So apparently there’s this big restaurant chain that’s sending mystery shoppers around to make sure that staff at restaurants in the chain act super-happy all the time. Matt Yglesias thinks this is a sign of some more profound issues:
There’s a fairly profound issue here, however. If you think about an increasingly mechanized, digitized, and roboticized economy the case for human labor will increasingly be one of affect. And the affect economy is really everywhere. People go to yoga classes instead of watching streaming videos at home. Bartenders work for tips. Journalists are increasingly expected to “engage” with the audience through a variety of social media and put forth not just informative copy but an appealing persona. A real estate agent improves upon a computer search algorithm in part by offering a modulated emotional response that makes the client feel happier and more confident about the process.
In all kinds of fields, people who intuitively “get” the considerable advantages of appropriate revenue-maximizing self-presentation will be advantaged. Normally you think, “okay there’s a valuable job skill that not everybody has so at some firms managers are teaching it to people—that’s good, that’s training.” But enforced-friendliness policies and mandatory personae seem like a violation of people’s integrity in a way that “operate the lathe this way, not that way” don’t. Worries about an economy in which robots displace human labor and everyone sits around unemployed watching The Real Housewives of Boise strike me as overblown. What you should worry about, instead, is the growing significance of the affect economy and the problems people will have with what Pret is trying to do here.
One thing to point out here is that there’s little reason to think we won’t one day be able to program robots to act super friendly and happy too. I suppose the rebuttal to that from Yglesias’ point of view is that people might prefer to have certain kinds of work done by “real people”* even if robo-personalities are indistinguishable from the real thing. It would be similar to how people are now willing to pay a premium for a handmade necklace bought off Etsy, even if you could set up a factory to make indistinguishable necklaces.
I still think Yglesias is underestimating the revolutionary potential of human-level AI, but it’s also interesting to think about the challenges we could face if things go the way he’s predicting.
*Apologies for the scare quotes, but I’m not actually sure what “real people” would mean in a posthuman future.