HarperCollins CEO's ideas for using AI sound just godawful

Murray notes that the company will mostly be using the tech for stuff like sales and marketing and behind-the-scenes work. (Including in editorial, which has a grimness all its own to countenance.) But he has apparently been pitched on some more overtly dystopian ideas, including the idea of a “talking book.” “Where,” according to an interview with Publishers Weekly, “A book sits atop a large language model, allowing readers to converse with an AI facsimile of its author.” At last: No more having to think about the meaning of complicated passages, or trace the lines of thought that got an author from A to B; now you’ll be able to just ask a computer, have it lie to you, and move on with your life. More prosaically, HarperCollins apparently has all the recipes from many of its previously published cookbooks loaded into the computers now, so it could just use AI to shit out a web site to allow people to browse them, which is boring, but at least marginally less evil.

The thing is—and you will notice this is a running theme any time CEOs start telling themselves, and us, that AI is the future—not even people like Murray seem to actually believe there’s money in this kind of computerized codswallop. Calling the AI products “kind of interesting,” Murray made it clear that “I don’t know how you market, price, or sell them.” Some people might take that factoid and posit that the products in question are hard to sell because they have no value, but guys like Murray clearly aren’t going to stop: “We’re spending a lot of time figuring things out.”