've had claude rewrite it and put it in this gist:
https://gist.github.com/abuisman/05c766310cae4725914cd414639...
An LLM guide would do better to avoid every one of those labels and examples, since the whole point is not to prime the pattern.
Instead each instruction should describe the positive shape of good writing – what a well-constructed sentence, paragraph, or piece actually looks like.
Following this line, here is Claude rewriting OP:
https://gist.github.com/abuisman/05c766310cae4725914cd414639...
// This post’s typography and Oxford commas by human hands.
Also, I sometimes find a sort of Streisand effect: when you tell the LLM to avoid something is starts doing it more. Like, if you say "don't use delve" it contains the words "use delve" which, amongst a larger context, seems to get picked up.
I have more success telling the LLM to write in the style of a particular author I like. It seems to activate different linguistic patterns and feel less generic.
Then, I make an "editor agent" comb through, looking for tropes and rewording them. Their sole focus is eliminating the tropes, which seems to work better.
I'll give some examples. Some will be from this list of "AI writing tropes" and some will be from prominent human-written (prior to 2020) sources. Guess which is which (answer at the bottom).
- "Let's explore this idea further."
- "workload creep"
- "Navigating the complex landscape of "
- "Let's delve into the details"
And I'm not going to get into how silly this is as a so-called LLM trope: "Every bullet point or list item starts with a bolded phrase or sentence." I remember reading paperbacks published before the first PC that used this style.
Fractal summaries is literally how composition is taught to students. Avoiding that style will make the writing more likely to sound less like a person wrote it.
I would suggest the author upgrade this to a modern version of Strunk & White and go on a mission to sell that. Call it Anti-Corpspeak or whatever. But don't pretend that these formulations only arrived in bulk in the last 2-3 years.
ANSWER KEY: these are all obviously prominent in text published before LLMs hit, as well as in the tropes doc. They are no more signs of LLM-generated text than is the practice of using nouns, verbs, and adjectives to convey ideas.
I see what you did there.
Roll thine eyes all thou wish’t but this I promise thee, if thou art lucky there will come a day on which thou shalt speak the single most important sentence in thy life, and that sentence will contain the word “thee”: “With this ring I thee wed”.
Feed that into a slop detector.
> Honestly? We should address X first. It's a genuine issue and we've found a real bug here.
Honorable mention: "no <thing you told me not to do>". I guess this helps reassure adherence to the prompt? I see that one all the time in vibe coded PRs.
But I feel like I’ve noticed an uptick in people using the adverb “genuinely” in what I genuinely believe to not be AI generated comments, articles, etc. Maybe it’s just me, I got similar vibes about the word efficacy a few years ago, before the ascent of GenAI (but after the pandemic — again, maybe just me).
"And honestly? That's rare"
I see this so often. Sometimes it’s just “no react hooks”, other times it gets literal and extra unnatural, like: “here’s <your thing>, no unnecessary long text explanation”. Perhaps we’re past AGI and this is passive aggressiveness ;)
That happens all the time if the previous discussion was about the other subject you don't want (tech in this case): LLMs (not just Gemini) go out of their way to reconcile the two topics.
As an example at some point I asked about the little shrooms people (the tiny people people do hallucinate all mostly the same when eating a particular mushrooms) to a LLM and forgot to begin a separate discussion and asked... About the root "-trinsic" in "intrinsic" and "extrinsic" and the city of "trinsic" in the Ultima game. Oh man... The LLM went wild. I totally forgot I asked about the little shrooms people hallucination but the LLM didn't forget and went totally nuts.
I think you'll get better result if you launch a new discussion and specify "Context: history" or "Context: cooking". Once it goes off the rail, asking it to "not do that" ain't really working: by that point it's just gone, solid gone.
You can give it additional instructions in the settings, but you have to be careful with that too. I've put my tech stack and code preferences in there to get better code examples. A while later I asked it about binary executable formats and it started ending every answer with "but the JVM and v8 take care of that for you."
Which is both funny in an "I, Robot" kind of way, and irritating. So I told it to ignore my tech stack. I have a master's in CS and can handle a bit of technical detail.
Turns out, Gemini learned sarcasm. Every following answer in that thread got a paragraph that started with something like "But for your master brain, this means..."
I noticed the "memory" too and it's turned Gemini into a useless syncophant for me, but so subtle that I almost didn't spot it.
The toggle by "Your past chats with Gemini"
https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing
Another one that seems impossible for LLMs to avoid: breaking article into a title and a subtitle, separated by a colon. Even if you explicitly tell it not to, it'll do it.
This is a problem, because you can easily get stuck in a self-reinforcing loop. You feel strengthened in your convictions that you're good at ferreting out LLM-speak because you've found so much of it. And you find so much of it because you feel confident you're good at it. Nobody ever corrects you when you're wrong.
Combine that with general overconfidence and you get threads where every other post with correct grammar gets "called out" as AI generated. It's pretty boring.
There's a similar effect with contentious subject. You get reams and reams of posts calling the other side out for being part of a Russian/Israeli/Iranian/Chinese troll network. There's no independent falsification or verification for that, so people just get strengthened in their existing beliefs.
Yes. People keep saying, in response to points like this, "oh but you/I can tell pretty easily." But it's not the detection, it's the verification! (see what I did there)
Where I'd push back is the idea that the problem is the boring "call out" discourse that follows each accusation. The problem of verifying human provenance is fundamental to the discussion of trust and argumentation, but the simple "the zone is flooded" problem is also an ecological one. There's terrible air/water/soil quality in the metro area I live in; people have to live with it w/o regard to how invested they are in changing it.
I cannot know what this says about my false negative rate, but at the very least I am confident in my false positive rate.
I honestly don’t know what sites like this will do when that happens and the only way of detecting LLMs is that they are subtly wrong or post too much, we’d be overrun with them.
Not sure if we should be hopefully or fearful that they will improve to be undetectable but I suspect they will.
There's precious little training material left that isn't generated by LLMs themselves.
Consider this to be model collapse (i.e. we might be at the best SOTA possible with the approach we use today - any further training is going to degrade it).
Percentage-wise this is quite exaggerated.
> Consider this to be model collapse (i.e. we might be at the best SOTA possible with the approach we use today - any further training is going to degrade it).
You consider this above factor to lead to model collapse? You’ve only mentioned one factor here; this isn’t enough. I’m aware of the GIGO factor, yes. Still there are at least ~5 other key factors needed to make a halfway decent scaling prediction.
It is worth mentioning one outside view here: any one human technology tends to advance as long as there are incentives and/or enthusiasts that push it. I don’t usually bet against motivated humans eventually getting somewhere, provided they aren’t trying to exceed the actual laws of physics. There are bets I find interesting: future scenarios, rates of change, technological interactions, and new discoveries.
Here are two predictions I have high uncertainty about. First, the transformer as an architectural construct will NOT be tossed out within the next five years because something better at the same level is found. Second, SoTA AI performance advances probably due to better fine-tuning training methods, hybrid architectures, and agent workflows.
> Percentage-wise this is quite exaggerated.
How exaggerated?
a) The percentage is not static, but continuously increasing.
b) Even if it were static, you only need a few generations for even a small percentage to matter.
> You consider this above factor to lead to model collapse? You’ve only mentioned one factor here; this isn’t enough. I’m aware of the GIGO factor, yes. Still there are at least ~5 other key factors needed to make a halfway decent scaling prediction.
What are those other factors, and why isn't GIGO sufficient for model collapse?
Similar to how you can watch one fantastic western/vampire/zombie/disaster/superhero movie and love it, but once Hollywood has decided that this specific style is what brings in the money, they flood the zone with westerns, or superhero movies or whatever, and then the tropes become obvious and you can't stand watching another one.
If (insert your favorite blogger) had secret access to ChatGPT and was the only person in the world with access to it, you would just assume that it's their writing style now, and be ok with it as long as you liked the content.
Overly focussed on style over content
Melodrama even when discussing the mundane
Attention grabbing tricks like binary opposites overused constantly
Overuse of adjectives and adverbs in particularly inappropriate places.
Lack of coherence if you’re generating large bits of text
General dull tone and lack of actual content in spite of the tricks above
Re your assertion at the end - sure if I didn’t know I’d think it was a particularly stupid, melodramatic human who didn’t ever get to the point and probably avoid their writing at all costs.
And yet people seem to still be terrible at that. Someone uses an em-dash and there's always a moron calling it out as AI.
> I honestly don’t know what sites like this will do when that happens and the only way of detecting LLMs is that they are subtly wrong or post too much, we’d be overrun with them.
My personal take is that it doesn't really matter. Most posts are already knee-jerk reactions with little value. Speaking just to be talking. If LLMs make stupid posts, it'll be basically the same as now: scroll a bit more. And if they chance upon saying something interesting then that's a net gain.
Personally, I think it will matter deeply if sites like this are overrun by bots. If you believe your description, why are you here?
That's how a trope starts. When a minority of writers are using a particular pattern, it's personalized style. When a majority of writers in a genre adopt the same personalized style, it's a trope.
We find AI tropes especially annoying because there are three frontier LLMs producing a sizable chunk of text we read (maybe even a majority of text, for some people) lately. It would also be annoying if a clique of three humans were producing most of the text we read; we'd start to find their personal styles annoying and overdone. Even before LLMs, that was a thing that happened in some "slop" fiction genres where a particularly active author would churn out dozens of novels per year in one style (often via ghostwriters, but still with a single style and repetitive plot pattern).
Puffery about "rich cultural heritage, a "tapestry" of sights "from the Colosseum to the Pantheon" and how they "serve as potent symbols" probably is better writing than "Rome is a city in the Lazio region of Italy with a population of 4m. It is the capital of Italy". Doesn't work quite so well when its trying to fit the pattern to the two competing diners of Bumfuck, Ohio and how the rich cultural heritage of its municipal library underscores its status as the third largest city in its county.
It makes a tremendous difference. Almost everything on this list is the emotional fluff ChatGPT injects to simulate a personality.
I can understand someone needing help with writing but getting an agent to do the job for you feels like a personal defeat.
Also whoever claims "no human writes like this" hasn't been to LinkedIn... though the humanity of those writers might be debatable. But all the vapidity, all the pointless chatter to fill up time and space, it learned that from us.
I wouldn't have delegated this to an AI. Human for human, human for AI.
With this I am able to get all my favorite subs onto my actual hard drive, with some extra awesome features as a result: I vibe coded a little helper app that lets me query the transcript of the video and ask questions about what they say, using cheap haiku queries. I can also get my subs onto my jellyfin server and be able to view it in there on any device. Even comments get downloaded.
All these streamers have gone too far trying to maximize engagement and have broken the social contract, so I see this as totally fair game.
sniff sniff I feel like I'm going insane.
Negative parallelism is a staple of briefs. "This case is not about free speech. It is about fraud." It does real work when you're contesting the other side's framing.
Tricolons and anaphora are used as persuasion techniques for closing arguments and appellate briefs.
Short punchy fragments help in persuasive briefs where judges are skimming. "The statute is unambiguous."
As with the em dash - let's not throw the baby out with the bath water.
‘It’s not mashed potato. Its potatoes lovingly mixed to perfection with butter and milk which quietly dominate the carrots beside them.’
The words are in the right order, th grammar is ok, but the subject is so banal as to undermine the melodramatic style chosen and they often insert several per paragraph.
Honestly, the easiest way to verify if a person wrote something is to look for apostrophe use.
This one hit home... the first time I ever saw Claude do it I really liked it. It's amazing how quickly it became the #1 most aggravating thing it does just through sheer overuse. And of course now it's rampant in writing everywhere.
"No rough handling. No struggles to accelerate. Just pure performance. The new Toyota GT. It's not just a car—it's a revolution."
Most of the tropes listed on this page give text a more "car ad" (or sometimes "movie trailer") quality. I wonder if magazine scans and press releases unduly weighted the training set.
You can test this quite easily, by checking and hopefully realizing that you in fact can understand written documents with syntax errors, emails with typos and road signs with improper casing or sentence construction.
If AI finally gets rid of the thing that drove me nuts for years: "leverage" as a verb mean roughly "to use"—when no human intervention seems to work, then I shall be over-the-moon happy. I once worked at a place where this particular word was lever—er, used all the damn time and I'd never encountered something so NPC-ish. I felt like I was on The Twilight Zone. I could've told you way back then that you sounded like a bot doing that, now people might actually believe me and thank god.
I will stick by the em dashes however. And I might just start using arrows too. Compose - > → right arrow. Not even difficult.
I hadn't noticed this - great point. To be fair the "home cooked meal" metaphor comes from 2020, predating genAI coding[1]. But even then, CPUs themselves are so normalised that we just kind of... forget how vertiginously complex the entire supply chain is.
“Leverage” feeling “impacted” much?
>> "How would you organize these LLM quirks, ontologically speaking? I have this notion that the better path is to identify what kinds of things are emerging and prompt to do those things better; accept it as something LLMs are going to do and treat it as something to improve on instead of something to eliminate."
The output is a bit better on blind prompting with applying the results. Here's the gist:
1. Compression artifacts — the model encoding structure implicitly
2. Attention-economy mimicry — the model trained on engagement-optimized writing
3. False epistemic confidence — the model performing knowledge it doesn't have
4. Affective prosthetics — the model simulating emotional register it can't inhabit
5. Mechanical coherence substitutes — the model managing the problem of continuity
Spot corrections are too spotty. Going higher levels with these kinds of problems seems to work better.
If you can convince people that SVO is a distinctly AI pattern it's an automatic win.
One I've seen Gemini using a lot is the "I'll shoot straight with you" preamble (or similar phrasing), when it's about to tell me it can't answer the question.
'you must be mad'. Aggressively hilarious. Love it!
All those tropes have their place in certain contexts. AI overusing them is because they have no memory across all they've written.
Each conversion is a new chat so it's like "I haven't used delve in a while, think I'll roll out that bad boy"
And then you try to fix this by telling it what not to do which doesn't work very well, so...
“We’ve all been there.”
“Your first instinct might be…”
“Now you have a…”
User: "This code gave this error."
GPT: "That's because you did X wrong. Do Y."
User: "You gave me X. And Y is wrong."
GPT: "Honestly, that's great feedback on X. Your Y is broken because Z."
Kind of like enforcing linting or pre-commit checks but for prose.
It does not seem like there are lots of people who are perversely inclined to write a story with all these tropes and words in it, but surely there must be some, because if you make something that beats the LLM (by being creatively good) using all the crap the LLM uses, it would seem some sort of John Henry triumph (discounting the final end of John Henry of course, which is a real downer)
When people use LLMs to generate text, they often ask it to write like a professional. (I haven't tried, but I assume that if you ask an LLM to write like a Reddit troll it will use a different set of forms.)
When you ask an LLM to write like a professional writer, it will aim to sound like a professional writer. They do in fact, and in speech, use words like "delve" and "robust" because they spend years cultivating their vocabularies.
Professional writers are comfortable with punctuation marks and know the difference between the em dash and the en dash, and when to use each versus other marks. (The typical non-professional cannot manage to use the apostrophe, much less the marks that require judgement.)
And a lot of them end up writing business content at some point in their careers. Which leads to an interesting mash where you may get "leverage" used as a verb alongside some of the other pattern tropes.
Because business writing is its own universe. LinkedIn has been swimming in content that would be flagged as LLM-generated for at least 10 years, long before ChatGPT landed.
But that being said, the problem I think is that people treat the output from LLMs as final.
It should be treated more as idea generation or early draft to get over the “staring at a blank page” and get the creative juices flowing and creating your own content.
Having purely AI generated content and eventually feeding the algorithms and soon enough every sounds the same (already does in a lot of places).
But the prompt is usually bereft of fully fleshed out ideas, so the LLM substitutes style in a futile attempt to amplify the signal.
Though maybe it’s not futile! HN voters eat this stuff up daily.
This alone can account for the seeming disparity. Though many people write poorly, they do not write much text for public consumption at all.
* The technical term is "mode collapse", see [1][2]
[1] https://en.wikipedia.org/wiki/Mode_collapse
[2] https://gwern.net/doc/reinforcement-learning/preference-lear...
Imagine a world (ha) where everyone writing on LinkedIn from cafes and couches starts disrupting AI by opting into rating ChatGPT responses.
How might that turn out?
More generally, it's interesting that many different LLMs have differences in their favorite tropes but converge on broadly similar patterns. Of course ChatGPT and its default persona (you can choose others in the settings, but most people don't do that) is overrepresented in these examples. For example, the article doesn't mention the casual/based tone of Grok that often feels somewhat forced.
Show HN: Tropes.fyi – Name and shame AI writing - https://news.ycombinator.com/item?id=47088813 - Feb 2026 (3 comments)
I hope ossa-ma sees this second round!
- “The Pledge”:…
- “The Turn”:…
- “The Prestige”:…
(For this particular example I used real terms from the stage magic world, at least according to Christopher Nolan’s film, as it captures the same meaningless-to-the-uninitiated quality.)
It's a bold strategy cotton. Bold of you to say that. Wild how mundane things get call wild. Thay're making calling things wild their entire personality. In that case, by your logic, (least generous misrepresentation of your logic).
I mean, "tapestry" is a great word for something that is interconnected. Why not use it?
No thanks, I hate this large scale social experiment
The post is moralizing theater masquerading as craft wisdom. “Just write it yourself” ignores the actual quality curve. Give a modern LLM a one-paragraph brief, ask for 600 words, then spend three minutes deleting the three most obvious adjectives and one “delve.” The result is already clearer, better structured, and more grammatically airtight than what 80 % of English-literate adults can produce in twenty distracted, coffee-spilled minutes. That isn’t speculation; it’s what every A/B test in every newsroom, ad agency, and SEO shop shows when copy is anonymized and editors pick winners. The average human twenty-minute draft loses—every single week.
I understand the sentiment. Meaning I think I understand some of the underlying frustration. But I don't care for the tone or the framing or the depth of analysis (for there isn't much there; I've seen the "if you didn't write it, why should I read it" cliché before *, and it ain't the only argument in town). Now for my detailed responses:
1. In the same way the author wants people to respect other people, I want the author to respect the complexity of the universe. I'm not seeing that.
2. If someone says "I wrote this without any LLM assistance" but do so anyway, THAT is clearly deceptive.
3. If you read a page that was created with LLM assistance, it isn't reasonable for you to say the creator was being deceptive just because you assumed. It takes two to achieve deception: both the sender and the receiver.
4. If you read a page on the internet, it is increasingly likely there was no human in the loop for the article at all. Good luck tracing the provenance of who made the call to make it happen. It might well be downstream of someone's job. (Yes, we can talk about diffusion of responsibility, etc., that's fair game -- but if you want to get into the realm of moral judgments, this isn't going to be a quick and tidy conversation)
5. I think the above comment puts too much of a "oh the halcyon days!" spin on this. Throughout history, many humans, much of the time, are largely repackaging things we had heard before. Unfortunately (or just "in reality") more of us are catching on to just how memetically-driven people are. We are both individuals and cogs. It is an uncomfortable truth. That brainwashed uncle you have is almost certainly a less reliable source of information than Claude.
6. The web has crappy incentives. It sucks. Yes, I want people to behave better. That would be nice, but I can't realistically expect people to behave better on the web unless there are incentives and consequences that align with what I want. The Web is a dumpster fire, not because of bad individuals, but because of system dynamics. Incentives. Feedback.
7. If people communicate more clearly, with fewer errors, that's at least a narrow win. One has to at least factor this in.
8. People accusing other people of being LLMs has a cost. Especially when people do it overconfidently or in a crude or mean manner. I've been on the receiving end. Why? Because I write in a way that sometimes triggers people because it resembles how LLMs write.
* I want to read high quality things. I actually care less if you wrote it as bullet points, with the help of an LLM, on a napkin, on a posterboard ... my goal is to learn from something suited to some purpose. I'm happy reading a computer-generated chart. I don't need a human to do that by hand.
The previous paragraph attempts to gesture at some of the conceptual holes in the common arguments behind "if you want a human to read it, a human should right it": they aren't systematically nor rigorously "wargamed" or "thought-experimented"; they are mostly just "knee-jerked".
I am quite interested in many things, including: (1) connecting with real people; (2) connecting with real people that don't merely regurgitate an information source they just ingested; (3) having an intelligent process generating the things I read. As an example of the third, I want "intelligent" organizations that synthesize contributions from their constituent parts. I want "intelligent" algorithms to help me focus on what matters to me. &c.
If a machine does that well, I'm not intrinsically bothered. If a human collaborates with an LLM to do that, fine. Whatever. We have bigger problems! Much bigger ones.
Yes, I want to live in a world where humans are valued for what they write and their intrinsic qualities, even as machines encroach on what used to be our biggest differentiator: intelligence itself. But wanting this and morally shaming people for not doing it doesn't seem like a good way to actually make it happen. Getting to that world, to my eye, requires public sense-making, grappling with the reality of how the world works, forming coalitions, organizing society, and passing laws.
Yes, I understand that HN has a policy that people write their own stuff, and I do. (See #8 above as well as my about page.)
Thank you to the approximately zero or maybe one person who made it this far. I owe you a beer. You can easily find me. I'm serious. But then we have to find a way to have a discussion while enjoying a beer on a video call. Alas.
I expect better from people -- and unfortunately a lot of people's output is lower quality than what I get from Claude. THIS is what pisses me off: that a machine-curated output is actually more useful to me than a vast majority of what people say, at least when I have particular questions to ask. This is one or many uncomfortable realities I would like people need to not flinch away from. As far as intelligent output is concerned, humans are losing a lot of ground. And fast. Don't shoot the messenger. If you don't recognize this, you might have a rather myopic view of intelligence that somehow assumes it must be biological or you just keep moving the goalposts. Or that somehow (but how?) humans "have it" but machines can't.
I don't write for sentimentality. I write so that my code designs can survive longer than my work on it.
No documentation is worse than deceit.
The emptiness and vastness of the void (entropy) is much deeper than humans or machines.
Google search says this philosphy is called https://plato.stanford.edu/entries/content-externalism/
If we want our systems to last, we would need the "process knowledge"—the actual mastery of the craft—to be in human hands rather than decaying in a dead system.
I don't think we can afford to process-knowledge-transfer many of our essential systems... without machine assistance.