CHATTING ABOUT CHATGPT AND OTHER A.I. PART 2

YESTERDAY, The New York Times offered several viewpoints about ChatGPT, Generative Pre-trained Transformer capabilities (and its potential threats). Today in Part 2, AAAS Science offers insights. And a previously unknown work of one of Spain’s greatest writers is discovered—thanks to A.I.

Fun, But…. In his January 27, 2023, Editorial in Science, Editor-in-Chief H. Holden Thorp writes, “ChatGTP is Fun, But Not an Author.”

Thorp relates an experience with ChatGTP: “I asked it to rewrite the first scene of the classic American play Death of a Salesman, but to feature Princess Elsa from the animated movie Frozen as the main character instead of Willy Loman.”

Thorp says, “The output was an amusing conversation in which Elsa—who has come home from a tough day of selling—is told by her son Happy, ‘Come on, Mom. You’re Elsa from Frozen. You have ice powers and you’re a queen. You’re unstoppable.’ ”

“Mash-ups like this are certainly fun,” Thorp says, “but there are serious implications for generative AI programs like ChatGPT in science and academia.”

The Challenge. Thorp reports, “In a recent study, abstracts created by ChatGPT were submitted to academic reviewers, who only caught 63% of these fakes. That’s a lot of AI-generated text that could find its way into the literature soon.”

“For years,” Thorp says, “authors at the Science family of journals have signed a license certifying that ‘the Work is an original.’ (italics added)…. Further, our authors certify that they themselves are accountable for the research in the paper.”

Drawing an A.I. Line: “Still,” Thorp says, “to make matters explicit, we are now updating our license and Editorial Policies to specify that text generated by ChatGPT (or any other AI tools) cannot be used in the work, nor can figures, images, or graphics be the products of such tools. And an AI program cannot be an author. A violation of these policies will constitute scientific misconduct no different from altered images or plagiarism of existing works.” 

Note, this does not preclude A.I. being used as a research tool: “Of course,” Thorp says, “there are many legitimate data sets (not the text of a paper) that are intentionally generated by AI in research papers, and these are not covered by this change.”

Felix Lope de Vega and A.I. A perfect example of the latter is described in “AI Reveals Unknown Play by One of Spain’s Greatest Writers in Library Archives,” CNN Style, February 1, 2023: “Researchers from Vienna and Valladolid universities used AI to transcribe 1300 anonymous manuscripts and books at the library, saving years of human effort, and also tried to discover their authorship by checking each work against a selection of words used by different writers.”

 CNN Style writes, “The National Library said on Tuesday that experts later confirmed that the Baroque playwright—one of the most prominent names of the Spanish Golden Age — wrote ‘La Francesa Laura’ (The Frenchwoman Laura) a few years before his death in 1635.”

A contemporary of Shakespeare, Lope de Vega was a key figure in the Spanish Golden Age of Baroque literature, considered second only to Miguel de Cervantes (another Shakespeare contemporary). 

What sets Lope de Vega apart is the sheer volume of his works: According to Wikipedia, “Some 3000 sonnets, three novels, four novellas, nine epic poems, and about 500 plays are attributed to him.” By contrast, the kid from Avon produced perhaps 38 plays and some 150 bits of poetry.

Add to Lope de Vega’s output La Francesa Laura, among the 1300 hitherto anonymous manuscripts in Spain’s National Library. 

“The plot,” CNN Style writes, “focuses around Laura, the daughter of the Duke of Brittany and wife to Count Arnaldo. The heir to the French throne is captivated by her, and although she rejects him, her jealous husband tries to poison her. Ultimately, Laura’s righteousness is proven and happiness is restored.”

CNN Style recounts, The National Library said the words used in the text were “closely aligned with Lope’s, and not with those of the other 350 playwrights who were part of the experiment. Experts then used traditional philological research resources to corroborate the findings.”

A.I.’s Role. Loosely, in examining the 1300 works, the A.I. algorithm asked itself, “Who’s written stuff like ‘To be or not to be’ and ‘Alas, poor Yorick’? Bingo!” ds

© Dennis Simanaitis, SimanaitisSays.com, 2023