How Google's Eerie Robot Phone Calls Hint At AI's Future

- Hello, how can I help you? - Hi, I’m calling to book a women’s haircut for a client. I’m looking for something on May 3rd. - Sure, give me one second. - Mmhmm.

The recording of this seemingly mundane phone call was played at one of the most talked about demonstrations at Google’s annual developer conference this week. What stirred the pot was the fact that the call wasn’t made by a human, but by Google Assistant enhanced with Google’s new Duplex technology for handling bookings and making appointments.

Not only did the bot manage to schedule a haircut, but it pulled off a natural-sounding conversation, pausing in the right places and adding “ummm” and “mmhmm” for realism. It did such a good job that a human working at a salon had no clue it was a robot on the other side of the phone.

Soon after the demonstration tech reporters, including journalists at the Guardian, Quartz, the Verge and WIRED sounded the alarm about the ethical consequences of Google fooling a person into thinking the caller was another human on the line. 

Soon after, Google clarified in a statement to the Verge that the AI assistant would identify itself as a robot before carrying on a conversation and played down Duplex as an early-stage development that might never make it onto the market. But is this enough to allay concerns?

I'm recommending WIRED's article because it looks beyond the issue of transparency and considers some other “what if” scenarios Google's human-sounding assistant gives rise to, like what if the bot makes a mistake when carrying out your restaurant reservation or if it's used as a machine for spammy calls. It probably still doesn't exhaust the subject, but it's a good starting point for considering the ethical and societal questions related to Google's new natural-sounding AI.