Whether or not your home contains one of the so-called “smart speakers,” I encourage you to spend 15 minutes with the cover story of The Atlantic’s November 2018 issue. Judith Shulevitz’s “Alexa, Should We Trust you?” serves as a captivating look at the present state of voice-activated virtual assistants and an evenly-weighted consideration of the future of devices like the Amazon Echo, Google Home, and Apple HomePod. (Full disclosure: I do not own any of these technologies, but I have seen several in operation in friends’ households.)
Before I share a few more comments about the article, it is useful to identify the distinctions between the pieces of hardware that may be sitting on your kitchen counter or bedroom dresser and the web-connected digital assistants that these gadgets allow users to interact with. The most popular line of smart speakers is the Amazon Echo, which — as of 2018 — 31 million U.S. users have plugged into a socket. The Echo hosts the upbeat female voice of Alexa. The bulk of Shulevitz’s article examines the relationship between human users and this admittedly robotic yet eerily articulate domestic oracle.
Less popular than the Echo/Alexa pairing — at least as measured by units sold — is the combination offered by Google, the search engine titan that powers 90% of web inquiries. Google’s devices, which number roughly 14 million active units in the United States, are identified as the Home series. The web-connected medium that they channel is the Google Assistant. Lastly, one of the most recent entries into the smart speaker market is the Apple HomePod. This cute, bulbous device lets users interact with Siri, the virtual assistant that many iPhone owners are familiar with.
With definitions out of the way, let’s move on to a few teasers in the hopes that you make time to read this worthy piece. The article’s introduction features some noteworthy statistics, namely that 8 million Americans own three or more smart speakers (i.e. that multiple devices are found in their homes) and that by the year 2021 estimates suggest that there will be nearly as many smart speakers in operation as the number of people on earth. That’s a stunning amount of domestic tech with ears eager to hear what we are up to. As Shulevitz notes, “They [the device manufacturers] want to colonize space. Not interplanetary space. Everyday space: home, car, office.”
And the fact that digital assistants like Alexa are constantly waiting for our input is one of the reasons that trust is a central theme that runs through the psychological aspects of human-computer interaction that Shulevitz explores. Users are, after all, not conversing with an entity located next to the kitchen sink; rather, they are speaking to a voice-recognition system whose computing “brain” is housed in gigantic server farms humming inside warehouses far from our homes. In reality, devices like the Amazon Echo serve as portals. Our words pass through them even though it feels like we are speaking with them.
When we employ the “wake word” that cues Alexa or Siri to pay attention, it prompts a system that inhabits a sprawling climate-controlled facility. Our queries about everything from the weather in Miami to the number of Ted Bundy’s victims are then tallied, analyzed, and categorized by the algorithm-infused AI infrastructure. These vast digital networks dispassionately learn everything they can about our likes, dislikes, and curiosities with the ultimate goals of (1) being able to predict what we might next desire, and (2) establishing metrics about how we — and others like us, based on demographics — think and behave. We are entrusting our personal data to Big Tech, and those firms are hungry to harvest it.
Although sharing information online is certainly not a new concept, especially given the billions of people worldwide who use Facebook and other social media platforms, the trust dynamic shifts when the devices we are interacting with can talk intelligently with us. When prompted by a seemingly humane entity, humans can be persuaded to divulge intimate questions and confessions. Large technology companies possess the power — and, in many cases, the incentive — to sell the fruits that grow from these sensitive topics to bidders who want to market products, services, and lifestyle aspirations to us. Consider this potentially unnerving fact from the article: if a web-connected Roomba sweeps crumbs and dog hair from your linoleum, then its manufacturer (iRobot) owns a partial floor-plan of your residence and knowledge of where your furniture is located. (!)
If the preceding paragraph suggests that The Atlantic’s cover story serves as an ominous deep-dive into the dark intentions of Silicon Valley, that would be inaccurate. Instead, the article is far more about our human nature and what our willingness (and desire) to direct questions to small devices that possess a voice — but no eyes, face, or conscience — means about our curiosities and vulnerabilities. We seek to be informed. And we don’t want to be lonely. Increasingly, it appears that we are willing to ask Alexa for assistance with both. What does it suggest about the human condition if we are more comfortable telling the black soup-can-like pod on the windowsill that we are feeling down than we are picking up the phone and discussing that same sadness with a friend?
Judith Shulevitz writes that, “We’ve been reacting to human vocalizations for millions of years as if they signaled human proximity.” And until now, they have. But today, in houses around the world, there are children, teens, and adults holding conversations with voices whose convenient accessibility and lack of moral judgement signal not a human proximity, but the nearness of a tantalizingly powerful amalgamation of trillions of data points arranged in the silhouette of a human form.
That form provides correct answers — at least, in many cases — but can it yield the right answers? And what does it mean if we believe that it can? It is the latter question that I think needs to be carefully considered. Are we nearing a cultural tipping point where humanity deems it desirable to not only request objective clarification from virtual assistants (e.g. “Alexa, how many acres can fit in a square mile?”) but also preferable to seek understanding or comfort from their AI brains (e.g. “Alexa, why am I struggling to make friends at my new school?”) instead of from fellow human beings?
By assessing one outcome of the quickening pace of technology advancement — and, in doing so, identifying a potentially alarming reality — Shulevitz remarks: “The line between artificial voices and real ones is on its way to disappearing.” If (or when) this distinction is fully realized, what will occur to the perceived value of face-to-face interactions? If discussing difficult subjects with a faceless digital assistant becomes good enough — or even better than talking to one’s closest confidants — will humanity have taken a step forward in sync with technology, or a step backward from genuine intimacy? Please, do not delay another minute. Follow this link to The Atlantic’s full article. You’ll be glad you did. Trust me.
Note — The images in this post, which were created by acclaimed illustrator Roberto Parada, are featured in The Atlantic’s on-line article and its November 2018 print edition.