
"The most common email messages I receive these days are obviously AI-generated pitches for guests to appear on my podcast. They all begin the same way, with a praising reference to one of my recent episodes-usually the second-to-last posted show. "Your recent interview with so-and-so was penetrating, and got to the heart of the problem of x or y." Then comes the crucial pivot: "John Dough's work takes that problem even further . . ." And then the pitch for John Dough to be on the podcast."
"The problem is not just that the publicist used AI to shotgun the known universe of podcasters with pitches artificially customized to their shows. It's that the comparisons and connections are really bad. "Your guest spoke so passionately about being a death doula, I think you would be so interested in an artist who makes Halloween napkins festooned with skeletons, which are usually of dead people.""
"I see the same thing happen with AI-generated reports and presentations. Someone gets some speculative idea and then asks Chat to justify it with a few case studies. On the surface, the case studies may sound like they're supporting the premise-but if you look any deeper, they don't really relate at all. They're analogous, but not truly relevant. Worse yet, they're sitting in what looks like a fully realized Powerpoint presentation."
AI-generated outreach frequently uses formulaic praise to create the appearance of personalization while connecting unrelated people or ideas. Such mass-produced pitches produce absurd or weak analogies that signal a loss of human judgment and result in publicists losing credibility. Treating AI purely as a productivity tool mirrors an Industrial Age approach that increases output while neglecting the nuanced human processes that constitute expertise. Similar problems appear in AI-generated reports and presentations, where superficially supportive case studies may be analogous but not truly relevant and are presented inside polished slides, creating misleading authority and reduced clarity.
Read at Fast Company
Unable to calculate read time
Collection
[
|
...
]