Wikipedia is abandoning its plan to test annotations to articles with artificial intelligence. Earlier this month, the platform announced plans to test this feature for about 10 percent of visitors from mobile devices. To say that editors didn’t take the plans well would be an understatement. The Wikimedia Foundation (WMF) changed its plans and canceled the test.
Artificial intelligence annotations would have appeared at the top of articles for 10 percent of mobile users. To see them, readers would have to register. The AI-generated summaries only appeared “on a set of articles” during the two-week trial period.
Editorial comments on WMF’s announcement (via 404 Media) ranged from “Ew” to “Smiling in horror.” One editor wrote: “Just because Google has released their AI-powered resumes doesn’t mean we have to beat them. I sincerely ask you not to test it on mobile devices or anywhere else. This will cause immediate and irreversible damage to our readers and our reputation as a trusted and serious source.”
“Wikipedia has in some ways become synonymous with sober boredom, and that’s great,” the editor continues. “Let’s not insult the intelligence of our readers and join the hype of releasing flashy summaries about AI.”
The editors were not happy with the idea. They also criticized the nonprofit for excluding them from the planning stage. “You also claim that this was ‘discussed,’ which is simply ridiculous, as the ‘discussion’ you refer to had only one participant, the author of the original poster, who is another WMF employee,” the editor wrote.
The organization does not rule out the use of artificial intelligence in the future. But they said that editors will not be left in the dark next time. “The introduction of generative artificial intelligence into the Wikipedia reading process is a serious decision with important consequences, and we intend to treat it as such,” a company spokesperson told 404 Media. “We do not plan to implement the summary function in Wikipedia without the participation of an editor.”
