A Google Home responds to a father and son’s mention of death. Could this have been intentional?
My husband and I have had quite an experience adapting to the Google Home in our home. We love its presence, truly, but there are continually times when we’re surprised at some of the interactions we’ve had with it.
For example, there have been several instances where we’d be chatting loudly and casually between the two of us, about something or other, when the Google Home would suddenly respond, despite the fact that neither of us had uttered the ‘OK Google’ hot key. It’s quite hilarious when it happens, though — we always laugh — and it sort of validates this running joke we share that our Google Home is actually just a roommate who lives here for free in exchange for all the knowledge it brings.
I didn’t think much of these random interactions until this week’s All About Android, where I’m a co-host. We had a video mail submission from a father who was playing a game with his son in a room where the Google Home was stationed. They had reached a part of the game narrative where they were too stuck to go on, and realized that the best way to complete the story in its entirety was to effectively kill their characters off and start anew. But then…
As we were moving on with the game, somebody spoke up and said, “I just want you to know, you’re not alone. Here’s the number for the suicide prevention hotline.”
It was our Google Home, and I think that when my son had said that we should kill ourselves, Google heard that, and that made it pop up with that message.
We already know that Google Home — and, by virtue, Amazon Echo — is constantly listening in, and that this actually a bit of an issue with owning one. And we already know that what you say could be used in a court of law. But it’s curious why it would respond in an instance where there was no specific query directed at it. Does that mean that Google Home is always listening for a chance to interject? Or does it listen for intonation in the voice; whether it’s conversational, and thus it must be conversing with it? Or, is it because even uttering the word Google will make its digital ears perk up in preparation for the next query?
Is Google Home listening, and could it act as an intermediary when someone is trying to harm herself?
For the most part, I thought it was interesting that Google Home responded after it somehow computed that someone was talking about suicide. I doubt it cared who was talking about it; it sounds like it was pre-programmed to listen for that specific phrase and reply in the event that it’s sitting idly in the room of someone who is truly considering the act. I wasn’t able to recreate this scenario at home, however, but as I was writing this article, Google Home responded. What’s triggering it?
This particular event also has me mulling over whether this is a glimpse at our robot future. Are artificially intelligent devices being programmed to be our friends? Or is this what Google merely considers an extra feature — an aid that maybe a programmer thought would be worth including considering the rates of suicide around the world?
What do you think? Has your Google Home ever responded in this manner?