The Post‘s Opinion section has already been fucked with—owner Jeff Bezos infamously laid out some strict editorial guidelines about “personal liberties and free markets.” Now, the revered paper will apparently expand the section to include writing that has already been published elsewhere—at other outlets or even from somebody’s Substack. There’s also a “final phase” to this expansion project where “nonprofessionals” could submit columns written with the help of an AI coach called Ember.
According to The Times, Ember basically sounds like a more advanced version of Grammarly or a similar tool. It doesn’t do the actual writing, but it “could automate several functions normally provided by human editors,” sources shared. Per the report, “Early mock-ups of the tool feature a ‘story strength’ tracker that tells writers how their piece is shaping up, with a sidebar that lays out basic parts of story structure: ‘early thesis,’ ‘supporting points’ and ‘memorable ending.’ A live A.I. assistant would provide developmental questions, with writing prompts inviting authors to add ‘solid supporting points,’ one of the people said.”
Submissions created with Ember will supposedly be reviewed by human editors, which begs the question, why does AI have to be involved at all? Back in this writer’s day, even “nonprofessionals” learned the above skills in high school English comp classes. It’s unclear what makes this fundamentally different from just using The Washington Post op-ed submission form—unless the AI is doing more of the writing than this initial report suggests. Even then, we have plenty of examples of AI writing being grossly inaccurate, so to publish something too AI based would leave The Post vulnerable to embarrassment, if not legal liability, which circles back to Ember not being a particularly useful tool at all. Perhaps all the money invested in AI tech would be better used elsewhere.
