UC Press Blog

by HSNS contributors Melinda Baldwin and Brigid Vance

In the third chapter of his now-classic book Lords of the Fly: Drosophila Genetics and the Experimental Life, Robert Kohler introduces his reader to Jack Schultz, a “bright, precocious, and independent” researcher whose gregarious, quick-witted personality made a noticeable impact on the culture of Drosophila genetics. However, Kohler tells us, Schultz’s reputation as a researcher was hampered by an inconvenient fact: “He hated writing and would do almost anything to avoid it.”[1]

We can’t help wondering what Jack Schultz would have thought if he had been alive in November 2022 when OpenAI launched the first version of its generative AI chatbot, ChatGPT. Would Schultz have begun feeding ChatGPT prompts to help him with his most hated scientific task, or would he have decried its prose as nonsensical garbage that he didn’t have the time to edit? Would the “moral economy” Kohler famously described for Drosophila geneticists have accepted the use of AI to draft scientific papers? Would such papers have been acceptable to scientific journals—and if they were, would Schultz have been obligated to list “C. GPT” as a coauthor? 

In short, how would Schultz and those around him have answered the questions that writers and editors of scholarly papers are confronting today?

This month, HSNS is running a special section on “AI and Scholarly Publishing,” with six essays reflecting on AI and scholarship from the perspective of the history and philosophy of science and technology. It seems to us that our field has something unique to offer the conversation about AI; as historians of science, we are poised not just to consider the potential impact of AI on our own scholarly writing and the training of students, but to see AI in the context of a longer historical conversation about the nature of knowledge, the rules of authorship, and human interactions with technology.

Joseph Martin, chair of the HSNS editorial board, explores the idea of “intelligence.” Alex Csiszar dives into the “authorship” question, considering large language models in a longer history of how we assign credit for scientific writing. And Damien Williams’s essay on the “objectivity” of GPTs challenges us to rethink assumptions that we may unconsciously hold not just as scholars, but also as administrators and teachers. Nicole Howard suggests that understanding the ambivalence surrounding authorship and trust in the early modern printing press may help us untangle the relationship between AI and authorial trust. Samuel Moore exhorts us to collectivize knowledge production to slow the flood of unaccountable AI technologies. Based on her decades’-long work in Japan, Jennifer Robertson recommends that we approach AI with both a historical perspective and ethnographical fieldwork. 

We think these six essays raise timely, pertinent questions about the role of AI in scholarship. They place our current debates in a broader context, consider a long history of concepts about authorship and knowledge production, and suggest ways we might rethink the here and now. And it is imperative we rethink in the here and now, too, so that we not only understand the technologies that we have created, but also the potential for impact, both positive and negative. As OpenAI CEO Sam Altman confessed to journalist Kara Swisher, “we are messing around with something we don’t truly understand.”[2] We think these essays indicate that one way to better understanding is to see generative AI as part of a long history of debates about technology, authorship, objectivity, and the nature of knowledge.

[1] Robert Kohler, Lords of the Fly: Drosophila Genetics and the Experimental Life (Chicago: U. Chicago Press, 1994), 127-8.

[2] Quoted in Kara Swisher, Burn Book: A Tech Love Story (New York: Simon and Schuster, 2024), 291.

We invite you to read the special section for free online for a limited time.

POD copies of HSNS's special issue 54.5 can be purchased on the journal’s site. For ongoing access to HSNS, please ask your librarian to subscribe and/or purchase an individual subscription.