[wheelinsupial]

There is a paragraph on this in the publicly accessible pre-print [1] of the article.

"Approved by the respective ethics boards of MIT and Maynooth University, the experiment amounted to a friendly stress-test of the potential vulnerability of judicial legal reasoning to the limitations of reliance on Wikipedia, notably, its ad hoc topic coverage and unknown author/editorship. The experiment featured Wikipedia entries authored by faculty and by law students under faculty supervision, who each had access, through their university library, to all the relevant primary and secondary legal materials available to judges and their clerks. This assurance of accuracy and of informed analysis in the content of the entries—though short of that offered by a specialist textbook—indicates that judges or lawyers would be unlikely to be misled by what they might read. However, as the authorship of Wikipedia articles is opaque, this fact would not be known to any legal professional when using them. From the users’ perspective, there was no particular reason to imagine that the creators of the relevant entry had any legal expertise—or even that they lacked an ulterior agenda."

I'm curious where I missed the part about the algorithm did the editing? I don't see any matches for "algo" when I control + f the article. "Algo" doesn't appear in the article either. The excerpted paragraph says who wrote the entries and it wasn't an algorithm.

[1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4174200

[mch82]

Thanks. You summarized the paper better than the article this discussion thread is about. On closer reading, CSAIL had people write the articles and used natural language processing to detect “linguistic fingerprints of the Wikipedia articles that they’d created”.