This article originally appeared in TidBITS on 2017-01-04 at 9:23 a.m.
The permanent URL for this article is:
Include images: Off

CES 2017: Tech Trends to Watch

by Jeff Porten

This year’s CES began with two unwelcome realizations: 1) the USB-C connector on my Nexus 5X phone is considerably grippier to the cable than to its internal connections, and 2) my middle-aged tuchus apparently exerts a more considerable lateral force on both than I realized.

So I find myself at the world’s largest tech show, where everyone is promising “futuristic” and “ubiquitous,” and I’m deprived of a smartphone for 48 hours. (A new one is on its way; there’s something to be said for $250 base prices.) It’s mundane to note how central a smartphone is to modern life, but that doesn’t change how devastating it is to be without one: it’s how I know where I’m going, how I get there, what I do along the way, and (when I’m working) a substantial part of what I do when I arrive. Instead, for at least two days, I’m going to be more reliant on a paper pad and pencil than I have since 1999. (And here I’ll pour one out for my beloved Palm VII and Sony Ericsson P800 [2].)

In any case, welcome back to CES (formerly the Consumer Electronics Show), the annual Las Vegas bacchanal of approximately 165,000 attendees, thousands of exhibitors, and around 20,000 newly announced gadgets, gizmos, and networks that aim to be your next “so important I’d be lost without it” technology. I’ve received press releases on gear ranging from smart spoons and trashcans to infrared iPhone cameras and artificially intelligent telescopes, and I’ll be writing about the slice of newsworthy technology that I’m able to see, because everything here happens in dozens of venues simultaneously spread out over the entire city.

I’ll also try to convey what it’s like to be here. CES is a closed show for industry and press, and while it’s exciting, more of my colleagues on Twitter seem to be expressing joy that they’re not coming than looking forward to working the show. It’s fun but also akin to surfing during a hurricane. CES is the kind of event where, when I took a break, I was thankful for the relative peace and quiet of a casino. This is the 50th anniversary CES, covering 2.6 million square feet of convention space; the Consumer Technology Association, which runs the show, estimates that CES will be covered by 1000 more journalists than attended the Olympics.

I kicked off the show with a presentation by CTA’s chief economist, Shawn DuBravac, who spoke about consumer technology trends in 2017. The first of these is the increasing adoption of voice technologies as a user interface for electronics. DuBravac discussed the switch to the GUI interface in the 1980s, then the use of smartphones in the past decade, and said that voice interaction removes the need for the screens that these earlier technologies required. Each of these interface approaches is, in its way, redundant; anyone who has worked with Terminal knows you can do nearly everything from the command line. But taking excess CPU horsepower and using it for a GUI made for a transformative experience. Likewise, you interact differently with technology when you’re talking with it, even if that conversation is an “unnecessary” additional way of doing things.

Voice recognition has made massive improvements in the past 20 years. In 1995, 100 percent of all input resulted in voice recognition errors, but today, technology is about as good as humans are at understanding what you say. (People have trouble recognizing what’s said about 5 percent of the time, apparently, so maybe you’re being unfair to Siri.)

Voice interaction is the emerging interface for robots and other technology similar to Amazon’s Alexa [3] or Google Home [4]. DuBravac called this a new era of “faceless computing,” where you can get things done while looking at other things; color me skeptical about whether it’s better. Yes, Alexa can call you an Uber, but I like having my phone show me that the driver is picking me up in front of my house, instead of the back alley that Uber keeps sending them to. That said, for casual interactions, voice is excellent. Look for new capabilities to make this better, such as services that can tell who is speaking based on their voiceprint, allowing for things like customized search results and parental control of their kids’ interactions.

DuBravac’s next trend is the infusion of artificial intelligence into a wider range of applications. The falling price of both chips and software allows for reasonable intelligence to be installed in just about anything, although I think there’s too much hype around both the terms AI and “smart,” which can apply to nearly anything. Is a refrigerator with sensors and a computer chip “smart” or “artificially intelligent?” No… but on the other hand, it might take care of small chores that no human would bother to do, like micro-adjusting its temperature to best suit the foods it’s storing.

Likewise, it’s a bit hard to tell when a new device is usefully smart, or when it’s taking “smart” a step too far. For example, it would be trivial for a washing machine to count the number of loads it has done, keep track of how much detergent you have left, and automatically order a refill when you need it. Alternatively, you can buy an Amazon Dash button [5] and push it when you need more Tide. Or you could just remember to buy detergent. DuBravac suggested that in the future, maybe 40 to 50 percent of all of our household purchases could be made by designated agents like smart washers.

I have no issues with removing minor annoyances from my life, but it seems a bit Brave New World to me if my appliances do my shopping without any interaction on my part. Perhaps I want to change detergent brands, or some friends gave me an extra bottle when they moved, so I’m not out when the washer thinks I am.

Beyond that, I note that this futuristic vision requires an affluent lifestyle that reliably remains affluent. I doubt the average person living paycheck to paycheck would be well served by having so many automatic deductions from their bank accounts, and woe to the laid-off person who suddenly needs to remember to cancel a hundred different automatic orders.

This leads us to the smart home, part of DuBravac’s next trend, that of widely connected everything-to-everything. He compared smart homes to the adoption of the dishwasher, which was invented in 1893(!), but took until the 1980s to make its way into half of U.S. homes (penetration is around 80–85 percent today). The jump in adoption didn’t happen until the necessary infrastructure was adopted after World War II: indoor plumbing, electrification, and the standardization of kitchen countertop heights.

DuBravac sees ubiquitous Wi-Fi and cellular broadband as the standardized countertop of the smart home; it’s the prerequisite that allows the smart home to be built up piecemeal, from many competing vendors, as opposed to the whole-house model that made little headway in the market previously. Smart home gadgetry will be a global $25 billion market this year, with $3.5 billion of that in the United States. On the downside, market competition that allows on-the-fly adoption also makes for competing standards, which slows mass market adoption.

Rounding out DuBravac’s home connectivity trend will be wearables that enable your house to monitor you. Whether it’s biometric feedback designed to improve your health or interactive environment sensing that lets your home know whether you’re there and where you are, technology you carry on your person will integrate you into a data set about your home, which will let your home be responsive to your needs and expectations.

The path to making smart things smarter comes from the interconnected nature of many of these devices; when one device senses an error condition or learns something new, it can improve the responsiveness of all other devices in its class. That’s how self-driving cars went from an 11-minute failure in 2004 to millions of test miles successfully driven on the roads today. This kind of aggregation puts many devices on an exponential learning curve, enabling rapid development from beta testing to seriously road-tested.

Of course, it’s one thing to predict overall trends; it’s another to see how 2017’s devices are living up to these models. And that’s one thing I’ll be sure to look for in the demos at CES.