Jeff Bezos’s Washinton Post Plans to Add Random Opinion Writers Edited by AI

Given how artificial intelligence’s pitfalls dominate the daily news cycle, you’d think that the media industry would take a cautionary approach to AI. Yeah, think again. In an effort to boost readership, the Washington Post plans to expand its opinion section to include newbie writers who will be guided by artificial intelligence.

Since April, the Post has worked on scaling an initiative to bring new voices to the paper. Multiple people confirmed with the New York Times that the program, dubbed Ripple, involves publishing established Substack writers and nonprofessionals. On the surface, it sounds similar to previous contributor models utilized (not necessarily successfully) by the Huffington Post and Forbes.

The biggest difference is that the Post won’t guide new writers with a human. That’s way too basic. Instead, the outlet will use Ember, an AI coach, to oversee the process. Per the Times, its current prototypes include a “story strength” tracker and a sidebar laying out basic components like “early thesis”, “supporting points,” and “memorable ending”. It also features a live assistant to deliver writing prompts and help shape content.

Each article, which will be published outside of the Post’s usual paywall, will still be reviewed by a human before it goes live. Per the Times, Ember will begin testing this fall. But why test it at all? We’ve seen this exact same scenario play out before. Just two years ago, CNET came under fire after publishing AI-guided articles with barely any disclosure. Although editor-in-chief Connie Guglielmo later stated that each story was “reviewed, fact-checked and edited by an editor with topical expertise”, Futurism broke down how one of the articles highlighted by Guglielmo was riddled with errors.

The Post has struggled to successfully adapt to the ever-changing media landscape. Last year, it launched a new subscription model in an attempt to boost revenue after losing $77 million. But a fancy subscription model can’t save you if you’re led by dickheads. In January, the Post laid off about 4 percent of its workforce in an effort to recoup losses. Just a month later, owner Jeff Bezos’ controversial overhaul of the opinion section (which prompted opinion editor David Shipley to immediately resign) cost the outlet over 75,000 digital subscribers.

Artificial intelligence could potentially be useful for journalism. Currently, the Times uses it to shift through large databases and recommend articles. Except Ember is being given a much more significant role. And given that journalists at multiple outlets, including the Atlantic and Politico, are pushing back against AI, the Post’s entire plan seems especially ridiculous. But hey, what the hell do we know?