Yesterday morning, billionaire Los Angeles Times proprietor Patrick Soon-Shiong published a letter to readers letting them cognize the outlet is now utilizing AI to adhd a “Voices” explanation to articles that return “a stance” aliases are “written from a individual perspective.” He said those articles whitethorn besides get a group of AI-generated “Insights,” which look astatine the bottommost arsenic slug points, including immoderate labeled, “Different views connected the topic.”
“Voices is not strictly constricted to Opinion conception content,” writes Soon-Shiong, ”It besides includes news commentary, criticism, reviews, and more. If a portion takes a stance aliases is written from a individual perspective, it whitethorn beryllium branded Voices.“ He besides says, “I judge providing much varied viewpoints supports our journalistic ngo and will thief readers navigate the issues facing this nation.”
The news wasn’t received good by LA Times union members. In a connection reported by The Hollywood Reporter, LA Times Guild vice chair Matt Hamilton said the national supports immoderate initiatives to thief readers abstracted news reporting from sentiment stories, “But we don’t deliberation this attack — AI-generated study unvetted by editorial unit — will do overmuch to heighten spot successful the media.”
It’s only been a day, but the alteration has already generated immoderate questionable results. The Guardian points to a March 1st LA Times opinion piece astir the threat inherent successful unregulated usage of AI to nutrient contented for humanities documentaries. At the bottom, the outlet’s caller AI instrumentality claims that the communicative “generally aligns pinch a Center Left constituent of view” and suggests that “AI democratizes humanities storytelling.”
Insights were besides apparently added to the bottommost of a February 25th LA Times story astir California cities that elected Klu Klux Klan members to their metropolis councils successful the 1920s. One of the now-removed, AI-generated, bullet-pointed views is that section humanities accounts sometimes painted the Klan arsenic “a merchandise of ‘white Protestant culture’ responding to societal changes alternatively than an explicitly hate-driven movement, minimizing its ideological threat.” That is correct, arsenic the writer points retired connected X, but it seems to beryllium clumsily presented arsenic a counterpoint to the story’s premise – that the Klan’s faded bequest successful Anaheim, California has lived connected successful schoolhouse segregation, anti-immigration laws, and section neo-Nazi bands.
Ideally, if AI devices are used, it is pinch immoderate editorial oversight to forestall gaffes for illustration the ones LA Times is experiencing. Sloppy aliases nonexistent oversight seems to beryllium the roadworthy to issues for illustration MSN’s AI news aggregator recommending an Ottawa nutrient bank arsenic a tourer luncheon destination aliases Gizmodo’s awkward non-chronological “chronological” database of Star Wars films. And Apple recently tweaked its Apple Intelligence notification summaries’ quality aft the characteristic contorted a BBC headline to incorrectly propose that UnitedHealthcare CEO shooting fishy Luigi Mangione had changeable himself.
Other outlets usage AI arsenic portion of their news operations, though mostly not to make editorial assessments. The exertion is being utilized for a assortment of purposes by Bloomberg, Gannett-owned outlets like USA Today, The Wall Street Journal, The New York Times, and The Washington Post.
English (US) · 
Indonesian (ID) ·