Let stochastic parrots squawk: why academic journals should allow large language models to coauthor articles

In late 2022, the large language model (LLM) known as ChatGPT was released for public use, and a few researchers began crediting it as a coauthor of publications. The academic reaction was overwhelmingly negative; experts condemned LLM coauthorship, and major journals banned LLM coauthorship. In this paper, I challenge this reaction. Specifically, my main thesis is that journals should allow LLMs to coauthor articles. My two sub-theses are that journals should allow “heavy LLM usage” and that journals should allow LLMs to be credited as coauthors when they are heavily used. However, I argue that journals should do this only in conjunction with adopting policies like the ones I propose. Unlike the existing publications that share my conclusion, this paper is extensive, covering the enormous number of prima facie plausible reasons for banning LLM coauthorship. As I show, the alleged downsides of LLM coauthorship are either exaggerated, similarly applicable to human coauthorship, heavily mitigated by existing journal policies and/or my proposed journal policies, or outweighed by the upsides of LLM coauthorship.