Grilled Cheese

ExploreLog inSign up

Vauhini Vara

@vauhinivara.bsky.social

434 Following1.5k Followers

author of SEARCHES, THIS IS SALVAGED, & THE IMMORTAL KING RAO. pulitzer finalist. journalist. mentor at the lighthouse writers workshop and look2justice. contact me: dm, signal (@vauhinivara.82), or my website (vauhinivara.com).

PostsRepliesMedia
@vauhinivara.bsky.socialOct 11, 2026, 3:45 PM

YES!! Thanks for saying that. I believe it really firmly!

@vauhinivara.bsky.socialOct 11, 2026, 2:42 PM

if you use the contact form on my website it will come directly to me! Or signal — my details in bio.

@vauhinivara.bsky.socialOct 11, 2026, 2:40 PM

I wrote about how the winner of another major literary prize appears to be AI and why we need to do something about all this. 😢

@vauhinivara.bsky.socialOct 9, 2026, 3:21 PM

I'm really annoyed by my repetition of the word "fascinating" in my last two posts.

@vauhinivara.bsky.socialOct 9, 2026, 3:05 PM

Thank you, @isabelcorpus.bsky.social, for your fascinating work and for responding here!

@vauhinivara.bsky.socialOct 8, 2026, 4:12 PM

Yeah, I saw that note in the study, and it makes sense! I'd be curious to know (including if there's any research you can point me to) whether the underlying finding of AI being more persuasive still holds — like, whether, as there's more exposure to it, it's becoming less effective at persuasion?

@vauhinivara.bsky.socialOct 8, 2026, 3:52 PM

I remember coming across that essay! I suspected the AI use early on and then skimmed the rest to see if there was a disclosure — and when there was, I didn't bother to read it ... I mention this because I think it's the point you were hoping to make with the gesture, in a sort of meta way.

@vauhinivara.bsky.socialOct 8, 2026, 3:36 PM

This is a fascinating study in the context of other research from @informor.bsky.social and others finding that AI models generate text that's more persuasive than human text. One potential reason: Readers suspected AI, so they trusted the content less.

@vauhinivara.bsky.socialOct 7, 2026, 6:43 PM

I inadvertently kinda ripped it off for the ending of one of my stories and when a journal editor pointed it out I let her take it out but then I put it back in for my book 🤪

@vauhinivara.bsky.socialOct 7, 2026, 6:35 PM

Best last line too

@vauhinivara.bsky.socialOct 7, 2026, 6:26 PM

Hi @tedunderwood.com, can you share the preprint here? I’m interested & maybe others will be too! Thanks for your analysis.

@vauhinivara.bsky.socialOct 6, 2026, 4:58 PM

Here's the paper that accompanies the demo. The researchers find that IdeaLens can fairly accurately flag (when Pangram can't) stories that are written by people based on "AI-generated plans."

@vauhinivara.bsky.socialOct 6, 2026, 4:58 PM

I checked an essay my dad wrote and had ChatGPT significantly edit. Pangram calls it 100% AI or AI-assisted. IdeaLens calls the prose 98% likely to have been AI-generated but gives the ideas just a 19% chance of having been AI-generated.

@vauhinivara.bsky.socialOct 6, 2026, 4:58 PM

The award-winning Commonwealth Prize story in Granta that Pangram flags as 100% AI comes out in IdeaLens as having a more than 99% chance of being AI for both prose and ideas. granta.com/the-serpent-...

@vauhinivara.bsky.socialOct 6, 2026, 4:58 PM

The critic @roncharles.bsky.social put AI-generated material based on an interview transcript in his Substack (deliberately, to make a point). The AI detector Pangram flagged it as 77% AI. IdeaLens says the prose has a 60% chance of being AI but the ideas 14%. roncharles.substack.com/p/will-human...

@vauhinivara.bsky.socialOct 6, 2026, 4:58 PM

This is a really fascinating research demo, from the lab of Mohit Iyyer at the University of Maryland: It predicts the likelihood that the ideas contained in a piece of text are AI-generated versus human-generated. A couple of examples in this thread.

@vauhinivara.bsky.socialOct 5, 2026, 6:22 PM

ok looks like we need to hear from octavia butler again. tl;dr: actually no one sees the future

@vauhinivara.bsky.socialOct 1, 2026, 3:05 PM

This isn't just a think piece saying they disagree with PEN America; they actually went back and examined at the data and reporting — including a key claim that a literary agent was selling far fewer books by Jewish and Israeli authors in the US — and found it lacking. Fascinating work.

DDan Sinykin@dan-sinnamon.bsky.socialOct 1, 2026, 1:49 PM

Working on this reported essay, just published, increasingly took over my life for the last two months. PEN America did a terrible disservice with its sloppy reporting about American Jews and Israelis in publishing in July. Josh Lambert and I tried to do better. jewishcurrents.org/what-pen-ame...

jewishcurrents.orgWhat PEN America Got Wrong About Antisemitism in PublishingMisleading anecdotes, flawed data, and oversimplified analyses are at the heart of a narrative of Jewish exclusion promoted by the free expression…
1
@vauhinivara.bsky.socialSep 30, 2026, 6:25 PM

What about Look2Justice? It’s a Washington-based org run by incarcerated writers and organizers.

@vauhinivara.bsky.socialSep 29, 2026, 9:40 PM

Thanks for sharing all this. I haven't looked closely at the company since last year, and, yeah, my recent quick check was for 2Hr Learning rather than Legacy of Education (which had been doing business as Alpha) itself. I updated my recent post to reflect that — I should have been more precise.

Older posts
Terms of UsePrivacy PolicyCommunity StandardsHelpGet the app

Grilled Cheese is a product of Village Compute

Version devBuilt at: 2026-10-11 23:30:56 EDT