As 2018 gets started, there are many dire warnings about the dangers facing the world, from a war with North Korea to an explosion of violence in the Middle East. But I have a more mundane fear: that the robots and algorithms that we are incorporating into our daily lives will suddenly go crazy.
You may think that I’ve been watching too many science-fiction movies over the holidays, but it wasn’t that.
It was a personal experience I had with Alexa, the Amazon Echo virtual assistant that my son gave me as a birthday present more than a year ago, and that has been in my living room ever since.
Alexa is a virtual assistant like Apple’s Siri that lives inside a speaker that looks like a cylinder and has a round light on its top. The device turns itself on when it hears the word “Alexa,” and you can ask it any question you want.
You can also ask Alexa to play a specific tune, give you the weather forecast or provide you with the latest news. Alexa can also order a pizza for you or buy a book from Amazon, its parent company’s site. Alexa is already in more than 20 million American homes, according to the company.
At home, we have been using it for its most basic functions, mostly to get the weather forecast, play music and get news briefs from National Public Radio. It has mainly been a good conversation piece, especially over dinner when we have visitors from other parts of the world where Alexa is not yet available.
But, while my overall experience with Alexa has been fun, I had a freaky incident with her a few months ago that makes me wonder about what will happen as we allow much of our daily lives to be aided – if not run – by virtual assistants, GPS navigation systems, medical diagnostic robots and other intelligent machines.
It happened while I was working in my home office. I suddenly heard a loud male voice in the living room. At first, it didn’t cross my mind that it could be Alexa. I was alone, my wife was traveling abroad and we have no pets. There was nobody at home who could have awoken Alexa by calling her name.
Scared, and frustrated by not finding any hard object that I could use as a weapon against a possible thief, I slowly walked into the living room, with my heart pounding. Once there, I found that there were no strangers – only Alexa, which had somehow turned herself on and was broadcasting the latest news from NPR.
I asked Amazon’s customer support department why this happened and was told by a friendly representative that it was an unusual incident that may have been caused by a “technical glitch.”
In a previous query to Amazon’s media department, I was told that the NPR program may have been on pause, and Alexa may have “thought” she heard the words “play” or “resume.”
I also learned that another possibility was that NPR was playing on another device at home with a story about Alexa, and she may have heard the wake word followed by “NPR.”
None of those possible explanations satisfied me, because Alexa had been off for many days, if not weeks, and because there was no radio or television set on at home at the time. I’m still corresponding with higher-ups at Alexa’s customer support staff, who say they are looking into what exactly could have happened.
Intrigued by my incident with Alexa, I re-read over the holidays the book “Homo Deus,” by Israeli historian Yuval Noah Harari, about the future of mankind in an increasingly robot-run world. He writes that, “Once Google, Facebook and other algorithms become all-knowing oracles, they may well evolve into agents and finally into sovereigns.”
But before intelligent machines become so smart that they can run the world, we should worry about a much simpler threat: that they simply run amok.
Andres Oppenheimer is a Latin America correspondent for the Miami Herald. He can be contacted at firstname.lastname@example.org.