Posted on 08/29/2026 9:04:49 PM PDT by SeekAndFind
A GHOST WRITER is haunting the English language. The linguistic spectre can turn its hand to prose, poetry, journalese and corporate jargon. It is frightfully versatile: you can get it to mimic Shakespeare’s sonnets or a schlocky beach read; Ernest Hemingway’s taut prose or the office-printer manual. It is frightfully fast, churning out thousands of words a minute. (Hemingway rarely produced as many in a day, and required much more booze.) Wordsmiths are spooked.
AI writing is everywhere. It is in your inbox and on your LinkedIn feed. It is all over the internet, drafting more than a third of new websites by one count. Large language models (LLMs) are helping students write essays and probably helping scientists write papers. Some allege AI-generated prose won the Commonwealth Short Story prize this year, with judges praising its “quiet authority”. (The Commonwealth Foundation denied the claim.)
LLMs have stylistic quirks. They are thought to maximise the use of long em-dashes—and the use of words like “maximise”. They like to “deep dive” (and, better yet, “delve”) into the “rich tapestry” of the world. AI writing is not about a single word or phrase, but a rich tapestry of things.
Spotting AI texts can be tricky. This is in part because you need evidence beyond a few words or dashes: claiming that a text is by an LLM because it uses the word “delve” is like claiming one is by Jane Austen because it uses “imprudence”. Bots also write in slightly different ways.
There is no single style of AI writing, explains Karolina Rudnicka, a linguist at the University of Gdansk in Poland, just as there is no single style of human writing. Writers have idiosyncrasies—Emily Dickinson, for instance, loved em-dashes—and bots may do, too. But there are a few ways to identify LLM-generated text.
One is to use detection algorithms that are trained to spot the texture of human or AI prose. Pangram, a leading firm, claims to have 99.98% accuracy. (It has partnered with Substack, a blogging platform, on such a tool.)
Detectors, however, are black-box algorithms that can give false positives. They do not give reasons for why they reach their conclusions.
Researchers have also tried scouring texts for suspicious words or comparing papers from before and after LLMs were made available to the public. But these approaches have drawbacks too, not least because it is hard to disentangle AI quirks from other language trends.
You can discover AI’s hallmarks by comparing the writing of man and machine. To do this you need a baseline that is distinctive and familiar. The Economist turned to prose that we’re sure is human and that readers will recognise: our own. We designed a study to ask top LLMs—OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini and xAI’s Grok—to write versions of our articles without consulting the web. (As a prompt, we gave them the AI-generated summaries that we have experimentally added to some of our articles.)
This gave us a corpus of human and AI creations and we compared them across 55,940 sentences and 1.2m words. To make sure we were detecting AI quirks rather than our own, we also checked the AI texts against journalism from CNN, the New York Times and the Washington Post. Excerpts from hit novels published between 1950 and 2022 offered another test.
Our findings are surprising. AI prose is distinguishable by word and punctuation choice as well as sentence and paragraph structure. But its hallmarks are not what you might expect, partly because its writing style has changed with software updates. That does not mean that LLMs are great writers: their prose lacks lucidity and elegance and is often formulaic. So those aspiring to be impressive (human) storytellers should avoid the following peculiarities in their own prose.
First, consider words. The vocabulary that bots overuse has changed: they no longer “delve” and there are not as many “tapestries”. Instead they offer a significant number of polysyllables like “significant”, “increasingly” and “consequences”. They use more rare words (“interdependence”, “reindustrialisation”) and scientific lingo (“parameter”, “methodology”) than humans, and are fond of nominalisations (making nouns from verbs, such as “expansion” from “expand”). All the LLMs in our study use such words, but particularly Gemini and Claude.
Much of this language could be described as what George Orwell called “pretentious diction”. He railed against writers who “dress up simple statements” with complicated words and jargon to sound clever. Such pontificating penmen, Orwell observed, also believe that “Latin or Greek words are grander than Saxon ones”. (Bots agree: more Latinate suffixes crop up in their writing than in human texts.)
Then look at punctuation. Many believe LLMs stuff their prose with em-dashes, but that is not true after the most recent updates. Today only Claude uses more em-dashes than human writers, with ChatGPT using markedly fewer than any other writer in our study. Humans rejoice—and start using dashes again.
A better way to spot AI-generated writing would be to look for texts without much punctuation at all. LLMs are very Joycean about it: they use fewer commas and semicolons than humans (and hardly any parentheses). They use less punctuation in part because they write longer sentences—“and” is their most overused word—and in part because they do not quote experts.
Finally, study the sentence. Bots’ sentences tend to be long; paragraphs are rarely interrupted with short, punchy statements. How dull. When LLMs want to make their sentences more lively, they often reach for a rhetorical device. Their favourites include: “not X but Y”, “not only but also” and the “rule of three”. (Grouping ideas in threes makes them more engaging, as we did just then.) ChatGPT and Claude use more of these constructions per 1,000 sentences than other LLMs and humans.
So if you want to spot AI writing, look for bland, pretentious prose lavished with Latinate words—at least for now. With every update, our study shows, AI writing is becoming more similar to human prose. Pangram successfully detected AI-generated copy, but may struggle in future. LLMs are trained on human writing and learn from human feedback, notes Tommie Juzek of Florida State University, picking up things people find impressive and dropping things they do not.
Bots learn fast, too. Take ChatGPT: until very recently it used an em-dash in almost every sentence. When your correspondent asked an older model whether it thought AI overused the dash, it said: “Ha—great question!” Ask the bot the same thing today and it soberly says “they’re best used sparingly.” Only a ghost could shapeshift so quickly. ■
|
Click here: to donate by Credit Card Or here: to donate by PayPal Or by mail to: Free Republic, LLC - PO Box 9771 - Fresno, CA 93794 Thank you very much and God bless you. |
Just ask it to put a word in white on white paper and you’ll know.
I have had a few people accuse me of being an AI Bot on Free Republic due to my lengthy posts.
I am guilty as charged for my lengthy posts, and it is true I occasionally (rarely) use anything from AI, and when I do, I try to explicitly say so.
But I find it somewhat humorous to be accused of being an AI Bot...:)
College professor busted 32 out of 35 students using AI on a final — after they all made the same lazy mistake
The article hints at a stylistic quirk I see (but more often hear) everywhere.
The pattern starts with an assertion:
“Left handed artichokes soon found out that they could alter the results”.
NOT by abcxyz <———
NOT by defghi <———
But by jklmnop
There are always two post-assertions starting with “not——
Notice it, and you’ll hear it everywhere in AI-generated video narrations.
Watch any YouTube video that claims to be a historical, narrated documentary of any sort. Between 5 and 8 of the assertions made follow this pattern.
Years ago I read something about how they can compare one's writing and style and match it with other stuff that is “anonymous”. I often write stuff on our “anonymous” Free Republic and think “Oh - they'll be able to pick me out from that sentence.”
That's okay - keeps me polite. And legal.
I am vindicated! I
ssldapq34[a-fwer[eqe54[t[g mccq34[toin trqyweuritokfnwfbdoq3w4diyyg;’welvrgkwrenj;kfbqwlekhjfbwkjn;elkm;ftjrnejbfen;trlkgmrt’;ylj,rt e6r.t5rkm lgewk;lektmyh’rlegknj;kj bhtkrgjnk;jg
[BREAK]
[BREAK]
[END]
[RESET]
Oops. Sorry! it gets away from me sometimes!
Use of the word “beloved” has become frequent.
An em dash is a long dash. It functions more like a strong break, where a regular dash or hyphen connects compound words.
They didn't start building it until 1963.
If you paid attention in grade school you should be able to pick out AI slop rather easily.
A lot of thread sources are AI, especially that from India. (India Times)
“...translate to Russian, then Korean, then back to English.”
So like the owner’s manual on my dishwasher?!
When calculators came out in the 70s my college math instructor told us that it’s game over. He said that while college math degrees will still be offered in the future, hand held calculators would now do 99% of all math calculations. 30 years later search engines and Google in particular started thinking for us. And now AI is here. This time, it’s really game over. Critical thinking will disappear in one generation.
So how long will it take for programmers to eliminate these AI tells? By the end of the year at the latest, they’ll be gone.
Figures with your name you would know that!
Thanks. I thought they were both just a dash and used for different purposes.
From AI (of course):
Typewriter Keyboard Limits Single key: Standard typewriters featured only a single hyphen/minus key (-).
No dedicated em dash: They lacked keys for the long em dash (—) or the medium en dash (–).
How Writers Made Long Dashes The double hyphen: Typists created a long dash by typing two hyphens side-by-side (—).
Publishers’ role: Professional typesetters and book publishers would look at the double hyphens in a typed manuscript and convert them into a single, proper long em dash during printing.
******************************
Had to look up on what the heck a “medium en dash” is.
Main Uses
Value Ranges: Used between numbers, dates, or times to mean “to” or “through” (e.g., 1980–1990 or pages 10–20).
Scores and Results: Used to separate competing scores or results in games (e.g., the Yankees won 5–2).
Connections and Directions: Used for paired or linked terms and routes (e.g., the London–Paris train).
**********************************************
I spotted AI writing on Amazon while buying cat food for Daisy. AI advised me not to buy an expensive brand because Daisy loves gravy, missing in new brand. . Bought it anyway, The minute I openedthe first can, she was on the counter licking out licious gravy. scads of it.
Score so far, Cats 100, AI 0
be fair, AMZ makes it clwear that AI is making statements.
(I was on an unscripted road trip from New England to burn up vacation time I was going to lose.
I was going to stay at a horse ranch on the Blue Ridge Parkway near Beford, VA (home to the tragic "Bedford Boys", 20 men from that small town who died on Omaha Beach on D-Day) that my wife and I had previously visited, but it rained so much that there was no horse riding to be done, and the owners even left me alone for that day and the next to leave town! (Heh, they agreed to let me bag out on the last day and not pay for it) and when I left, the rain was coming down so hard that in places I nearly had to pull off the road because I couldn't see, the James River was the color of light milk chocolate with all the soil being washed into it, and I saw waterfalls and beer cans floating down the white water on each side of the Blue Ridge Parkway!
Anyway, I was on my way to the Four Corners Monument, but...as I ate dinner in Louisville, I realized I just didn't have enough time. So I closed my eyes and put my finger on a map of Nebraska, and landed on Lincoln. So when I started to plan, everything I read about Lincoln seemed to indicate it was best known for a "transgender" from that city who was murdered in that state.
I picked up my finger and put it down with my eyes closed on Tecumseh! It was far more interesting, being as small as it was, and named after the famous Indian from the original name of Francis, supposedly the mistress of Col. Richard M Johnson, but the priggish townspeople changed it out of spite against him!
Or, at least that is the story.
Anyway, when I passed the St. Louis arch, it was far, far smaller than my brain had visualized it! (I was warned not to stop there at night, as it was a dangerous place to get out of the car, so I saw it from the highway!
Heh, for some reason, I felt cheated which probably isn't fair!
I was thinking that same thing. How long will it be before using AI on a test in a college course will be accepted?
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.