But the social media allegations have turned into a scandal by now, and if the New York Times is writing about it now, so can I.
And if you have reached a blog post about a short story, you can also read the short story and form your own opinion. It is called The Serpent in the Grove and is credited to author Zamir Nazir. It is not paywalled, and is available on the Granta website.
How did you feel when you read the sentence, “Outside, little Putty – three years old, dark as the sun, with shining eyes – was chasing a bird through the dust, her laughter like water on pebbles”? I’m guessing you made the joke with several AI tropes in mind. Even if you came into the story disinterested, you might have felt them, but even then you might not have felt that way. Let’s be honest: If there were no scandal, you probably wouldn’t be reading a short story today.
But here’s a section that seems less likely to be written by an AI:
Putti walks there carrying his father on his shoulders and his mother on his shoulders, when work bothers him. He stops before the ring out of respect due to his habit. He hears: the twinkling language of the leaves, the faint whisper of the sun, a creaking where the wood learns to pretend to be a board and gets tired of pretending.
It’s too stylized and fiddly with grammar to be a normal AI output. But what does this really mean for the person who thinks the AI model wrote the story? Does this refute the entire notion of AI authorship? Does this mean that the human author embellished some parts? or do you Perhaps Do you think you could get an AI model to write like this, especially if you gave it an example?
Granta’s publisher, Sigrid Rausing, issued a puzzled, vague statement about the AI allegations, writing, “It may be that the judges have now awarded the prize for an example of AI plagiarism – we don’t know yet, and perhaps we will never know.” But their statement also said they put the story in the cloud, and it focuses on the more human-seeming parts, adding that they have “unusual specificity” and that AI could potentially be used to “elaborate” on those parts.
But then, wow, Who really cares what Claude thinks about this?
Razmi Farooq, director general of the foundation that administers the Commonwealth Prize, spoke to The New York Times, and kept things fairly vague, saying that his organization “has taken stock of the comments,” and has done some internal introspection to see whether we feel our process so far has been robust enough. While their foundation is “confident in the rigor” of its AI-checking process, they note that it is an “evolving technical environment.”
There are now other stories on the Granta website accusing AI online plagiarism, and Granta has added a note to all Commonwealth Award winners, saying in part, “The suggestion that writers have not presented their content authentically is an allegation we take seriously, but until definitive evidence emerges, we will keep these stories on our website.
But despite some early allegations to the contrary, Zamir Nazir appears to be a real person living in Trinidad and Tobago. If they used AI to write the story, the “prestigious” prize was worth more prestige than money. He received £2,500 for his trouble, as he was the Caribbean regional award winner. The all-around winner, who receives £5,000, will not be announced until June 30.
Judging by the intensity of discussions online, especially on book subreddits, it appears that eventually, someone is going to track down Nazir and ask him to either confess or make him write a signed legal affidavit swearing in blood that he wrote it all himself.
At any rate, it’s doubtful that those upset by this will get the retribution they want. Even if Nazir is guilty, he may deny it, or – in keeping with what people do in this situation – claim that he has taken suggestions here and there from LLM, but he is still the true author.
Meanwhile people certainly have a lot of strong opinions about the short story. And if the truth is that Nazir writes like LLM, then what is the way to find out.
<a href