Good morning.
I wrote last month that the development of ChatGPT may turn out to be the most significant news event of 2022. If I’m right, then kudos to marketing mavens Richard Bowman and David Boyle for producing the first book on the topic: Prompt: A practical guide to AI-powered brand growth with ChatGPT. I read it over a long airplane ride this weekend, and it crystallized my thinking about what makes this technology revolutionary: It produces informed bullshit.
I mean that in the best way. Like many, I had tended to view A.I. as a tool for calculating reliable, data-based predictions, point solutions, or probabilities. But of course, that’s not what the human mind does. There’s ample research showing human beings are exceptionally bad at data-driven predictions and probabilistic thinking. What we are good at is synthesizing information to quickly detect interesting patterns and insights that may—or may not—survive more rigorous scrutiny. In short, we are good bullshitters. Which makes ChatGPT the most human form of A.I. yet. As Bowman and Boyle put it:
