Given how synthetic intelligence’s pitfalls dominate the day by day information cycle, you’d suppose that the media business would take a cautionary strategy to AI. Yeah, suppose once more. In an effort to spice up readership, the Washington Post plans to broaden its opinion part to incorporate beginner writers who might be guided by synthetic intelligence.
Since April, the Post has labored on scaling an initiative to deliver new voices to the paper. Multiple individuals confirmed with the New York Times that this system, dubbed Ripple, includes publishing established Substack writers and nonprofessionals. On the floor, it sounds much like earlier contributor fashions utilized (not essentially efficiently) by the Huffington Post and Forbes.
The greatest distinction is that the Post gained’t information new writers with a human. That’s means too primary. Instead, the outlet will use Ember, an AI coach, to supervise the method. Per the Times, its present prototypes embrace a “story energy” tracker and a sidebar laying out primary parts like “early thesis”, “supporting factors,” and “memorable ending”. It additionally contains a stay assistant to ship writing prompts and assist form content material.
Each article, which might be revealed exterior of the Post’s typical paywall, will nonetheless be reviewed by a human earlier than it goes stay. Per the Times, Ember will start testing this fall. But why check it in any respect? We’ve seen this very same state of affairs play out earlier than. Just two years in the past, CNET got here beneath hearth after publishing AI-guided articles with barely any disclosure. Although editor-in-chief Connie Guglielmo later acknowledged that every story was “reviewed, fact-checked and edited by an editor with topical experience”, Futurism broke down how one of many articles highlighted by Guglielmo was riddled with errors.
The Post has struggled to efficiently adapt to the ever-changing media panorama. Last 12 months, it launched a brand new subscription mannequin in an try to spice up income after shedding $77 million. But a elaborate subscription mannequin can’t prevent if you happen to’re led by dickheads. In January, the Post laid off about 4 % of its workforce in an effort to recoup losses. Just a month later, proprietor Jeff Bezos’ controversial overhaul of the opinion part (which prompted opinion editor David Shipley to instantly resign) price the outlet over 75,000 digital subscribers.
Artificial intelligence may doubtlessly be helpful for journalism. Currently, the Times makes use of it to shift by massive databases and advocate articles. Except Ember is being given a way more important position. And provided that journalists at a number of shops, together with the Atlantic and Politico, are pushing again towards AI, the Post’s whole plan appears particularly ridiculous. But hey, what the hell do we all know?
