Wikipedia Contributors Clash Over AI-Generated Summaries

Where does “AI” fit within Wikipedia? The specialized site 404 Media reported on a heated discussion that took place in early June among some English-speaking contributors to the online encyclopedia regarding the announcement of a test related to generative artificial intelligence (AI). The experiment was originally planned for two weeks: up to 10% of readers could have seen, when opening a Wikipedia page, a summary of a few lines automatically generated by AI.

This sparked a minor revolt. Some contributors expressed their refusal to see such texts appear as such on the encyclopedic portal. They point to the risk of “disinformation,” as AI generating texts can be guilty of “hallucinations.” The danger: discrediting the information contained in the original article but also, more broadly, the usual work of editing, verification, and adding links carried out by all Wikipedia contributors. Some requested that any AI-generated text be supervised by a “human” editor before publication, and be labeled as such. The experiment was put on hold the day after these reactions, 404 Media reports.

---

Enjoyed this post by Thibault Helle? Subscribe for more insights and updates straight from the source.