Literary Clusterfucks 2013

Beginning to think maybe a rolling thread might be good. Anyway:

So Alisa Valdes just publishes this memoir about this cowboy of hers and how he's a man's man and now's she's a woman like never before and etc. That link's to Hanna Rosin's review, and she's essentially going "Um...you sure?"

And then Valdes publishes this today.

That said, a lot can happen in two years, especially when you’re in a relationship with a man as complicated and volatile as the cowboy.

"what I actually wrote was a handbook for women on how to fall in love with a manipulative, controlling, abusive narcissist."

just what the world needed. like a poke in the eye.

jesus

https://www.youtube.com/watch?v=FORoSB5JxCU

"Cowboy up."

none of these people are really writers

― puff puff post (uh oh I'm having a fantasy), Friday, June 4, 2010 1:18 PM (2 years ago) Bookmark Flag Post Permalink

wtf @ that whole story

polo shirt under a jacket, tho

I see by your outfit that you are a cowboy.
You see by my outfit that I'm a cowboy, too.
We see by these outfits that we are all cowboys.
If you buy a cowboy outfit, you can be a cowboy, too.

damn nm I just actually read this shit how f'd up

i keep reading her name as alida valli

jesus

I don't know if we do this around here but there's some srs abuse and sexual assault triggers in Ned's second link

it's probably a good thing that she's posting a picture of her rapist on the internet, now we can watch out for him

man that's a tough read

cowboyfucks

How much you wanna bet that Mr Cowboy is gonna be a MRA talking head

In 2001, Valdes emailed a 3400-word resignation letter to her superiors at the Los Angeles Times. The letter was widely circulated on the Internet[ and reprinted in the St. Petersburg Times. In the letter she accused the newspaper of racism and discrimination, especially in its synonymous use of the word "latino" with "Spanish-speaker", a practice she equated to genocide.

buzza are you suggesting that alisa valdes is hysterical or otherwise to be dismissed

this is not the elizabeth wurtzel thread

And so, even though I was 43 years old and have Lupus

WHY DO THESE FANFICCY CRAP ROMANCE NOVELISTS ALWAYS HAVE LUPUS OR FIBRO WTF.

hm prob should have read the whole post of hers before making light. still, this shit brings the whole 50 shades bullshit into its awful, true light.

Considering Valdes wrote one of the single most amazing demolitions of a horrible person ever, reading/seeing all this...yeah.

okay that was awesome. thanks for linking that Ned

still cannot get over her story with the cowboy. so fucked up. I mean, just that it reads so familiarly, is so sad to me.

this is not the elizabeth wurtzel thread

oh, okay then

“An irresistible, post-feminist Taming of the Shrew. Don’t be scared by the premise. This is not a story about a woman relinquishing her identity. Quite the opposite. It is a riveting tale about how a brilliant, strong-minded woman liberated herself from a dreary, male-bashing, reality-denying feminism.”

– Christina Hoff Sommers, author of The War Against Boys; How Misguided Feminism is Harming Our Young Men

a practice she equated to genocide

ugh fuck this, there is like a 100% chance she was referring to cultural genocide, a term used for decades and not meant to imply the actual murder of a group of people

http://www.sptimes.com/News/110300/Floridian/The_language_of_genoc.shtml

buzza idgi are you trying to damage the credibility of the woman who basically just announced she wrote a book about a man who raped her

apparently so?

it's tough when you can only speak in the form of revived threads

seemed like zachylon wanted the context of the wiki quote so i provided it?

thank you for posting it

she does make a clear distinction between the two types of genocide tho she doesn't mark it with "cultural" or something similar. she does 'equate' the two but that's sort of the idea, while the wiki editor left out any of that context and framed it like "she compared this one tiny linguistic choice with the holocaust", fuck wiki

"literary"

I have a lot of thoughts about this whole thing and also some feelings but none are organized enough to share except for, Jesus, Lady--at least when I did that I didn't write a book about it.

considering that one of the major goals of feminism was to protect women from the power imbalances present in domestic relationships, it's not super surprising that valdes' paean to how feminism got romance wrong and how there's something special about a real man turned out to be about an abusive asshole. i don't mean to suggest that she deserves what happened in the least, but there is a sort of irony that the very political principles she decried in the context of this relationship turned out to be especially relevant to her needs.

xp On second thought that makes it sound like my experience was as extreme as hers: it was not. I also didn't put it in those terms of submission etc or posit that it revealed anything about how feminism has failed us. And I didn't have to jump out of a moving truck although after getting hit by an actual car frankly I'd take another one of those accidents over another of those relationships.

there is a sort of irony that the very political principles she decried in the context of this relationship turned out to be especially relevant to her needs.

It's not like that's a coincidence. She decried them because she was being told to.

really want some blogger to try to get a reaction out of christina hoff summers

I usually assume "How I did X and Changed My Life" memoirists are flighty, superficial and unrealistic people, because shit just doesn't work like that. This is a particularly egregious example.

This is a horrible horrible story.

btw, I regret my above post, having apparently made it without really reading most of the story in her blog post.

However, the blog post is now gone.

?!

this just got a little clusterfuckier

without the followup blog post this is kind of incoherent

she wrote a fluffy romance novel that seems to spend half its time scolding modern feminism, then revealed that the man she was writing about raped and abused her and (this is where things are fuzzy to me) the whole novel was a double-feint?

it's a memior!

ergh

so, replace "romance novel" with "memoir"; is the rest accurate?

Yup.

huh

Given some of the things she was also saying about her publisher I wonder if that had something to do with it.

I feel like we are rapidly approaching the literary clusterfuck equivalent of "where is the downbeat." Fascinating to read what other people read (or don't)

― Cattedrale metropolitana di Santa Maria de Episcopio, Wednesday, 20 May 2026 bookmarkflaglink

William Empson.jpg

You jest, but I think a lot of the AI LLM junk is dealt with very neatly of Empson's classic diss from 7 Types of Ambiguity:

They admired the poetry of previous generations, very rightly, for the taste it left in the head, and, failing to realize that the process of putting such a taste into a reader’s head involves a great deal of work which does not feel like a taste in the head while it is being done, attempting, therefore, to conceive a taste in the head and put it straight on to their paper, they produced tastes in the head which were in fact blurred, complacent, and unpleasing.

*dealt with very neatly by

It has been bizarre to see people claiming that the author is a fake identity set up as part of a long-con by an AI company. There’s an interview with him from 2018, which is also potentially quite relevant to the discussion around ‘standard’ and ‘vernacular’ English.

It really does look like a guy who left a career a while back to do a bit of consulting work, got bored and started writing vast amounts of poetry.

https://i.postimg.cc/SNNS7W2w/IMG-0739.jpg

Where's the birth certificate, though?

Thanks for all the context/info these past few days, ShariVari.

The only time Indo-Trinidadians are ever in the news is when Nicki Minaj does something stupid so it’s fun to have something else to talk about!

I don't think the author is fake, I think the author isn't very good at what he does. The poems that I've read are...not good.

Absolutely.

Supposing that to be what the author meant by "apologizing to furniture"

Oh, no. I mean, I'm only inferring what the author meant, but I meant literally apologizing to furniture - you know, an instinctive 'sorry' after knocking into an empty chair? You have possibly not seen it, or not noticed, but it codes much as you might imagine: deference, shyness, caution. What I took from that phrase is that she's not self-conscious about her size - as you say she'll probably knock more furniture than Sita, but she'll apologise less, she's very comfortable in being her. And then we get to her acting and talking, and that fills in the details, but within that very clear impression that I first got from that phrase. Now could some of that be fitting to an unintended reading? Sure of course, that's always a possibility.

(sorry, had that open in a tab somewhere all day)

"walking that made benches become men" is an incredible line

if it was a line in a john ashbery poem or garielle lutz story i wouldn't bat an eye

Dude totally exists and has had a book of his poetry for sale on Amazon since 2018 with reviews of it from readers in 2018.

Doesn't prove if this particular work was A.I. or not but I'm kind of sick of everyone appointing themselves internet cop. I too was baffled by the inane number of BlueSky posts gleefully insisting this guy doesn't exist

kind of interesting to see his pro-AI facebook posts tho

He went into IT consulting when he left the Civil Service and I think a lot of people use pretty generic AI talking points on Facebook and LinkedIn to make it sound like they’re up to speed with transformation, etc.

The only semi-interesting thing that has come out recently is that he writes on his phone using voice to text due to ill health / disability.

I think that's a choice tbf, not everyone does that. LinkedIn is a choice. Though I haven't really followed this guy, tbh I could say more on this thread, especially about the ideas people have about MFAs and "MFA writing" and how swiftly they change according to the conversation that's being had, while making some criticisms of my own, but feel we kind of did it with Ocean Vuong.

I am fairly sure, from memory, if you look back at this prize over the years loads of the winners are as bad as this. Or if you look at certain other literary magazines.

I think there's sometimes a p interesting disconnect between the deeper more independent parts of literary fiction and the rest of humanity, even people who read books constantly but don't actively seek to be part of that scene or engage with it.

There are some seriously hallowed works and styles that provoke dismay, boredom or horror even among people who like reading, even those who would define their tastes as literary fiction, but... again we probably touched on this before.

I think the fiction industry (and I mean industry) is more deeply insecure and confused than other parts of the arts and so AI is gonna be tricky when agents themselves are saying they haven't a clue what people like (or what's good).

I think this clusterfuck, whether you like the piece or not, is a good example of how the industry sells a certain type of writing based on the identity of the writer. It is not so keen to let writers deviate from that. That has a weird effect on the internet where people can treat that type of writing as a proxy for the identity of the writer.

I know the discussions here are less Bluesky than that but I still think it's an interesting subject.

Again, kinda touched on before and prob itt.

LinkedIn is a choice.

brb going to get some t-shirts printed

Could promote them on LinkedIn.

LinkedIn
LeanedIn
LockedIn

million seller

https://www.reddit.com/r/RSbookclub/comments/1thab8t/comment/ommnmcs/
https://www.reddit.com/r/RSbookclub/comments/1thvvkw/comment/omqmb8h/

Can't stop thinking about these guys, I love them so much - Pareidolia is obviously a thing with anything relating to AI, it warms my heart that they are committed to the same business but the other way around - don't let the machines take away the jobs of misidentifying things as AI!

There was an insane substack post where the author said he could identify where a bunch of three word phrases in the Granta story had been plagarized from--but the sources were things like "Better Call Saul" fanfic.

Genuine question Andrew, how much A.I. generated literary prose have you read?

Remember when it came out that Dylan’s Chronicles book had lifted weird bits from loads of different sources, including rando self-help books?

Dylan was scraping before AI

Love and theft >>> just theft

"She caused a lot of *beaches* to become men" could be in a possibly promising sequel to/takeoff from "The Girl From Impanema": an example of AI's own possible usefulness, if you tweak it.

Also, I heard something on npr a while back: a woman was trying to write a profile of her late sister, and use AI to help get though depressive fog--AI kept inserting, "And we used to go to soccer!"---but she ditched all such idiot fabulism, repeating her themes, getting closer to the kind of phrasing she wanted, teaching the AI, and it began to give her some clarifying choices, useful points of departure for her own responses, But I think that she said she was never going to use it again, didn't otherwise feel the need.

stylometrics has always seemed like massive bs to me. it's interesting that (ai text detector) pangram seems to work for now, but you gotta wonder how long that can last

doesn’t seem very discriminative to me tbh

saw someone use it on a new yorker article that had it at 100%

There was an insane substack post where the author said he could identify where a bunch of three word phrases in the Granta story had been plagarized from--but the sources were things like "Better Call Saul" fanfic.

If it's the same thing I read, the claim was that AI had randomly hoovered up those passages and spat them back out. But I'm not sure that's how it works, does it? To the extent that it generates language through a kind of predictive text mechanism, doesn't it use the entire set of training data to produce plausible word sequences based on the statistical likelihood of those words appearing together, rather than cobbling together unique strings of words?

This is the Substack:
https://tuhinchakrabarty.substack.com/p/ai-slop-grantagate-and-bad-writing

Yeah, the reactions I saw to it were mostly “This seems like BS.”

So here's something funny from the world of music books.

Melvin Gibbs (bassist for Harriet Tubman, formerly of the Rollins Band, Ronald Shannon Jackson's Decoding Society, and many, many other situations) wrote a book called How Black Music Took Over The World. It's really good - part memoir, part cultural history, part hands-on guide to playing/making groove-based music. I loved it, and wrote about it here.

Allen Lowe, a musician and scholar whose professional frustrations (nobody buys his records, and even fewer people read his books) have turned him into kind of a crank, thought a book called How Black Music Took Over The World should be about the history of black music - that it should start with minstrel shows and progress through Louis Armstrong, Duke Ellington, and blah blah blah. He grumbled about this on Facebook, then published a review for JazzTimes that wasn't a review at all - instead, it was a description of the book he, Lowe, would have written under that same title.

Gibbs took note, and got JazzTimes to publish a rebuttal. And Allen Lowe has been crashing out ever since. I think he's posted nine or ten times on Facebook since Gibbs's piece went live, and has even posted a photo of himself with duct tape over his mouth. For he is being silenced, you see.

Allen Lowe has a very large record collection, but he is a gaping asshole animated by two core beliefs: first, that he knows more about American music and history and race than you do, and second, that every other writer on earth is a bought-and-paid-for shill in thrall to the Jazz Industry. So seeing someone tell him, "Shut up, dude," is very welcome.

doesn’t seem very discriminative to me tbh

― k3vin k., Tuesday, 26 May 2026 18:13 (yesterday) bookmarkflaglink

saw someone use it on a new yorker article that had it at 100%

― k3vin k., Tuesday, 26 May 2026 18:18 (yesterday) bookmarkflaglink

everything evaluation i’ve seen has been very impressive. this paper did an independent evaluation of it and competitors:

Artificial intelligence (AI) tools are increasingly used for written deliverables, e.g., ensuring assignments were completed by students, product reviews written by actual customers, etc. This has created demand for distinguishing human-generated text from AI-generated text at scale. A decision-making aiming to implement a detector in practice must consider two key statistics: the False Negative Rate (FNR), which corresponds to the proportion of AI-generated text that is falsely classified as human, and the False Positive Rate (FPR), which corresponds to the proportion of human-written text that is falsely classified as AI-generated.We evaluate four leading detectors—Pangram, OriginalityAI, GPTZero, and RoBERTa—on their performance in minimizing these statistics using a large corpus spanning genres, lengths, and models. Commercial detectors outperform open-source, with Pangram achieving near-zero FNR and FPR rates that remain robust across models, threshold rules, ultra-short passages, "stubs" (≤ 50 words) and ’humanizer’ tools.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5407424

xp Lowe is his own worst enemy. I appreciate his encyclopedic knowledge and he’s led me to discover some rewarding black music I never would have on my own. But yes he gets into snits.

^this

Man that Gibbs rebuttal was brutal

The majority of people using AI detection software are using whatever cheap programs they find after a cursory Google, plus I can't get past the hypocrisy of using AI tools to determine if AI was used.

I've read many a post from fatigued TAs who said they've had AI detectors turn up 50% AI-written results on papers that were written before LLMs existed.

Part of the issue i have is the lack of falsifiability. The studies above involve an experiment where the conductors know the AI status of the works used. However, unless some watermark or digital artifact is left behind in the document definitively proving it was generated by an LLM,the tools are basing the results on style-based assumptions and not actual evidence.

As such, unless you kept meticulous records of your writing process from start to finish, how can you disprove it? And even if you have THAT, how do you prove you didn't use an LLM to write it and then transfer it to your word processing software.

I've heard stories of a Ph.d candidate who had to redo his dissertation 5 times because the AI detection tool kept popping "likely AI" results even though he swore he keyed every character manually.

LLMs are a big problem but I feel like we've started nuking innocent people in efforts to fight it. Like there's the presumption of guilt in academia where students have to prove their innocence rather than vice versa

And then there's this

AI-Detectors Biased Against Non-Native English Writers | Stanford HAI https://share.google/ZxZarB6POr30yohrw

plus I can't get past the hypocrisy of using AI tools to determine if AI was used.

takes one to know one

I'm not a duck but I'm fairly sure I could identify one

Actually I might be a duck

RETVRN (TO ORAL TRADITIONS)

xp try this game someone made where you distinguish real vs ai texts and post your score https://fakewriters.onrender.com/

i got 5/10

5/10 here, too.

I got 7/10. Not quite Voight-Kampff but I'll take it. I couldn't identify AI Edith Wharton, AI Agatha Christie, or AI Mary Shelley

Interesting statistics on those results. Most folks can tell the difference between real and fake Tolstoy, but Arthur Conan Doyle fools 3 out of 4.

9/10. TBH, the only one of those authors I've read semi-recently was Melville a couple of years ago, it seems like (this) fiction AI has a noticeable style that shows through. No matter who it's trying to imitate there's a kind of clipped cadence like it was trained on a bunch of bad Hemingway knockoffs.

also is the AI trying to mimic Tolstoy via Constance Garnett or Pevear/Volkhonsky or etc.