Sometimes the future shows up so fast it hits us in the face, like a brick wall in a VR headset. Other times the miraculous promises of technology—the rearrangement of our very DNA, the blockchain-enabled toppling of Facebook—are frustratingly slow to arrive. But either way, the future is coming, and we should be ready. In the following pages we lay out a series of predictions, starting with some changes that are immediately upon us. Then, looking down the road, we get ever-bolder in our prognostications, year by not-so-far-off year. —The Editors
Cyberattacks Will Hit a Power Grid Near You
Hackers who disable power grids and detonate gas pipelines have been the villains of popular cyberparanoia for decades. Until recently, they were as easy to dismiss as Die Hard’s Bruce Willis–thwarted infrastructural cyberpocalypse, circa 2007. But hackers are starting to catch up with Hollywood.
Over the past few years, each of these hacker attacks on infrastructure could have caused serious physical disruption or destruction.
CrashOverride (aka Industroyer)
This Russia-linked malware switched off up to a third of the electricity in Kiev. Its modular structure means it could easily be adapted to other countries—and potentially planted on many targets at once.
Designed to disable systems that regulate conditions in physical plants, this malware instead triggered a shutdown of an oil and gas facility in Saudi Arabia last year.
Russian hackers recently gained access to multiple US power utilities. The intruders tunneled deep enough into some networks to screenshot control-system interfaces. According to security firm Symantec, they could have started flipping switches.
Potential cost to the insurance industry of a cyberattack on the US electrical grid.
In just the past year, researchers have uncovered two unprecedented pieces of malicious code that targeted industrial control systems. One, linked to Russian hackers, cut off up to a third of the electricity to the Ukrainian capital in 2016. Another, of more mysterious origin, shut down an oil and gas facility in Saudi Arabia last year. That second digital weapon had the ability to silently turn off safety systems, which could have led to a lethal fiasco. It’s been almost a decade since the NSA reportedly sabotaged an Iranian nuclear facility with malware called Stuxnet; now the rest of the world is entering the arms race.
Those recent incidents, along with dozens of other infrastructure attacks that used more traditional hacking tools, are part of an “extreme uptick” in government-sponsored hacker groups targeting industrial control systems, says Robert M. Lee, founder and CEO of security firm Dragos and a former NSA analyst. Last year, his analysts counted five new groups of nation-state hackers focused on infrastructure targets—including at least one Russian group that the US government believes gained deep access to a handful of US utilities. Digital sabotage of electrical grids, water systems, and petrochemical facilities can send threatening signals and test tools that might come in handy in future wars. As countries work to keep up with their adversaries’ hacking techniques, “there will be a rush for everyone to build these capabilities,” Lee says. “The losers will be civilian infrastructure owners.” Oh, and anyone who uses electricity too. —Andy Greenberg
Robots Will Roam Abandoned Big-Box Stores
The retail carnage in America feels unstoppable, as empty malls turn the suburbs into a hellscape of bleak parking lots and shuttered buildings. But, if they haven’t yet gone bankrupt, a few big-box operators and mall stalwarts see a glimmer of light: They’re going dark. Instead of abandoning nonviable stores, though, these companies are thinking about turning them into warehouses and fulfillment centers. Because we’ll still require stuff in our mobile, I-want-it-now, same-day-delivery future, and that stuff requires logistics. After closing 63 of its outlets this year, for example, Sam’s Club has begun converting around 10 of them into distribution centers for its ecommerce operations.
3 Shades of Dark Store
Full conversion to a warehouse. This option accepts that the future of shopping is largely online.
Some stores could pull double duty. By day, they’re open for business; by night, robots and packers fill online orders.
Chieh Huang of Boxed sees a far-off future in which—if physical stores survive at all—robocarts operate alongside shoppers and employees. Workers could help with the ecommerce in between stocking shelves and assisting customers. It would eliminate the need for nighttime pickers— so long as customers don’t kick over the robots. Or vice versa.
Many of these retailers are considering using robots to assist with the picking and packing. Amazon already has its somewhat reliable warehouse bots, and competitors are experimenting with their own setups. Boxed, a wholesale ecommerce startup, has built autonomous guided vehicles, essentially robot carts, that roam its warehouse aisles while a crew of humans pick items from shelves. When a Boxed cart scoots up to an item it needs and flashes a red light, a worker drops the object onto the cart. These picker carts have led to “extreme savings on capital expenditures,” such as acquiring permits and installing traditional conveyor belts, says CEO Chieh Huang. Legacy retailers have asked to license the technology, but Boxed has demurred.
Other startups are building similar tools, including Kindred, an outfit that makes warehouse robots and is piloting its system with Gap. And former executives of Kiva Systems, which Amazon acquired in 2012 to create the robot system it uses today, now have a new startup called 6 River Systems, which has developed a rolling warehouse robot named Chuck—the personal shopper you’ll never meet. —Erin Griffith
We’ll Share Our Emotional State as Willingly as We Share Our Photos
By now we all know that sharing our personal data is the price of free stuff online. We clicked Yes to terms of service (that we never read), agreeing to relinquish our photos, locations, the things we “like”—all just to log on to a social network or take some (seemingly benign) quiz. In exchange we got free email, music, searches, a parking spot in the cloud. But the cost turned out to be steep, making us marks for Russian instigators, hackers, and targeted ads that seem to know us a little too well.
In Your Face
Facebook has explored facial-recognition cameras for brick-and-mortar stores that scan the crowd and relay emotion to clerks—potentially also matching your face with your Facebook profile. If you’re trustworthy enough, the system could automatically unlock a secure display case.
Amazon will reportedly analyze your voice to recognize your mood and develop Alexa’s conversational skills. It could recognize frustration, not just from the words you use but from your tone of voice.
Sonos, the speaker company, filed a patent for technology that could customize playlists based on the emotion in your voice or biometric data obtained from a wearable device, like perspiration or heart rate, all cross-referenced with your listening history.
The information we share is about to get even more intimate. For years, companies have been trying to pry into our minds, looking for clues in our facial expressions. It was easy enough to recognize a smile but much harder to figure out what it meant. Advances in deep learning, however, have automated the process of finding signals that indicate subtle changes in mood.
So while early “emotion AI” was used in the lab by advertisers to learn how to make stuff more irresistible, now companies are preparing to use the tech directly on you. Consider iPhone X’s Animoji feature, which turns your face into a dynamic emoji (like a fox or a panda) that mimics your expressions. Beneath that cutesy tool, Apple is analyzing more than 50 facial-muscle movements. Meanwhile, emotion-recognition companies like Affectiva, Beyond Verbal, Kairos, and EmoShape are finding new applications in self-driving cars, health care, personal robotics, and gaming.
In theory, we’ll click Yes on the “emotion” terms of service because the data will hyper-personalize our products. Siri and Alexa will be better conversationalists; shopping experiences will feel ultra-tailored. Is it invasive? Prone to abuse? Absolutely. But if the deal we’ve struck with social media giants is any indication, we’ll consider sharing our emotions a bargain, at least until the wrong app gets a peek inside our head. —Nitasha Tiku
We’ll Crispr the Hell Out of Things—but Not, at First, the Way You Think
In the brief history of Crispr, nothing has gotten more attention than the possibility of using the gene-editing technology to cure disease. Human trials are just beginning for Crispr therapies that would fight cancer and treat sickle cell anemia. But those won’t be the first medical applications of Crispr to hit the market. No, that honor will go to diagnostics that are similar to a home pregnancy test, except that when you pee or spit or bleed on them, they reveal whether you’ve contracted a virus or which mutations are driving your cancer.
The very precise Crispr technology, which snips DNA to introduce or excise genetic material, can be programmed to target a genetic sequence known to reside only in a specific genome. With a bit of extra engineering, the system can also trigger a flare that says, “Hey! There’s dengue in here!” (Or Zika or, in theory, drug-resistant staph.) Since last year, pioneers in the Crispr field, Jennifer Doudna and Feng Zhang, each published papers showing this detection capability. Now their groups are commercializing the tests, which promise to be way more sensitive than some common diagnostic techniques and deliver results in hours, all remarkably inexpensively.
A Different Diagnosis
The lowest concentration of genetic material in a sample that Sherlock can pick up, making the diagnostic much more
sensitive than traditional assays.
Sherlock’s typical time to disease detection (a colored line appears on a paper strip).
In April, Doudna joined a team of researchers in launching Mammoth Biosciences to begin prototyping such tests. The company is pursuing partnerships with liquid-biopsy companies to pinpoint cancer mutations, but it plans to eventually expand Crispr’s detection capabilities to agricultural and industrial applications.
Zhang’s group is exploring a licensing strategy to bring its system, called Sherlock, to the developing world as cheap tests for infectious diseases. International trials will begin later this year in Nigeria, which has seen a recent outbreak of Lassa Fever. Computational geneticist Pardis Sabeti’s team of virus hunters plans to use the tests to track the spread of the disease and, hopefully one day, help local health authorities contain similar outbreaks. It might not be a therapy, but it could be the first time Crispr saves a human life. —Megan Molteni
Robotrucks Will Crisscross the Country
The long-haul trucking industry is rumbling toward a crisis. The general crumminess of the job—weeks at a time away from home and family; long, boring hours—makes it so hard to find and keep drivers that, right now, this $676 billion industry needs about 50,000 more humans just to keep up with rising demand from a growing population and improving economy. By 2026, that number could hit 174,000, according to the American Trucking Associations, the industry’s largest trade group.
Humans in the Loop
Companies building robotrucks have different ways to plug in the human for the first and final miles.
As promoted by Uber, a self-driving truck traverses the country, pulling off at its exit into a special zone where a human-piloted cab hooks up to the trailer and takes over.
A startup called Starsky wants humans to steer through narrow streets and difficult turns when needed—but from up to 500 miles away, sitting in front of what looks like a racing videogame setup.
Volvo and Daimler like platoons, in which a convoy of semis is connected by a wireless system. When a driver leading the pack brakes, all the other trucks automatically brake. This lets them run close together, cutting fuel-wasting wind resistance. The next logical step: eliminating human drivers and letting the robots convoy unassisted.
You know who doesn’t get bored or sleepy, miss their family, or demand a salary and benefits? Robots. And the highways on which trucks spend nearly all their time make for the easiest kind of computer-controlled driving. Just stay between the lines, don’t hit the car up ahead, and keep on rolling.
So it’s no real surprise that traditional automakers Volvo and Daimler, tech titans Uber and Waymo, and Elon Musk’s Tesla are all working on self-driving trucks. (To give itself a head start on understanding the trucking industry’s complex logistics, Uber has even started a service to connect human drivers to cargo that needs hauling.) Each has its own approach for how the trucks would drive on more challenging streets in cities and suburbs, where human navigators will still be needed, but the machines are already on the job. In several pilot runs, robotrucks have hauled beer across Colorado, refrigerators through the Southwest, and water for hurricane relief in Florida. And as soon as federal regulations make it easier for them to move between states, they’ll be Jack Kerouac-ing their way down the open road, (almost) no human drivers required. —Alex Davies
You’ll Go to Work in Virtual Reality
Companies spent $4.7 billion on virtual-, augmented-, and mixed-reality business applications in 2016. That’s projected to more than triple to $16.3 billion by 2020, with AR/MR overtaking VR.
First will come the ability to use your usual software and apps in VR. That’s already arriving: Apps like Bigscreen and Virtual Desktop allow you to interact with your desktop windows in VR environments. You can use your regular keyboard (fast, if you can touch-type) or “press” letters on a floating virtual keyboard (very, very slow). Logitech is even developing a kit that will let users track their IRL keyboards in VR.
One hurdle is that it takes a beast of a computer to handle the CPUsurping demands of high-end VR headsets. The laptops that can do this? Ginormous. But one company, QuarkVR, has managed a Silicon Valley–like feat of compression. It offloads all the processor-melting stuff about VR to the cloud—and kicks it to your headset through something as small as a Macbook Pro.
Meanwhile, VR headsets are going wireless. The self-contained ones available this year and next won’t replace your work computer—they’re more like smartphones with an overachiever complex—but systems on a chip continue to beef up. Qualcomm’s newest VR development platform can handle eye tracking and maps your room so you can move around in it. And crazy augmented- and mixed-reality headsets aren’t far behind. Hell, Microsoft’s Hololens is already self-contained, even if it’s got a tiny field of view.
So you’re looking at five years before you can ditch that monitor and strap on a wearable device. It might be VR, it might be AR; it might even do both. It’s still gonna be on your face, though, so take it off before your meatspace meetings. If you have those anymore. —Peter Rubin
The Blockchain Will Rebuild the Internet as We Know It
For internet idealists, the blockchain is an intoxicating technology. Blockchains support cryptocurrencies, yes, but these high-concept databases also allow many online services to have distributed ownership—that is, they don’t have to be controlled by a central authority. Want to ditch Uber? At least three groups are building ride-sharing apps using the blockchain. Care to stop shopping on Amazon? A rudimentary decentralized marketplace called OpenBazaar is available. Need to store some data? Sure, you can tap into other users’ spare disk space using the blockchain-based Storj. No other architecture offers such a radical alternative to how things work today—one that could bring data- and power-hoarding megacorporations crashing down, to be replaced by decentralized alternatives.
The approximate number of US households that could run on the energy consumed by the bitcoin network. New bitcoin transactions require a computing-intensive calculation called “proof of work.” Future blockchain systems will need to be more efficient; one alternative is known as “proof of stake.”
“We now have a way to build data systems and applications that doesn’t create this monster at the center, this Uber, this eBay, this Twilio—this central hub that we are all sharecroppers on,” says Brian Behlendorf, executive director of the blockchain consortium Hyperledger (and the chief engineer behind wired’s first website). Think of it this way: Whereas Facebook stores your data on its servers, users of a social network like Steemit record their posts in a blockchain. Some users can contribute spare computing power to help maintain that blockchain, and get compensated for the work in the service’s token, Steem. Facebook makes money by exploiting your personal data; the people who maintain Steemit’s blockchain earn rewards for keeping the platform secure.
Once blockchain technologies mature, they will recede into the background to become one of the fundamental systems powering the internet. That may sound boring, but for a technology seemingly beset by Ponzi schemes and hucksters, boring may be a blessing. —Sandra Upson
This article appears in the June issue. Subscribe now.