TaxProf Weblog

Benjamin Alarie (Toronto; Google Scholar) & Arthur J. Cockfield (Queen’s; Google Scholar), Will Machines Exchange Us? Machine-Authored Texts and the Way forward for Scholarship, 3 L., Tech & People 5 (2021):

We current right here the primary machine-generated legislation assessment article. Our self-interest motivates us to consider that data staff who write complicated articles drawing upon years of analysis and energy are protected from AI developments. Nonetheless, how cheap is it to persist on this perception given current advances in AI analysis? With that matter in thoughts, we prompted GPT-3, a state-of-the-art AI, to generate a paper that explains “why people will all the time be higher attorneys, drivers, CEOs, presidents, and legislation professors than synthetic intelligence and robots can ever hope to be.” The ensuing paper, with no edits other than giving it a title and bolding the headings generated by GPT-3, is reproduced beneath. It’s imperfect in a humorous manner.

Paradoxically, it’s publishable “as-is” solely as a result of it’s machine-generated. Nonetheless, the ensuing paper is sweet sufficient to present us some pause for thought. Though GPT-3 is lower than the duty of changing legislation assessment authors at present, we’re far much less assured that GPT-5 or GPT- 100 may not be as much as the duty sooner or later.

https://taxprof.typepad.com/taxprof_blog/2022/02/will-machines-replace-us-machine-authored-texts-and-the-future-of-scholarship.html