Robot service

Everyone knows that customer-service agents can seem incredibly stiff. But what if they’re actually robots? A writer at the San Franciso Chronicle got a letter from a reader complaining that the online help folks for SBC Yahoo’s internet service seemed overly fakey. So the writer logged on to talk to the help himself — someone who identified himself as Floyd:

I was asked, “How can we assist you today?”

“Floyd,” I wrote, “do you really exist?”

A few minutes passed. Then came the reply: “Yes, I am a person like you.”

Was it just me, or did that sound eerily like something Hal might have said in “2001”?

“So this is really a live session?” I asked. “I’m not talking to a machine?”

“Yes, you are talking to a live person,” Floyd wrote back after another lengthy interval.

“Prove it,” I replied. “What is the current level of the Dow Jones industrial average?”

Duck that one, I thought, as the minutes ticked by.

“Please be assured that this is a live and interactive chat session and I am a human being like you,” Floyd finally answered. “May I know the issue you are facing related to SBC Yahoo Internet services?”

“I just want a little confirmation,” I responded. “How about this: What did you think of the new ‘Lord of the Rings’ movie?”

The guy’s supposed to be a techie, right? No way he hasn’t seen “The Two Towers.”

“Please know that we are not authorised” — note spelling — “to discuss anything other than the technical issues related to SBC Yahoo Internet services,” Floyd wrote back.

I love this stuff. Obviously, as the writer points out, Eliza-like bots have been around for years — faking conversation by taking a person’s statement and reformatting it as a question, a la Rogerian psychology (“I hate my mother.” “Why do you hate your mother?”) That’s part of the stuff I wrote about in my profile of the creator of the A.I. chatbot Alice. And many of you probably know that companies like ActiveBuddy have been making customer-service ‘bots for a while. Indeed, some of the ‘bots are sufficiently good that, according to ActiveBuddy employees I’ve interviewed, customers sometimes just wind up having regular conversations with the ‘bots. Unlike the SBC/Yahoo ‘bot, which seems pretty crappy, some of these customer-service A.I. constructs have tens of thousands of possible conversational gambits. “If you’re bored at work, they’re sometimes more interesting to talk to than, say, your co-workers,” said one ActiveBuddy executive.

The irony here is that actual customer-service agents — the real, live ones — are nowadays increasingly forced to read from scripts and never deviate from them. I’m not sure why; maybe it’s got to do with liability issues. But either way, talking to these people is incredibly awkward. A while ago, I called Sprint to get some help with my mobile-phone account. Every question I asked, I could the guy rummaging through online Q&A guides to find the appropriate pre-canned answer, which he’d numbly recite with an almost Brechtian lack of emotion. When I finally ended the call and said good-bye, he said “At Sprint we try to offer you the finest quality service. Have I offered you the finest quality service today?” Well, no, man — because it’s kind of creepy to talk to someone who sounds like a robot!

Which is precisely the point, of course. In the modern service economy, robots behave like people, and people behave like robots. Would you like fries with that?

(Thanks to Plastic for this one!)

blog comments powered by Disqus

Search This Site


I'm Clive Thompson, the author of Smarter Than You Think: How Technology is Changing Our Minds for the Better (Penguin Press). You can order the book now at Amazon, Barnes and Noble, Powells, Indiebound, or through your local bookstore! I'm also a contributing writer for the New York Times Magazine and a columnist for Wired magazine. Email is here or ping me via the antiquated form of AOL IM (pomeranian99).

More of Me


Recent Comments

Collision Detection: A Blog by Clive Thompson