Hustling is the default mode of the 21st century, and I'm not above listing my adorable split-level Victorian on Airbnb during my out-of-town weekends. Need to rent a car for the day? Take mine—I wasn't using it anyway. But whoring out my bed—my own private sanctuary, complete with sweat-stained sheets and raggedy stuffed elephant named Elephant—on Recharge, the “Airbnb for naps”? I'd rather sell a kidney. The tech industry thinks that every last inch of my personal space should be for hire, that strangers should be able to rent it, on demand, by the hour, at their convenience. I call it, with eye roll heavily implied, the sublet economy. Initial moves toward the micropersonal seemed sane enough: Share the extra storage space in your garage (Spacer) or the empty parking spot in your driveway (Pavemint, CARMAnation) or that boat you spent way too much money on (Boatbound, Antlos). But now we can't look past our own noses without seeing dollar signs and feeling the guilt of unmonetized potential. Nothing is sacred, not even your laundry room (Laundromatch—now defunct, ha!). With all those student loans, can you really afford to leave your kitchen vacant instead of entrusting it to someone else's dinner party (Feastly)? You know that very relatable problem where you have a toilet that's just sitting there, not generating any revenue, most hours of the day? Put it on Airpnp, the “Airbnb of toilets”! When I volunteer to host my most intimate spaces, I sacrifice some of my basic human dignity. What's next, a service for renting out my fresh, youthful blood? What's mine is not yours. Unless you'd like to help me scrounge up the cash for a down payment on a house. Did I mention I had a kidney for sale?
Guidebooks highlight San Francisco’s Hayes Valley neighborhood for its lively bars and restaurants, nurtured by the removal of an earthquake-damaged freeway and swelling tech industry salaries. At Uber’s headquarters nearby, data scientists working on the company’s food delivery service, Uber Eats, view the scene through a more numerical lens.
Their logs indicates that restaurants in the area take an average of 12 minutes and 36 seconds to prepare an evening order of pad thai—that’s 3 minutes and 2 seconds faster than in the Mission District to the south. That stat may seem obscure, but it’s at the heart of Uber’s bid to build a second giant business to stand alongside its ride-hailing service.
Uber is fighting other well-funded startups and publicly listed GrubHub in the fast-growing market for food delivery apps. Winning market share and making the business profitable depend in part on predicting the future, down to the prep time of each noodle dish. Getting it wrong means cold food, unhappy drivers, or disloyal customers in a ruthlessly competitive market.
The mobile apps of Uber Eats and competitors such as DoorDash list menu items from local restaurants. When a user places an order, the delivery service passes it along to the restaurant. The service tries to dispatch a driver to arrive just as the food is ready, drawing on a pool of independent contractors, like in the ride-hailing business. Meanwhile, the customer is shown a prediction, to the nearest minute, of when their food will arrive.
“The more detail with which we can model the physical world, the more accurate we can be,” says Eric Gu, an engineering manager with Uber Eats’ data team. The company employs meteorologists to help predict the effect of rain or snow on orders and delivery times. To refine its predictions, it also tracks when drivers are sitting or standing still, driving, or walking—joining the growing ranks of employers monitoring their workers’ every move.
Improved accuracy can convert directly into dollars, for example by helping Uber combine orders so that drivers carry multiple meals without any getting cold. Drivers get a small bonus for ferrying multiple orders on one trip. “We can save on delivery costs and pass back some savings to the eater,” Gu says.
Four blocks away, Uber rival DoorDash has its own team of data mavens working on an AI-powered crystal ball for food deliveries. One of their findings is that sunset matters. People tend to order dinner when it’s dusk, meaning they eat later in summer and shift their habits when the clocks change in spring and fall. Like Uber, the company keeps a close eye on sports schedules and weather patterns, while also tracking prep times for the dishes offered at different restaurants. Company data indicates that pad thai takes 2 minutes longer to prepare Friday through Sunday than during the rest of the week, because kitchens are busier.
Rajat Shroff, vice president of product, says DoorDash data also clearly shows the connection between accurate delivery predictions and customer loyalty. “That’s driving a big chunk of our growth,” he says. The company was valued at $7 billion this month by investors who plowed in $400 million of fresh funding.
DoorDash has also been working to better understand what happens in restaurants, for example by connecting its systems with Chipotle’s in-house software so orders can be sent in more smoothly, and DoorDash can track how they’re progressing. The company has built a food-delivery simulator in which past data is replayed to test different scheduling and prediction algorithms. Both DoorDash and Uber use their data to offer drivers more money to head to areas where demand is expected to be strong.
Analytics company Second Measure says credit card data shows that DoorDash overtook Uber Eats for second place in US market share in November, behind GrubHub. As of January, the company says, GrubHub took 43 percent of food-delivery sales, compared with 31 percent for DoorDash and 26 percent for Uber Eats. DoorDash is a customer of Second Measure.
Still, DoorDash says it gets orders to customers in an average of 35 minutes. That’s slightly slower than the 31 minutes Janelle Sallenave, head of Uber Eats for the US and Canada, says her service averages for the US.
Uber’s data scientists have a potentially big advantage over their competitors: the rich live and historical traffic data from the company’s ride-hailing network. The company is also digging more deeply into its data on restaurants and Uber Eats drivers.
One project involves analyzing the language on restaurant menus. The goal is to have algorithms predict prep times for dishes it doesn’t yet have good data about by pulling data from menu items that involve similar ingredients and cooking processes.
Chris Muller, a professor at Boston University, says the data-centric view of dining taken by Uber Eats and its competitors is helping to drive a major upheaval of the restaurant business. “This is the biggest single transformation since we saw the growth of fast casual” chains like Chipotle that promise speedy meals of higher quality than fast food.
Joe Hargrave, who grew a farmers’ market stand into five Bay Area taco shops, is living through the food app transformation. He designed his Tacolicious stores for people who share his love of good food you can eat with your fingers while watching baseball. Now, more of his customers are eating their tacos at home, and delivery has become a lifeline.
Orders via apps including DoorDash and Caviar make up about 12 percent of Hargrave’s business, he says. They’ve helped revenue grow 8 percent over the past year, even while in-store business shrank by roughly a quarter. He appreciates what the apps do, but accommodating the delivery boom hasn’t been easy.
“I’ve spent my whole career trying to figure out how to put the best product in front of people,” Hargrave says. “Now I’ve been thrown this curveball where I have to put it in a box.” Tacolicious switched its register system to better handle delivery orders without compromising in-store service. There’s now often a person in each restaurant working exclusively on packaging and checking delivery orders.
Muller and Hargrave say the app-and-algorithm approach to dining can squeeze conventional restaurants and could even put some out of business. Uber’s standard cut of each order is 30 percent, a significant bite in a traditionally low-margin industry. Even restaurants like Tacolicious that accommodate delivery services must also serve people who walk in the door.
That’s one reason Uber is encouraging the development of “virtual restaurants,” which operate out of an existing restaurant’s kitchen but sell only via its app. Uber said last year that it was working with more than 800 virtual restaurants in the US; many operate during hours when a restaurant’s main business is slack or closed, allowing more efficient operation and use of the property.
Uber and DoorDash also work with so-called dark kitchens, operations that serve only via delivery apps and can be more efficient and predictable than conventional restaurants. DoorDash operates a 2,000-square-foot kitchen space in the Bay Area that it rents to such operators.
Muller likens the arrival of Uber Eats and others to how online travel sites shook up the hotel industry, forcing hoteliers to adapt their business models to a market where consumers are more engaged, driving more visits, but at lower prices.
How lucrative this new form of restaurant business will be is unclear. Uber has previously said its service is profitable in some cities, but financials released for the last quarter of 2018 didn’t offer detail about Uber Eats. In all, the company said it lost $940 million, 40 percent more than the previous quarter. In the third quarter of 2018, the company said Uber Eats accounted for 17 percent, or $2.1 billion, of its worldwide gross bookings.
GrubHub has been consistently profitable since it went public in 2014 and sold $1.4 billion worth of food in the final quarter of 2018, an increase of 21 percent over the previous year. But it also reported a small loss after a big jump in marketing spending. GrubHub’s management told investors that competition wasn’t harming growth, but analysts interpreted the company’s results as showing how the rise of DoorDash and Uber Eats will put all the delivery apps under pressure.
Uber and DoorDash both declined to provide more detail about their businesses but are rapidly expanding their reach. DoorDash says it covers 80 percent of the US population, and Uber Eats claims to have reached more than 70 percent, in addition to serving more than 100 cities in Africa, Asia, and Europe. Sallenave, the Uber Eats head for the US and Canada, predicts eating via app will become the norm everywhere, not just in urban areas. “We fundamentally believe we can make this business economically viable, not only in large cities but also in small towns and in the suburbs,” she says.
Ari Walker had been working in the wine business for a few months when the dreams started. He didn’t know much about wine; he’d left college and taken a job at a distributor because his wife was pregnant and they needed money. But the more he tasted and read, the more entranced he became. Soon she was shaking him in his sleep, telling him he was mumbling about food pairings. “You mentioned Nebbiolo,” she’d say, referencing an Italian grape variety. “And blood sausage.”
In 2001, after a few more jobs in the industry, Walker started an import and distribution company with a partner: Kevin Hicks, an entrepreneur who’d made a fortune with an online rating system for doctors and hospitals. By the time we met, at a wine event in Boulder a few years later, he had amassed an impressive portfolio and was living an enviable lifestyle. But his business was going broke.
Walker was spending much of his time tracking down unusual wines from viticultural regions around Italy. They had singular flavors and compelling stories. But the vast majority of American wine drinkers, he’d come to understand, have little interest in those stories. They want wine that tastes good and doesn’t cost much. So Walker and Hicks created a cheap brand that could be sold at volume to subsidize the imports, but that didn’t work, either. There were too many in the market already, all trying to solve the same problem with a mediocre product. “The question we tried to answer was, how do we make these generic wines better?” Walker says. “We looked at all sorts of stuff but had a hard time moving the needle.”
The breakthrough started with baby food. In 2012, Hicks was about to become a father. He started wondering what, exactly, was in the organic, premium-priced products that he and his wife were planning to feed their newborn, so he sent samples off for laboratory analysis. “If you know Kevin,” Walker says, “you understand that that’s just totally something he would do.” When the bills—as much as $1,500 for a single sample—started to add up, Hicks created a lab of his own, which he dubbed Ellipse Analytics. He had a bigger plan. He invested several million dollars in equipment and hired a team of scientists and technicians and before long, Ellipse had enticed paying clients to commission chemical breakdowns of entire consumer categories, like protein powders and sunscreens. Walker saw the potential for wine, and he pushed Hicks to use his technology for their own business.
Like anything else, wine is a combination of chemicals. Ellipse can test for some 500 different attributes and measure the results at the parts-per-billion level. Hidden in that data, Walker realized, were the precise combinations of esters and acids and proteins and anthocyanins and other polyphenols that make a wine taste creamy or flinty, or give it aromas of blueberries or vanilla or old leather—the chemical compositions of America’s most popular wines. Walker also knew that most wine gets a boost from additives such as Mega Purple (for color), oak extract (for tannins and flavoring), and similar chemistry-set concoctions. Using cheap surplus wines readily available on the bulk market and blending in natural additives, he thought, it might be possible to make some pretty convincing copies of popular premium wines.
In 2015, Walker and Hicks started Integrated Beverage Group and set out to duplicate wines that they knew Americans already liked. They planned to do this in plain sight, naming their brand Replica and urging consumers to compare their products with well-known names that usually cost as much as double the price. It didn’t take long before they realized that, in most cases, even professional critics couldn’t distinguish their facsimiles from the originals.
In a gray concrete building, part of a grim-looking industrial complex north of downtown Denver, four glasses of wine are lined up at each place around a conference table. It’s a Tuesday afternoon at the IBG offices. Walker sits across from me, wearing a trim beard and a sweater over a button-down shirt. Next to him is a scruffy man in his thirties who has the chemical structure of dopamine tattooed on his left arm. That’s Sean Callan, a PhD chemist who runs the Ellipse Lab.
Brett Zimmerman, one of fewer than 250 certified Master Sommeliers in the world, is at one end of the table to my right, typing notes into a laptop. To my left is IBG’s winemaker, Everett, who has just arrived from California. I’m calling him Everett and not his real name because he also works as an enologist for a large American wine producer, precisely the kind of industry giant that IBG is looking to undercut. “If they found out I was doing this,” he tells me, “it wouldn’t turn out well.”
Two years after IBG was started, Replica wines are sold in 49 states (everywhere but Iowa), in major retailers such as Publix in the south and Winco in the west. Because of the substantial investments needed to build the lab and start the brand, the company isn’t yet profitable. But by other metrics, its concept has been a remarkable success. Sitting in the conference room, I watch the process unfold.
The sample to the far left at everyone’s place is the one the IBG team is trying to match, a 2015 Far Niente Chardonnay. A highly respected name in California wine, Far Niente makes Chardonnays that sell for $60 in retail shops and $100 on wine lists. The wines have a singular style that places them somewhere between the robustness of most Napa Chardonnays and the nuanced flavors of white Burgundy. A West Coast retail chain has placed an order with IBG for a proprietary Napa Chardonnay with attributes that track Far Niente’s. The deadline to ship the wines, I’m surprised to learn, is only two weeks away. Yet the IBG team is still in the preliminary stages of tasting and blending. “We’ll get it done,” Walker assures me. He has already named the wine Per Sempre. That’s “forever” in Italian. More important, it sounds like Far Niente.
The previous week at his office in Sonoma, Everett had tasted through more than 70 lots of Chardonnay that are on offer from a Napa-based wine broker for purchase in varying volumes— some only a few hundred gallons, others several thousand. He chose two that seemed as though they might be a fit in a potential blend. One was from a boutique vintner in St. Helena. The other came from a massive producer, one of America’s most famous, that grows grapes all over California and makes millions of bottles of wine each year for its numerous brands. He bought as much of the first wine as he could, and about the same amount of the second wine.
Glasses of the two of them also sit before each of us now, beside the benchmark Far Niente. To the far right in the lineup is a preliminary blend of the two potential components in roughly equal proportions. To help the team understand where it should be aiming, a graph projected on a screen identifies more than a dozen aromatic attributes present in most California Chardonnays, and to what degree the half-dozen or so most popular brands contain each of them. The data is culled from an Ellipse analysis, and the results from the various brands match almost exactly.
What consumers want in the category, it turns out, is remarkably consistent. In several areas, though, Far Niente is an outlier: most notably in the presence of citrus and the absence of butter and coconut. The Far Niente also has a far higher level than the other wines of malic acid, which is found in lime juice, and there’s a reason for that. Unlike both lots of the purchased wines (and the vast majority of Napa Chardonnays), it hasn’t undergone the secondary fermentation that transforms malic acid into lactic acid and changes the taste of green apple into cream or butter. I don’t get how Everett will be able to combine two wines that have undergone malolactic fermentation into one that tastes like it hasn’t, but he’s not concerned. “We can add back malic acid in the blending,” he says.
The coconut is harder. All three wines have been fermented in barrels rather than steel tanks. But different kinds of oak have different characteristics. “The Far Niente shows clove and raw wood,” Everett says, “rather than the caramel, vanilla, and coconut I’m getting from the others.” Back home, Everett has a table covered with vials of wood flavoring that he might be able to use to nudge the profile of the wine closer to Far Niente’s by literally blending it into the wine. “But barrel fermented wines are hard,” Zimmerman points out. “That’s probably the hardest thing we do.”
With certain red wines, which are easier to replicate than whites, IBG has come within a few percentage points of matching the components at a parts-per-billion level. That includes a whole lot of attributes that a wine drinker will never detect. What’s more crucial is nailing the handful of attributes that define the wine for the casual drinker, those points of difference that deviate from the norm. It’s like Alec Baldwin playing Donald Trump on Saturday Night Live. When you see him on the screen, scowling and doing that little turn of his wrist, you’re not fooled into thinking that’s really Trump. It obviously isn’t. But it’s equally obvious whom he’s imitating.
Walker and the IBG team try to do the same with wine. If they can hit the most blatant elements of a popular bottling using inexpensive bulk wine and a bunch of additives, they’ll be in business. Some perceive this as undermining those ineffable elements that make wine different from, say, toothpaste. In a full-page feature, the Santa Rosa Press-Democrat, a newspaper based in Sonoma County that is perceived as the voice of the California wine industry, characterized Replica’s products as “Frankenstein wines.” “While Replica wine doesn’t begin in a petri dish,” it said, “it is created, to a large degree, in a lab.” This doesn’t trouble Walker. “You could say it’s weird,” Walker says of IBG’s process. “Or you could say it’s our point of difference.”
The boutique Chardonnay matches many of the attributes of Far Niente. But adding too much of the other wine to it, all agree, pulls the blend in the wrong direction. We try a blend of three-quarters boutique and one-quarter mass producer, then one closer to 85 percent boutique. The result still doesn’t seem right. Eventually, Zimmerman, who has the best palate of the group, argues for not including any of the wine from the major producer. “Even in the small amount, it takes away the zippy tone that we’re getting from the Far Niente,” he says. That “zippy tone,” a palpable sense of energy coursing through the wine, is one of the most recognizable attributes of the Far Niente. If you’re imitating Trump, that’s the jut of his lower lip. If you don’t have that, you don’t have a match.
Everett admits that no combination of the wines matches the Far Niente better than the boutique wine by itself. But there just isn’t enough of it to make the number of bottles that the retailer has ordered. They need to blend that with something that’s less ideal, but not so much of it that it moves the wine noticeably away from the benchmark. “So the question becomes,” he says, “how much can you feather it up before you have a deal-breaker?”
The following afternoon, Callan walks me through the Ellipse facility, which consists of a single room in the same concrete building, down the hall from Walker’s office. As labs go, it isn’t a particularly large one. Along one wall, a woman in a ski hat is prepping samples of dietary supplements for one of Ellipse’s clients. In a storage nook off the main room, I spot a cart loaded with bottles of wine. “Those are California Pinot Noirs,” Callan says. “We’re doing that next.”
Over the past four years, Ellipse has analyzed thousands of wines, a formidable chunk of the American marketplace. More than anyone else, it is safe to say, the IBG team can scientifically define what the most popular wines taste like. “Only we, uniquely, have this data to say, ‘If you like Goldeneye, we know exactly why you like Goldeneye,’” Walker says about a California Pinot Noir that IBG will soon be trying to replicate. “And we know what else you’re likely to like. And what you won’t.”
IBG can’t replicate every wine. Those with singular attributes, like wines made from grapes grown in a specific vineyard or from a hard-to-find variety, are far more difficult, bordering on impossible, for the simple reason that all of IBG’s wines start with those surplus lots being sold on the bulk market. When I ask Zimmerman if they could replicate a small-batch Shiraz from a producer in Australia’s Adelaide Hills that is a particular favorite of mine, he rolls his eyes and says there’s no way.
But the world’s most popular wines—from Kendall-Jackson Chardonnay to Dom Perignon—are made hundreds of thousands of bottles at a time, enough volume that their grapes are sourced from a range of vineyards. “The reason K-J is so successful,” Hicks says about Kendall-Jackson, “is that it tastes consistent, year after year, bottling after bottling. You know what you’re going to get, like Coca-Cola or Campbell’s soup.” If Kendall-Jackson is using what seems like a fairly exact recipe to make each vintage of its wines, Hicks figures, there’s no reason that IBG, with its reams of scientific data, can’t match it.
Two weeks later, I get on a plane to California and visit Everett at the office he uses in downtown Healdsburg, in Sonoma County. The building happens to be an old winery where some of the earliest California Zinfandels were made in the late 1800s. Everett is still making wine there, in a sense, doing things like adding malic acid to the Per Sempre blend to simulate the energy of the original. He then overnights the results back to Colorado for Zimmerman to taste and Ellipse to analyze. That back-and-forth gets them closer and closer to their target. When I arrive, he pulls out a bottle of the finished Per Sempre, which is labeled and ready to be sold, and pours each of us a glass. “It’s much more Far Niente-ish, don’t you think?” he says. “See how the addition of the malic brought back a little of the flintiness?”
I do. Back at IBG, I’d submitted to a blind comparison test involving another of their wines, called Label Envy, which is meant to replicate La Crema Pinot Noir. Callan had poured two glasses of one wine and one glass of another and asked me to identify which two were similar. I thought I knew, but I was wrong. They’d hit their target precisely. The simulacrum Everett handed me seemed maybe a little clunkier, a little less graceful, than the actual Far Niente I had in Denver. But it definitely was closer. And without a glass of the original in front of me for comparison, or perhaps even with one, I could easily be convinced that it was the same wine.
If you’re an average drinker, you aren’t interested in parsing the flavors in your Napa Chardonnay or constructing a critical analysis. You just want to sip a glass of something nice with roast chicken. For that, the Per Sempre might serve the purpose just as well as the Far Niente. The packaging looks handsome enough that you wouldn’t hesitate to bring it to a dinner party, and it would cost $25 as opposed to the $50 you’d pay for the Far Niente “I’d drink it,” Everett pronounces.
After tasting the wine, we drive toward Everett’s house on the outskirts of Healdsburg. We cross a small bridge, then continue down a gentle hill. We rumble down a country lane, past small vineyards where the buds are just starting to break. Being here, you can’t help but feel the attraction of the tales people tell about wine, including how they explain the attributes in each bottle that make it a topic for contemplation and not just consumption.
That freshness? It’s from the difference in temperature between the warm nights and cool days in this particular valley. That resistance in the mouth, the little push-and-pull that can taste like a tea bag left too long in hot water but provides a framework to help offset the plush fruit? That comes from the ocean wind that toughens the skin of these grapes. “When you get out in the country, there’s a certain amount of romance,” Everett says. He’s quiet for a moment. “But there is also a chemical and scientific aspect to this, too,” he continues. “It’s the juxtaposition of those things that’s attractive to me.”
We step inside a cottage beside his house, which is where he does much of the tasting and blending for Replica. A table off the kitchen is covered with small bottles, samples of bulk lots of Pinot Noir available for purchase that will be used to help match Goldeneye. In the work he does for his day job, Everett starts with grapes as his raw material. He oversees the fermentation process, which lets him make decisions that will go a long way toward determining how a finished wine tastes. But he’s also at the mercy of what his employer’s vineyard holdings have given him. When he starts with wine that already has been made, he gets to taste dozens of possibilities from sites all over Napa, Sonoma, and beyond. And then he works with only those that he chooses.
“This is someone’s reject wine,” he says, pouring out a Pinot Noir sample for me to taste. “But why was it rejected? Was it great wine that just didn’t match stylistically with what they were trying to do? Was it second-best to the other lots that they had? Did they just have too much of it? And at the end of the day, does it matter? Not to me.” He holds the glass against a piece of white paper to get a better look at the color of the wine. He takes a sip. “Now, maybe this one is just a little too tart for whoever made it,” he admits. “That’s why they sold it off. But we can fix that, too.”
Before leaving California, I stop at Far Niente. I’ve visited before, and I never get tired of it. The setting is delightful, an 1885 winery building surrounded by gingko trees and plum blossoms. The late Gil Nickel, who renovated the disused winery and created Far Niente in the 1970s, started in Oklahoma as a horticulturist. He wanted to make Far Niente the gem of Napa Valley, as beautiful as any winery in the world.
That landscaping needs to be maintained, of course. So does the collection of classic sports cars in an adjacent barn. Far Niente employs several winemakers, and also a team of gardeners, and chefs who prepare the lunches it serves to wine club members in a clearing overlooking the vineyards. When you buy a bottle of Far Niente, you’re paying for the whole package: the wine itself, its reputation, and the enticing site that makes the narrative possible.
Yet, in a sense, the wines made at Far Niente are no more authentic than Replica’s. Even the finest wines exist as the sum of hundreds of decisions in the vineyard and the winery, each designed to help steer the wine—or manipulate it, if you want to use that word—in a desired direction. I have no idea whether Far Niente’s enologists typically acidify their wines to freshen them in warm vintages or add tannins to help balance soft fruit flavors, but plenty of wineries do—such additions are perfectly legal. It’s also within the rules in California to blend in as much as 15 percent of wine from a different vintage than the one on the label, and different grape varieties, and even grapes from somewhere else entirely.
And that doesn’t even get into other standard practices of winemaking, such as jump-starting fermentation with commercial yeasts and reducing alcohol levels by sending the wine through a contraption called a spinning cone column. None of these figure in the romantic narrative of letting nature make the wine. But if they make the wine we’re drinking tonight taste better, few of us would argue against them.
When I enter the old stone Far Niente winery, I’m offered a glass of Chardonnay off a silver tray. It’s the 2016, not the 2015 that Per Sempre is modeled on, but that same energy is on full display. I carry the glass out to a balcony off the main building and stand in the afternoon sunshine. The wine is delicious. There are vines below and olive trees and a view of Oakville and the hills beyond. I take a moment to notice that particular green-apple taste and the hint of cinnamon aroma that comes from the particular barrels used to ferment and age the wine. Did the Per Sempre I tasted with Everett have that too? Perhaps it did. The truth is, I can’t remember.
Oenophiles wax poetic about the look, feel, smell, and taste of red wine. But what’s actually inside the drink? Dozens of complicated molecules from the grape’s juice, seeds, and skin. Oh, and alcohol.
Two of the top 10 cookbooks in 2017 were devoted to the appliance, according to . Melissa Clark’s (Clarkson Potter, $13.93) has sold 150,000 copies since its October release; the columnist estimates that her latest hit outsold her previous 39 cookbooks, combined.
Mention the appliance to chefs and you’re most likely to draw a blank stare.
“What’s an Instant Pot?” asks Alex Stupak, chef and co-owner of Empellón in New York.
Officially, the device shouldn’t be in professional kitchens at all. “Our current products are designed and certified for household use only,” Yi Qin, vice president of product management at Instant Pot in Ottawa, Ontario, told Bloomberg by email.
One obvious reason the appliance hasn’t been embraced by the restaurant community is scale. The largest Instant Pot holds 8 quarts—a drop in the (stock) pot for most restaurants.
One of the few professional chefs who admits to having an Instant Pot in his restaurant is Jonny Hunter of the Madison, Wisc.-based Underground Food Collective. In fact, he has five. Hunter is a fan of the compressed cook times and precision that the device offers.
“Traditionally, it takes about 40 days to make black garlic,” he says, referring to the intensely sticky Asian flavoring. “I can do it in six hours.”
Most dishes can’t be sped up so rapidly by the Instant Pot, but Hunter argues that even modest time savings will add up for a busy cook. Take hard-boiled eggs, for example: “It takes you eight minutes in an Instant Pot; the regular way takes 12 minutes,” he says. “Chefs say, ‘Who cares about that difference?’ But I save four minutes each time, and they’re perfectly cooked.”
Garrison Price, of New York’s il Buco Alimentari, routinely does 250 covers a night, yet he still finds the low-yielding appliance useful for making goat-milk yogurt. He serves it as an accompaniment to leg of lamb with wild watercress and anchovies, as well as spice-roasted spring carrots with green almonds. Making yogurt the traditional way is “tricky,” Price says. “You don’t have to baby sit yogurt you’re making in an Instant Pot.”
Price believes chefs don’t use the Instant Pot because of the message they associate with it. “I think it’s the infomercial-ness,” he says.
In Houston, James Beard award-winning chef Chris Shepherd is experimenting with an Instant Pot to create batches of pho “dressing” for a carpaccio dish at his upcoming 80-seat restaurant, UB Preserv. “I got the idea from my manicurist; she’s a big Instant Pot fan,” says Shepherd. He first used one at a previous restaurant when he ran out of Korean-braised goat and dumplings. “My cook said: ‘We should bust out that Instant Pot we have in storage.’ We had more goat ready in 45 minutes.”
At the Latin restaurant Público in St. Louis, Mike Randolph cooks almost everything on an open hearth. Yet his Instant Pot has been used to produce items ranging from vegan chorizo stock to dulce de leche. Randolph agrees that a drawback for some chefs is perception. “There’s a hesitation in having a brand name like that in your kitchen. A lot of chefs want to keep things more traditional, with a stovetop,” says Randolph.
Why indeed doesn’t Instant Pot create a bigger model for professionals? The company appears to have already asked the same question. “We are looking into all opportunities to expand the electric pressure multi-cooker market. Currently, we don’t have a commercial offering. But nothing is impossible,” said Qin by email.
(Corrects the name of the restaurant in the 13th paragraph.)
Alexis Rivas opens his Mac laptop and zooms in on a 3D rendering of a house in Echo Park, a hip neighborhood in Los Angeles. Set off from the main house, there’s a small, modern structure that his company, Cover Technologies Inc., hopes to build. “You’ve got the kitchen here, a little stovetop, fridge,” Rivas says as he navigates around the 502-square-foot unit with his cursor. “And then we can take a walk around and go into the bedroom.”
It’s the kind of design that would typically cost a few thousand dollars in architecture fees, says Rivas, who co-founded Cover Technologies in 2014. The Los Angeles outfit can put together a proposal for just $250, using software to determine whether a specific property meets local and state requirements for adding a backyard unit. If building is allowed, the company designs one of its modular, factory-built structures to fit the plot. Homeowners often hesitate to take on a project like this, Rivas says over the whir of a drill in his company’s workshop, because “they’re expected to put a lot of time or money into the process without really getting a clear picture of what they can build.”
The housing crunch in many West Coast cities has revived interest in an old idea: the granny flat. Often called “accessory dwelling units,” or ADUs, the free-standing structures can be manufactured off-site and plunked in a backyard for about $150,000, including permits and site work. Some housing experts are promoting ADUs as a small way to address the affordability crisis in high-cost places such as Seattle, the Bay Area, and Los Angeles.
Lawmakers are warming to the concept, approving legislation to make it easier and cheaper to install ADUs. And unlike some other efforts to increase housing density, these measures generally haven’t been met with fierce opposition from antidevelopment groups. Perhaps that’s because ADUs can blend into single-family neighborhoods and let homeowners profit by owning rental units. “They might be the single most promising means of upping the housing supply that is also politically feasible,” says Issi Romem, chief economist at BuildZoom, a company that mines building permit data to help homeowners find contractors.
Seattle, Vancouver, and Portland, Ore., have all seen applications for ADU permits climb after issuing rules relating to their construction. California is playing catch-up: The state’s legislature passed laws in 2016 and 2017 removing parking requirements for ADUs, eliminating some utility connection fees, and streamlining the approval process. Los Angeles issued 721 permits for ADUs last year, a fivefold increase from 2016, according to Attom Data Solutions. San Jose, San Francisco, Santa Barbara, and Oakland also saw upticks last year.
While that interest is notable, ADUs aren’t a panacea for a state that for years has failed to keep pace with housing demand. California’s economy added 2.3 million jobs over the past five years. But the state issued permits for fewer than 480,000 new residential units over the same period, or about one home for every five additional workers.
Building enough backyard units to narrow the gap between supply and demand in any noticeable way will be challenging. An ADU is “a construction project that needs to go through zoning, regulation, financing,” says David Garcia, policy director at the University of California at Berkeley’s Terner Center for Housing Innovation. “The typical homeowner’s not prepared for that.” Many who are considering a backyard unit, he says, will want a “one-stop shop.”
A Portland-based startup offers a turnkey solution. Dweller Inc. covers the upfront costs of installing an ADU in return for a 25-year ground lease on the land where it sits. The company is responsible for finding a tenant and captures 70 percent of rental income. “We have the potential for this to be a very commonplace thing,” says Chief Executive Officer Patrick Quinton.
Dweller’s business model is untested—the company won’t install its first company-financed unit until June—as are those of several startups targeting the market. Seattle’s CityBldr started a service in March that streamlines the design and permitting process for ADUs. Cover, which has raised $1.6 million from Khosla Ventures, General Catalyst, and Fifty Years, has built only one of its backyard units, though Rivas says it has several in the pipeline.
As these businesses ramp up, they’re likely to run into a problem vexing more experienced builders: competition for materials and labor. Steve Vallejos, whose Valley Home Development has been installing prefabricated units in the Bay Area for more than a decade, is building his own factory after his manufacturing partners got busy with bigger projects. Studio Shed, a Boulder, Colo., company that’s installed more than 1,000 backyard units, including dwellings and workspaces, is concentrating on developing a network of builders, electricians, and plumbers to install ADUs. “There’s almost no upper limit in terms of the available places where people could put them,” says Jeremy Nova, the company’s co-founder. “That’s an opportunity for our business, but it’s very hard to find contractors right now.”
BOTTOM LINE – Seattle, Los Angeles, and Portland, Ore., have logged sharp increases in permits for accessory dwelling units following changes to zoning laws.
From the street, you can hear children at play. Inside the one-story house in Fremont, California, a fish tank gurgles by the front door. A plastic bin filled with Legos sits in the sun room. Renuka Sivarajan, 37, runs a home daycare here. Her path to this point has been like the stock market of late.
When Sivarajan first came to the US from India, in 2003, she worked for a tech company in Phoenix. After she married, she commuted each weekend to the San Francisco area, where her husband worked as an engineer. When she became pregnant with her son in 2007, she moved to California, giving up her job—and work permit. For three consecutive years, she applied for the same work visa that her husband holds, an H-1B. Each year, she was not picked in the random lottery that allocates these visas. She became depressed.
“There would be days when I would be so dull that I wouldn’t even want to play with my child, my own child,” she says. Sivarajan decided to go back to school, completing courses in early childhood education at a local community college. In 2015, after a years-long push from activists, the Obama administration allowed spouses of certain H-1B holders to obtain work permits. Sivarajan opened her daycare. More recently, she thought about expanding outside her home, to a new center. Then, in December, the Trump administration indicated it may eliminate the work permits for spouses.
“There are days when I can’t sleep properly because it bothers me,” Sivarajan says. “It bothers me to think about the future.” Without a job, she worries about holding onto the house, paying bills. She doesn’t know if finances will force the family to move back to India.
“Legally, I’m not allowed to work if I don’t have a work permit, which means all the 16 children whose families who are dependent on me right now, I have to let them all go. They have to find another provider for themselves. My three employees will lose their jobs,” she says. “Thinking about it makes me sad.”
The Trump administration says it plans to soon end the program allowing Sivarajan and more than 100,000 others to work in the US. Called H-4 EAD, or employment authorization document, the permit is available to the spouses of workers on H-1B visas who are in line for permanent US residency. Many tech companies sponsor and apply for H-1B visas, which are given to high-skilled foreign workers, often engineers.
The Obama administration started the program in 2015 partly due to a backlog in the green card process. Because of a per-country cap, people from populous countries such as India and China must wait years before gaining residency. That also meant spouses had been waiting years before they were eligible to work.
When it established the rule, US Citizenship and Immigration Services said the program would benefit the American economy. Last fall, USCIS indicated it was considering revoking the rule as part of President Trump’s “Buy American, Hire American” executive order.
Shah Peerally, an immigration attorney in Newark, California thinks the government will end the program. “I hope I am wrong, but I think it’s on the way,” he said. The government is still facing a 2015 lawsuit from a group alleging the program is illegal and takes jobs from US citizens. (The group is represented by attorneys from two organizations that the Southern Poverty Law Center classifies as hate groups.)
The Trump administration initially suggested it wanted to end the H-4 EAD program by February, but recently delayed that plan, saying it needed to conduct a new economic analysis. It now hopes to issue a proposal in June. Peerally thinks the administration has already made up its mind. He predicts that unless a lawsuit bogs down the process, the program “will be gone, basically in the next one or two years.” USCIS said in a statement that it is undergoing a “thorough review of employment-based visa program” and said it has not made a final decision.
Meanwhile, in Congress, Senator Orrin Hatch, (R-Utah), introduced an immigration bill that would keep the H-4 EAD program in place. Tech companies have stayed largely quiet on the issue, but an industry group wrote to USCIS in support of the program.
In the last few months, holders of the work permit—a vast majority of whom are women, many with advanced degrees—have ramped up a social-media campaign. They’ve gone to congressional town hall meetings and lobbied members of Congress at their offices, encouraging them to pressure USCIS to halt the rule change and to protect the program through legislation such as Hatch’s. At a recent town hall meeting with US Representative Ro Khanna, (D-California), a staffer asked the several hundred in the crowd how many were affected by the H-4 visa issue. At least two-thirds of the auditorium stood up.
"Most of us are spouses who have not been able to work for so long," Sivarajan says. "And not all of us know how to advocate for ourselves."
In conversations, several visa holders in Silicon Valley and one in Georgia speak of uncertainty prompted by the potential repeal of the policy, and question the government’s logic. Because of these work permits, people have bought homes, and moved around the country. Children have been brought into the world. From a range of backgrounds and professions, the visa holders talk of the sacrifices they’ve made, their hopes for the future, and the dignity of work, its inseparability from identity.
Tanya Madan, 28, Mountain View, California
Madan is a recruiter for a staffing company in San Jose. Every day, she helps Americans find work as cashiers, baristas, and clerks. We’re not talking high-profile tech jobs. “Most of them are looking for jobs in retail, restaurants, like, you know, for instance, Starbucks, Macy’s, something like that,” she says.
Madan came to the US in 2015, then spent a year at home before her work authorization came through. “When I got my worker permit authorization, it’s like, you know, ‘I got my wings back,’” she says. “I will work to give back something to this country because they’ve given me this great opportunity to work,” she decided. So she volunteered for a year at nonprofit news company as a recruiter. Money was secondary. First came identity. But when she and her husband moved in 2017 from New York to California, mammon reared its ugly head. “God, this area is too expensive!” she realized. So she got a job.
Madan looks back, sadly, on the year when the government did not allow her to work. “The whole year I didn’t know what to do,” she says. “Since I was in India I have been working since I was 18. So sitting at home, devoting my life to kitchen? Household chores? That’s not what I ever dreamed of. I’m a free bird.”
"I have a plan, that probably 10 years from now I would have a separate organization where there would be no fees charged when it comes to looking for a job, or looking for a candidate. It would be a free service. For everybody. Irrespective of what country you’re coming from. So I want to do something. I don’t know. It’s just something in my mind. (Laughs.) But I want to do something. Where people have free access to do their job. I mean, of course, there are staffing companies that make money. I don’t want to do that."
Sampada Khanapurkar, 37, Cumming, Georgia
Khanapurkar came to the US in 2004. She got her second masters degree at Virginia Tech in analytical chemistry; her first, in India, was in organic chemistry. In Boston, she worked on an H-1B visa for almost six years as a scientist at a biotech firm involved in cancer research. After her son was born in 2013, she quit her job, which changed her visa status to an H-4. The couple moved to Georgia for its lower cost of living. Khanapurkar had a hard time finding a position as a scientist on an H-1B there. By 2016, she had had another child because the new H-4 work permit program meant the couple could again have two incomes. She switched careers. Now Khanapurkar works as a project manager for a workflow-solutions company whose clients do a lot of printing, such as the Boston Globe. Another client is the insurance company known as American Family.
"We are looking at options. Should we move to Canada, or should we move to New Zealand? Or any other country that’s accepting of us? Right now we just feel not accepted here. Like we are not wanted here. And we are, you know, contributing so much to the economy and being legal citizens, legal immigrants. In spite of that, it’s the addition of being looked down upon. It’s not a good feeling, you know? I mean, I’ve been here so long, I just thought, ‘These people are mine.’ And now people aren’t accepting me. It’s not a good feeling. I told my husband yesterday, if this is how we’re feeling and this is how we’re going to be feeling every single day of our lives, living in fear, never know when our visas will be revoked, never know when we’ll be accepted here legally, in spite of being legal, we might as well go to a place where people are accepting of us."
Teenu Sharma, 31, Milpitas, California
In India, Sharma worked in insurance. She thought when she came to the US in 2014 that she would be able to find an employer willing to sponsor her for an H-1B visa. She soon realized it was next to impossible without being in tech. (Most petitions for these types of visas are for workers in STEM fields.) Sharma sought to study, but didn’t have the money. When the work permits became available in 2015, she had to return to India for four months for a family crisis, and is still waiting to receive her H-4 EAD. Her dream is to open a restaurant.
"I was totally independent when I was in India. Now I am dependent on everything on my husband. Let’s say if I want to buy a gift for my husband on his birthday, on Valentine’s Day, I cannot. Because I don’t have my own bank account. I don’t have my own debit card, credit card. I don’t have anything. So for his gift, I have to ask for money. From him. He’s a wonderful guy, that’s not the problem. But it hurts my dignity. It hurts my independence. I have dreams too. I have skills. I want to realize it for my own sake, for my own career, for the economy as well. And we love America as much as we love our home country."
In the space of an hour on a recent evening, a couple dozen cars refilled their gas tanks at a Valero service station just off the Redwood Highway in Mill Valley, a northbound stop on the way into California’s Marin County. During that time, the only pumping station that sat mostly unused was the cobalt blue one supplying hydrogen fuel. The hydrogen pump received only three visitors: two of the 3,800 hydrogen-powered sedans on California’s roads, each looking for a quick fill-up, and one old station wagon that parked there for a few minutes. An attendant who’s worked at the Valero for three years says that’s a pretty busy day for the hydrogen pump, which usually fuels one car per hour.
That’s not much of a return on the roughly $100 million California has spent over the past several years to build fueling stations for hydrogen vehicles. Each of the 31 hydrogen pumps around the state cost at least $2.5 million and was heavily subsidized with funds from the public and from Toyota Motor Corp., Honda Motor Co., and other automakers. Demand, however, remains so low that even with subsidies, they aren’t busy enough to turn a profit. (A typical fill-up costs customers about $45, but that’s heavily subsidized, and most lessors cover fuel costs.)
At Governor Jerry Brown’s direction, the state is spending more than $2.5 billion in clean energy funds to accelerate sales of hydrogen and battery vehicles. That includes $900 million earmarked to complete 200 hydrogen stations and 250,000 charging stations by 2025. A larger hydrogen network will help make the market more sustainable, the thinking goes—part of a kitchen sink approach to reducing carbon emissions alongside electric cars. Brown’s office referred requests for comment to the California Energy Commission, which said in a statement that the governor aims to have 5 million zero-emission vehicles on state roads by 2030, and that hydrogen is a part of that calculus.
The question is whether the money would be better spent on charging stations and other support for the state’s 360,000 plug-in electric vehicles. “There are still some kinks to work out,” says Joe Gagliano, infrastructure development manager at the California Fuel Cell Partnership, a group of state agencies and hydrogen proponents. “Hydrogen is coming out of left field for most consumers.”
So far, California is the only place in North America where drivers can buy a fuel cell vehicle and have some confidence they can get it refueled. “Competition is pretty scarce right now,” says Shane Stephens, chief development officer of FirstElement Fuel Inc., which built 16 of California’s 31 stations with $28 million in state grants. Most of the existing single-pump stations can fill 50 tanks to 60 tanks a day. Sometimes there’s a wait at the pumps, but mostly the nozzles hang limp, devoid of cars to feed.
On the occasions when two or more cars line up to fill at the same time, the second car typically has to wait an extra few minutes after the first car is done, because gas pumped under high pressure begins to freeze a single pump’s nipple, condensing moisture from the air into ice. Several newer, two-pump stations have helped address this problem, but California still needs more trucks to deliver supplies of hydrogen for the stations. (The state is subsidizing the refills for eight years.)
For decades, hydrogen advocates have called the Earth’s most abundant element the future of clean energy. But the occasional pilot programs around the world and those fueling stations in California have consistently failed to find traction, typically because of high costs and limited usability. Meanwhile, plug-in electrics have leapt ahead, thanks to Elon Musk and swift declines in lithium ion battery costs. California’s network of public charging stations now tops 14,000, well beyond the 9,000 gas stations that power the state’s roughly 25 million fossil fuel vehicles.
Stephens says the gap is beginning to narrow, if slightly. With continued government support, he says, FirstElement will be profitable in a couple of years and won’t need subsidies a decade from now. “We’re seeing uptake accelerate,” he says. Gagliano, who drives a hydrogen car himself, says that once people master the learning curve, they’ll appreciate hydrogen cars’ longer range (400 miles, compared with 300 for a typical Tesla) and shorter fill-up time (about four minutes, vs. at least a half-hour for the fastest battery-charging stations). “Filling up your first time is a little intimidating, but after that it’s no different than a gas pump,” Gagliano says.
Jim Collins, an investor who commutes to San Francisco from Tiburon in a hydrogen-powered Honda Clarity, says there are enough stations in the Bay Area to keep his car’s fuel cell running on the local trips he needs, but he wouldn’t have leased it if the state hadn’t put up $5,000 to lower the $400-a-month lease and kept subsidizing the fuel. Even brief disruptions of the refueling network, he says, can be a serious problem. When a station wasn’t working a week earlier, he switched to driving an old Saab. “This technology obviously isn’t ready for prime time,” he says.
California’s hydrogen power advocates say the next phase of the state’s fuel pump development will remain an early step along a path that’s likely to take decades. “We understand the skeptics,” Gagliano says. “What we’re doing now is just the tip of the spear. This is about getting to zero emissions by 2050.” But there’s little to suggest that consumers will be satisfied with similar standards 30 years from now, says Claire Curry, a technology analyst for Bloomberg New Energy Finance. “I’m skeptical,” she says. “I’d be surprised if we aren’t all getting around in a battery electric by then.”
BOTTOM LINE – Governor Brown has earmarked millions more dollars of state funds to subsidize hydrogen pump installations and refills. That may not be the easiest way to cut carbon emissions.
Warren Buffett auctions a lunch date for charity every year, and the winning bid usually stretches to seven figures. He twice sold his used cars to fans for multiples of their Kelly Blue Book value. Someone once even paid more than $200,000 to purchase his old wallet. (It had a stock tip inside.) For those who venerate one of the world’s best investors, money is usually no object when buying a piece of the legend.
A year ago, Buffett put his vacation home in Emerald Bay, a gated enclave next to Laguna Beach, Calif., up for sale. He bought the property in 1971 at the urging of his first wife, Susan, for $150,000—the equivalent of a bit less than $1 million today. At the time, he didn’t think of it much as an investment, he told the last year. Laguna was less developed back then, more surfer-and-hippie paradise than multimillionaire’s haunt. The couple and their family often spent summers at the home, as well as time around Christmas, when Buffett would hole up in the master bedroom working on his closely followed annual letter to Berkshire Hathaway Inc. shareholders.
Over the years, Buffett upgraded the home, which was built in 1936. He even bought an adjacent property for guests to sleep in. When Susan died in 2004, Buffett—who lives most of the year in Omaha, Neb.—stopped going to the house as much, which is why he finally decided to list it. His asking price: $11 million.
“For the first time in nearly 50 years the legendary ‘Oracle of Omaha’s’ home 27 Emerald Bay is now available!” the listing began, before going on to describe the almost-3,600-square-foot property’s six bedrooms and sea views. Staged photos of the interior played up the association with the billionaire investor. A living room shot, for instance, is decorated with a life-size cutout of Mary See, whose face adorns boxes of chocolates sold by Berkshire’s See’s Candies. In another room, an open bottle of Coca-Cola and a folded rest on a side table in front of a TV tuned to CNBC. (Berkshire is Coca-Cola Co.’s biggest shareholder, and Buffett is a frequent guest on the business news channel.)
The finishes and amenities appear to be relatively modest. One bathroom is wallpapered in old covers, and most of the kitchen counters are white laminate. The garage fits just one car, and the ocean view is partially obscured by other homes.
After the initial buzz and several write-ups in the press, interest in the property seems to have fizzled. It’s now been on the market for about five months longer than the median listing time for similarly priced homes in the same ZIP code, according to data compiled by Redfin. Bill Dolby, the listing agent, and Buffett didn’t respond to requests for comment about the home or how they decided on a price.
At the high end of the market, sales can be idiosyncratic—there’s a limited number of potential buyers, and they can be demanding. Still, other local real estate agents who work with high-end clients say it’s clear why the property hasn’t sold yet. “No one’s going to rehab a house from 1936,” says Eliisa Stowell, a realtor with Surterre Properties in nearby Corona del Mar. “It might have sentimental value to someone, but then it’s gone” when the home gets torn down to build something more modern, she says. Bill Cote, a Newport Beach-based realtor with Coldwell Banker, is more blunt: “It’s dreadfully overpriced,” he says. The Buffett name, he adds, probably doesn’t do much for buyers in the area. “If you said it was Bette Midler’s house, it might have some cachet,” Cote says.
Then again, buyers who do know Buffett might be nervous about being on the other side of a negotiation with him. One secret to Buffett’s success as a businessman and investor has been his unwillingness to bend on price. Whether he’s looking at a stock or a whole company, the billionaire has a strong idea for what he thinks things are worth and isn’t afraid to walk away if his terms aren’t met. Having more money than he’ll ever need means he’s never a forced buyer or seller. As Cote says, “It’s hard to get into a conversation with a seller who’s got more money than God.”
BOTTOM LINE – Buyers don’t seem interested in Buffett’s beach house, built in 1936. It may be overpriced—and the seller isn’t exactly desperate.
Granted, some tech is better than other tech. No one needs a Wi-Fi-connected juice press that doesn’t actually juice anything. Gadgets that offer real utility—like a smart oven or open source furniture—stand a better chance of becoming ubiquitous. If you’re skeptical, think of it this way: In-home refrigeration was the crazy, newfangled invention of 1913. Now, few among us can imagine living without it.
What will the home of the future look like? We took stock of the most exciting tech-forward home products on the market. It’s only a matter of time until at least some of these come standard in every American home.
The High-Tech Living Room
Thirty-nine million Americans now have a smart speaker in their homes—that’s 1 in 6 people—and all signs indicate this figure will only creep higher with time. In the living room of the future, smart speakers will be a central feature, with newer models connected to every element in your home, from the lightbulbs to the lock on your front door to the thermostat. They will become so essential you won’t think twice about plunking down $400 for one.
Watching TV and movies will be a wildly different experience. Why devote precious square footage in your living room to a giant screen when you could have one that effortlessly rolls up away and out of sight, like the one LG Display debuted at this year’s CES? Or you may choose not to have a TV at all and opt instead for a superhigh-resolution short-throw projector that turns any white wall into your own personal movie theater. Sony’s new $30,000 model would fit the bill, assuming the price tag comes down.
In the coming years, it’ll be much easier to design your living space. Apps and online platforms such as Modsy and Hutch will use virtual and augmented reality to help you visualize how a couch or chair will look in your home. You’ll have lots of options: Modular, open source furniture will dominate interior design trends, taking the lead from Ikea’s Tom Dixon-designed Delaktig couch, which has more than 97 different configurations. Choose wisely, because you’ll be spending more time on the couch than ever: Facebook Inc.’s forthcoming living-room-geared video chat device will reportedly use smart camera technology to make people on both ends feel like they’re sitting in the same room.
Also, expect your living room to be even more of a central hub than it already is. Deliveries will arrive here instead of on your front porch, thanks to Amazon.com’s new Prime service, which will let verified delivery persons carry goods right into your home.
Ultimately, the goal of kitchen technology won’t be to do the cooking for you. It’ll just make you a better cook. Smart ovens such as those from June will be outfitted with cameras and digital thermometers, helping you monitor your food as it bakes. And instead of just hoping the “medium-hot” setting on your gas range is hot enough, smart skillets will take guessing out of the equation by sizzling food at a precise temperature, which you’ll set on a connected app.
Once you’re up and moving, it’s time to get dressed: Your closet will be filled with clothes you don’t just wear. They will actually interact with you, tracking health markers and habits. Among them: MadeWithGlove’s still-in-development smart gloves, which promise to detect skin temperature and provide heat accordingly. Your clothes might even change shape or color based on your feelings, as will the Sensoree mood sweater, now available for preorder.
And if you want a new wardrobe, you won’t have to even leave the house to find the best-fitting clothes: Amazon’s patented mirror will let you virtually try on outfits from the comfort of your own bedroom.
Sound far-fetched? Remember a decade ago, few of us could have imagined being so attached to our smartphones, let alone ordering groceries off the internet or barking commands at a digital assistant. With time, even the strangest things can become normal.
When the masses descend upon snow-covered Davos for the World Economic Forum’s annual gathering, Esther Heldstab is determined to swim against the tide.
The purveyor of tourist mementos on the town’s main street is one of the few shopkeepers who won’t hand over their keys to multinational conglomerates who transform shoe stores and bakeries into lounges and cocktail party space for the week.
“On this mile, you can pretty much ask what you want,” said Heldstab. No one interviewed for this story would speak on the record about rental costs, though Heldstab said she could double her money if she rented out her shop in a half-timber house festooned with Swiss crosses. Still, “it’s so nice when the people from all over the world come — it’s like Hollywood.”
This year actress Cate Blanchett, singer Elton John and Bollywood icon Shah Rukh Khan kick off the WEF with the Crystal Award ceremony on the evening of Jan. 22. They join the ranks of Angelina Jolie, Charlize Theron and Matt Damon. John Kerry visited Heldstab’s shop when he attending as U.S. Secretary of State, she said over lunch earlier this month.
Others local businesses have opted for the money instead. Schneider’s bakery and cafe is being rented out to the Indian delegation. Furniture is shipped in and the chocolatier’s 95-person staff takes care of the catering and service, said owner Urs Wipraechtiger. Right next door is Russia House. Now in its fourth year, it has this year decamped to this larger space close to the conference center, according to organizer Roscongress. Energy Minister Alexander Novak will speak at an event.
Just a few steps away, M&G Investments will be hosting an evening of live jazz, drinks and canapes, while ConsenSys, a Brooklyn company that develops blockchain applications, is setting up an “Ethereal Lounge.”
Making its debut on the strip is Ukraine House, which is taking over the Timberland shoe store. The location has been booked since last summer, and both Ukraine’s President Petro Poroshenko and former world boxing champion Vitaliy Klitschko, now mayor of Kiev, will make appearances with the hope of enticing investors.
“We realized that there’s this tradition of having the country houses in Davos,” Chopivsky said. He wants to showcase cuisine and hired a Ukrainian catering company to make borscht and golubsti (meat and rice-filled cabbage rolls). Wine, cognac, vodka and chocolate from Lviv will be flown in.
While the University of Chicago is holding a reception with Microsoft CEO and alum Satya Nadella at the five-star Steigenberger Grandhotel Belvedere, business school INSEAD has opted for the Kirchner Museum, close to the conference center.
Each year the gallery installs a kitchen in its basement, and the library and meeting space are transformed into sitting rooms with sofas and lamps.
“These events are a big source of income,” said Dolores Mark, who heads the private museum’s administration and declined to comment on the rental fee.
And while red wine can’t be served for risk of damaging the works of the German expressionist to whom the museum is dedicated, that’s not putting off potential tenants. “Many book immediately after the event for the next year,” said Mark. “One or two companies have three-year contracts.”
With ad-hoc sites sprouting up like mushrooms, to locals Davos seems utterly transformed during the week of the forum.
“If a WEF participant came back during the summertime, he or she wouldn’t recognize the town,” said Kirchner’s Mark.
In early 2014, Srikanth Thirumalai met with Amazon CEO Jeff Bezos. Thirumalai, a computer scientist who’d left IBM in 2005 to head Amazon’s recommendations team, had come to propose a sweeping new plan for incorporating the latest advances in artificial intelligence into his division.
He arrived armed with a “six-pager.” Bezos had long ago decreed that products and services proposed to him must be limited to that length, and include a speculative press release describing the finished product, service, or initiative. Now Bezos was leaning on his deputies to transform the company into an AI powerhouse. Amazon’s product recommendations had been infused with AI since the company’s very early days, as had areas as disparate as its shipping schedules and the robots zipping around its warehouses. But in recent years, there has been a revolution in the field; machine learning has become much more effective, especially in a supercharged form known as deep learning. It has led to dramatic gains in computer vision, speech, and natural language processing.
In the early part of this decade, Amazon had yet to significantly tap these advances, but it recognized the need was urgent. This era’s most critical competition would be in AI—Google, Facebook, Apple, and Microsoft were betting their companies on it—and Amazon was falling behind. “We went out to every [team] leader, to basically say, ‘How can you use these techniques and embed them into your own businesses?’” says David Limp, Amazon’s VP of devices and services.
Thirumalai took that to heart, and came to Bezos for his annual planning meeting with ideas on how to be more aggressive in machine learning. But he felt it might be too risky to wholly rebuild the existing system, fine-tuned over 20 years, with machine-learning techniques that worked best in the unrelated domains of image and voice recognition. “No one had really applied deep learning to the recommendations problem and blown us away with amazingly better results,” he says. “So it required a leap of faith on our part.” Thirumalai wasn’t quite ready—but Bezos wanted more. So Thirumalai shared his edgier option of using deep learning to revamp the way recommendations worked. It would require skills that his team didn’t possess, tools that hadn’t been created, and algorithms that no one had thought of yet. Bezos loved it (though it isn’t clear whether he greeted it with his trademark hyena-esque laugh), so Thirumalai rewrote his press release and went to work.
Thirumalai was only one of a procession of company leaders who trekked to Bezos a few years ago with six-pagers in hand. The ideas they proposed involved completely different products with different sets of customers. But each essentially envisioned a variation of Thirumalai’s approach: transforming part of Amazon with advanced machine learning. Some of them involved rethinking current projects, like the company’s robotics efforts and its huge data-center business, Amazon Web Services (AWS). Others would create entirely new businesses, like a voice-based home appliance that would become the Echo.
The results have had an impact far beyond the individual projects. Thirumalai says that at the time of his meeting, Amazon’s AI talent was segregated into isolated pockets. “We would talk, we would have conversations, but we wouldn’t share a lot of artifacts with each other because the lessons were not easily or directly transferable,” he says. They were AI islands in a vast engineering ocean. The push to overhaul the company with machine learning changed that.
While each of those six-pagers hewed to Amazon’s religion of “single-threaded” teams—meaning that only one group “owns” the technology it uses—people started to collaborate across projects. In-house scientists took on hard problems and shared their solutions with other groups. Across the company, AI islands became connected. As Amazon's ambition for its AI projects grew, the complexity of its challenges became a magnet for top talent, especially those who wanted to see the immediate impact of their work. This compensated for Amazon's aversion to conducting pure research; the company culture demanded that innovations come solely in the context of serving its customers.
Amazon loves to use the word flywheel to describe how various parts of its massive business work as a single perpetual motion machine. It now has a powerful AI flywheel, where machine-learning innovations in one part of the company fuel the efforts of other teams, who in turn can build products or offer services to affect other groups, or even the company at large. Offering its machine-learning platforms to outsiders as a paid service makes the effort itself profitable—and in certain cases scoops up yet more data to level up the technology even more.
It took a lot of six-pagers to transform Amazon from a deep-learning wannabe into a formidable power. The results of this transformation can be seen throughout the company—including in a recommendations system that now runs on a totally new machine-learning infrastructure. Amazon is smarter in suggesting what you should read next, what items you should add to your shopping list, and what movie you might want to watch tonight. And this year Thirumalai started a new job, heading Amazon search, where he intends to use deep learning in every aspect of the service.
“If you asked me seven or eight years ago how big a force Amazon was in AI, I would have said, ‘They aren’t,’” says Pedro Domingos, a top computer science professor at the University of Washington. “But they have really come on aggressively. Now they are becoming a force.”
Maybe the force.
The Alexa Effect
The flagship product of Amazon’s push into AI is its breakaway smart speaker, the Echo, and the Alexa voice platform that powers it. These projects also sprang from a six-pager, delivered to Bezos in 2011 for an annual planning process called Operational Plan One. One person involved was an executive named Al Lindsay, an Amazonian since 2004, who had been asked to move from his post heading the Prime tech team to help with something totally new. “A low-cost, ubiquitous computer with all its brains in the cloud that you could interact with over voice—you speak to it, it speaks to you,” is how he recalls the vision being described to him.
But building that system—literally an attempt to realize a piece of science fiction, the chatty computer from Star Trek—required a level of artificial intelligence prowess that the company did not have on hand. Worse, of the very few experts who could build such a system, even fewer wanted to work for Amazon. Google and Facebook were snapping up the top talent in the field. “We were the underdog,” Lindsay, who is now a VP, says.
“Amazon had a bit of a bad image, not friendly to people who were research oriented,” says Domingos, the University of Washington professor. The company’s relentless focus on the customer, and its culture of scrappiness, did not jibe with the pace of academia or cushy perks of competitors. “At Google you’re pampered,” Domingos says. “At Amazon you set up your computer from parts in the closet.” Worse, Amazon had a reputation as a place where innovative work was kept under corporate wraps. In 2014, one of the top machine-learning specialists, Yann LeCun, gave a guest lecture to Amazon’s scientists in an internal gathering. Between the time he was invited and the event itself, LeCun accepted a job to lead Facebook’s research effort, but he came anyway. As he describes it now, he gave his talk in an auditorium of about 600 people and then was ushered into a conference room where small groups came in one by one and posed questions to him. But when he asked questions of them, they were unresponsive. This turned off LeCun, who had chosen Facebook in part because it agreed to open-source much of the work of its AI team.
Because Amazon didn’t have the talent in-house, it used its deep pockets to buy companies with expertise. “In the early days of Alexa, we bought many companies,” Limp says. In September 2011, it snapped up Yap, a speech-to-text company with expertise in translating the spoken word into written language. In January 2012, Amazon bought Evi, a Cambridge, UK, AI company whose software could respond to spoken requests like Siri does. And in January 2013, it bought Ivona, a Polish company specializing in text-to-speech, which provided technology that enabled Echo to talk.
But Amazon’s culture of secrecy hampered its efforts to attract top talent from academia. One potential recruit was Alex Smola, a superstar in the field who had worked at Yahoo and Google. “He is literally one of the godfathers of deep learning,” says Matt Wood, the general manager of deep learning and AI at Amazon Web Services. (Google Scholar lists more than 90,000 citations of Smola's work.) Amazon execs wouldn’t even reveal to him or other candidates what they would be working on. Smola rejected the offer, choosing instead to head a lab at Carnegie Mellon.
“Even until right before we launched there was a headwind,” Lindsay says. “They would say, ‘Why would I want to work at Amazon—I’m not interested in selling people products!’”
Amazon did have one thing going for it. Since the company works backward from an imagined final product (thus the fanciful press releases), the blueprints can include features that haven’t been invented yet. Such hard problems are irresistible to ambitious scientists. The voice effort in particular demanded a level of conversational AI—nailing the “wake word” (“Hey Alexa!”), hearing and interpreting commands, delivering non-absurd answers—that did not exist.
That project, even without the specifics on what Amazon was building, helped attract Rohit Prasad, a respected speech-recognition scientist at Boston-based tech contractor Raytheon BBN. (It helped that Amazon let him build a team in his hometown.) He saw Amazon’s lack of expertise as a feature, not a bug. “It was green fields here,” he says. “Google and Microsoft had been working on speech for years. At Amazon we could build from scratch and solve hard problems.” As soon as he joined in 2013, he was sent to the Alexa project. “The device existed in terms of the hardware, but it was very early in speech,” he says.
The trickiest part of the Echo—the problem that forced Amazon to break new ground and in the process lift its machine-learning game in general—was something called far field speech recognition. It involves interpreting voice commands spoken some distance from the microphones, even when they are polluted with ambient noise or other aural detritus. One challenging factor was that the device couldn’t waste any time cogitating about what you said. It had to send the audio to the cloud and produce an answer quickly enough that it felt like a conversation, and not like those awkward moments when you’re not sure if the person you’re talking to is still breathing. Building a machine-learning system that could understand and respond to conversational queries in noisy conditions required massive amounts of data—lots of examples of the kinds of interactions people would have with their Echos. It wasn’t obvious where Amazon might get such data.
Far-field technology had been done before, says Limp, the VP of devices and services. But “it was on the nose cone of Trident submarines, and it cost a billion dollars.” Amazon was trying to implement it in a device that would sit on a kitchen counter, and it had to be cheap enough for consumers to spring for a weird new gadget. “Nine out of 10 people on my team thought it couldn’t be done,” Prasad says. “We had a technology advisory committee of luminaries outside Amazon—we didn’t tell them what we were working on, but they said, ‘Whatever you do, don’t work on far field recognition!’”
Prasad’s experience gave him confidence that it could be done. But Amazon did not have an industrial-strength system in place for applying machine learning to product development. “We had a few scientists looking at deep learning, but we didn’t have the infrastructure that could make it production-ready,” he says. The good news was that all the pieces were there at Amazon—an unparalleled cloud service, data centers loaded with GPUs to crunch machine-learning algorithms, and engineers who knew how to move data around like fireballs.
His team used those parts to create a platform that was itself a valuable asset, beyond its use in fulfilling the Echo’s mission. “Once we developed Echo as a far-field speech recognition device, we saw the opportunity to do something bigger—we could expand the scope of Alexa to a voice service,” says Alexa senior principal scientist Spyros Matsoukas, who had worked with Prasad at Raytheon BBN. (His work there had included a little-known Darpa project called Hub4, which used broadcast news shows and intercepted phone conversations to advance voice recognition and natural language understanding—great training for the Alexa project.) One immediate way they extended Alexa was to allow third-party developers to create their own voice-technology mini-applications—dubbed “skills”—to run on the Echo itself. But that was only the beginning.
By breaking out Alexa beyond the Echo, the company’s AI culture started to coalesce. Teams across the company began to realize that Alexa could be a useful voice service for their pet projects too. “So all that data and technology comes together, even though we are very big on single-threaded ownership,” Prasad says. First other Amazon products began integrating into Alexa: When you speak into your Alexa device you can access Amazon Music, Prime Video, your personal recommendations from the main shopping website, and other services. Then the technology began hopscotching through other Amazon domains. “Once we had the foundational speech capacity, we were able to bring it to non-Alexa products like Fire TV, voice shopping, the Dash wand for Amazon fresh, and, ultimately, AWS,” Lindsay says.
The AI islands within Amazon were drawing closer.
Another pivotal piece of the company’s transformation clicked into place once millions of customers (Amazon won’t say exactly how many) began using the Echo and the family of other Alexa-powered devices. Amazon started amassing a wealth of data—quite possibly the biggest collection of interactions of any conversation-driven device ever. That data became a powerful lure for potential hires. Suddenly, Amazon rocketed up the list of places where those coveted machine-learning experts might want to work. “One of the things that made Alexa so attractive to me is that once you have a device in the market, you have the resource of feedback. Not only the customer feedback, but the actual data that is so fundamental to improving everything—especially the underlying platform,” says Ravi Jain, an Alexa VP of machine learning who joined the company last year.
So as more people used Alexa, Amazon got information that not only made that system perform better but supercharged its own machine-learning tools and platforms—and made the company a hotter destination for machine-learning scientists.
The flywheel was starting to spin.
A Brainier Cloud
Amazon began selling Echo to Prime customers in 2014. That was also the year that Swami Sivasubramanian became fascinated with machine learning. Sivasubramanian, who was managing the AWS database and analytics business at the time, was on a family trip to India, when due to a combination of jet lag and a cranky infant daughter, he found himself at his computer late into the night fiddling with tools like Google’s Tensorflow and Caffé, which is the machine-learning framework favored by Facebook and many in the academic community. He concluded that combining these tools with Amazon’s cloud service could yield tremendous value. By making it easy to run machine-learning algorithms in the cloud, he thought, the company might tap into a vein of latent demand. “We cater to millions of developers every month,” he says. “The majority are not professors at MIT but developers who have no background in machine learning.”
At his next Jeff Bezos review he came armed with an epic six-pager. On one level, it was a blueprint for adding machine-learning services to AWS. But Sivasubramanian saw it as something broader: a grand vision of how AWS could become the throbbing center of machine-learning activity throughout all of techdom.
In a sense, offering machine learning to the tens of thousands of Amazon cloud customers was inevitable. “When we first put together the original business plan for AWS, the mission was to take technology that was only in reach of a small number of well-funded organizations and make it as broadly distributed as possible,” says Wood, the AWS machine-learning manager. “We’ve done that successfully with computing, storage, analytics, and databases—and we’re taking the exact same approach with machine learning.” What made it easier was that the AWS team could draw on the experience that the rest of the company was accumulating.
AWS’s Amazon Machine Learning, first offered in 2015, allows customers like C-Span to set up a private catalog of faces, Wood says. Zillow uses it to estimate house prices. Pinterest employs it for visual search. And several autonomous driving startups are using AWS machine learning to improve products via millions of miles of simulated road testing.
In 2016, AWS released new machine-learning services that more directly drew on the innovations from Alexa—a text-to-speech component called Polly and a natural language processing engine called Lex. These offerings allowed AWS customers, which span from giants like Pinterest and Netflix to tiny startups, to build their own mini Alexas. A third service involving vision, Rekognition, drew on work that had been done in Prime Photos, a relatively obscure group at Amazon that was trying to perform the same deep-learning wizardry found in photo products by Google, Facebook, and Apple.
These machine-learning services are both a powerful revenue generator and key to Amazon’s AI flywheel, as customers as disparate as NASA and the NFL are paying to get their machine learning from Amazon. As companies build their vital machine-learning tools inside AWS, the likelihood that they will move to competing cloud operations becomes ridiculously remote. (Sorry, Google, Microsoft, or IBM.) Consider Infor, a multibillion-dollar company that creates business applications for corporate customers. It recently released an extensive new application called Coleman (named after the NASA mathematician in Hidden Figures) that allows its customers to automate various processes, analyze performance, and interact with data all through a conversational interface. Instead of building its own bot from scratch, it uses AWS’s Lex technology. “Amazon is doing it anyway, so why would we spend time on that? We know our customers and we can make it applicable to them,” says Massimo Capoccia, a senior VP of Infor.
AWS’s dominant role in the ether also gives it a strategic advantage over competitors, notably Google, which had hoped to use its machine-learning leadership to catch up with AWS in cloud computing. Yes, Google may offer customers super-fast, machine-learning-optimized chips on its servers. But companies on AWS can more easily interact with—and sell to—firms that are also on the service. “It’s like Willie Sutton saying he robs banks because that’s where the money is,” says DigitalGlobe CTO Walter Scott about why his firm uses Amazon’s technology. “We use AWS for machine learning because that’s where our customers are.”
Last November at the AWS re:Invent conference, Amazon unveiled a more comprehensive machine-learning prosthetic for its customers: SageMaker, a sophisticated but super easy-to-use platform. One of its creators is none other than Alex Smola, the machine-learning superstar with 90,000 academic citations who spurned Amazon five years ago. When Smola decided to return to industry, he wanted to help create powerful tools that would make machine learning accessible to everyday software developers. So he went to the place where he felt he’d make the biggest impact. “Amazon was just too good to pass up,” he says. “You can write a paper about something, but if you don’t build it, nobody will use your beautiful algorithm,” he says.
When Smola told Sivasubramanian that building tools to spread machine learning to millions of people was more important than publishing one more paper, he got a nice surprise. “You can publish your paper, too!” Sivasubramanian said. Yes, Amazon is now more liberal in permitting its scientists to publish. “It’s helped quite a bit with recruiting top talent as well as providing visibility of what type of research is happening at Amazon,” says Spyros Matsoukas, who helped set guidelines for a more open stance.
It’s too early to know if the bulk of AWS’s million-plus customers will begin using SageMaker to build machine learning into their products. But every customer that does will find itself heavily invested in Amazon as its machine-learning provider. In addition, the platform is sufficiently sophisticated that even AI groups within Amazon, including the Alexa team, say they intend to become SageMaker customers, using the same toolset offered to outsiders. They believe it will save them a lot of work by setting a foundation for their projects, freeing them to concentrate on the fancier algorithmic tasks.
Even if only some of AWS’s customers use SageMaker, Amazon will find itself with an abundance of data about how its systems perform (excluding, of course, confidential information that customers keep to themselves). Which will lead to better algorithms. And better platforms. And more customers. The flywheel is working overtime.
With its machine learning overhaul in place, the company’s AI expertise is now distributed across its many teams—much to the satisfaction of Bezos and his consiglieri. While there is no central office of AI at Amazon, there is a unit dedicated to the spread and support of machine learning, as well as some applied research to push new science into the company’s projects. The Core Machine Learning Group is led by Ralf Herbrich, who worked on the Bing team at Microsoft and then served a year at Facebook, before Amazon lured him in 2012. “It’s important that there’s a place that owns this community” within the company, he says. (Naturally, the mission of the team was outlined in an aspirational six-pager approved by Bezos.)
Part of his duties include nurturing Amazon’s fast-growing machine-learning culture. Because of the company’s customer-centric approach—solving problems rather than doing blue-sky research—Amazon execs do concede that their recruiting efforts will always tilt towards those interested in building things rather than those chasing scientific breakthroughs. Facebook’s LeCun puts it another way: “You can do quite well by not leading the intellectual vanguard.”
But Amazon is following Facebook and Google’s lead in training its workforce to become adept at AI. It runs internal courses on machine-learning tactics. It hosts a series of talks from its in-house experts. And starting in 2013, the company has hosted an internal machine-learning conference at its headquarters every spring, a kind of Amazon-only version of NIPS, the premier academic machine-learning-palooza. “When I started, the Amazon machine-learning conference was just a couple hundred people; now it’s in the thousands,” Herbrich says. “We don’t have the capacity in the largest meeting room in Seattle, so we hold it there and stream it to six other meeting rooms on the campus.” One Amazon exec remarks that if it gets any bigger, instead of calling it an Amazon machine-learning event, it should just be called Amazon.
Herbrich’s group continues to push machine learning into everything the company attempts. For example, the fulfillment teams wanted to better predict which of the eight possible box sizes it should use with a customer order, so they turned to Herbrich’s team for help. “That group doesn’t need its own science team, but it needed these algorithms and needed to be able to use them easily,” he says. In another example, David Limp points to a transformation in how Amazon predicts how many customers might buy a new product. “I’ve been in consumer electronics for 30 years now, and for 25 of those forecasting was done with [human] judgment, a spreadsheet, and some Velcro balls and darts,” he says. “Our error rates are significantly down since we’ve started using machine learning in our forecasts.”
Still, sometimes Herbrich’s team will apply cutting-edge science to a problem. Amazon Fresh, the company’s grocery delivery service, has been operating for a decade, but it needed a better way to assess the quality of fruits and vegetables—humans were too slow and inconsistent. His Berlin-based team built sensor-laden hardware and new algorithms that compensated for the inability of the system to touch and smell the food. “After three years, we have a prototype phase, where we can judge the quality more reliably” than before, he says.
Of course, such advances can then percolate throughout the Amazon ecosystem. Take Amazon Go, the deep-learning-powered cashier-less grocery store in its headquarters building that recently opened to the public. “As a customer of AWS, we benefit from its scale,” says Dilip Kumar, VP of Technology for Amazon Go. “But AWS is also a beneficiary.” He cites as an example Amazon Go’s unique system of streaming data from hundreds of cameras to track the shopping activities of customers. The innovations his team concocted helped influence an AWS service called Kinesis, which allows customers to stream video from multiple devices to the Amazon cloud, where they can process it, analyze it, and use it to further advance their machine learning efforts.
Even when an Amazon service doesn’t yet use the company’s machine-learning platform, it can be an active participant in the process. Amazon’s Prime Air drone-delivery service, still in the prototype phase, has to build its AI separately because its autonomous drones can’t count on cloud connectivity. But it still benefits hugely from the flywheel, both in drawing on knowledge from the rest of the company and figuring out what tools to use. “We think about this as a menu—everybody is sharing what dishes they have,” says Gur Kimchi, VP of Prime Air. He anticipates that his team will eventually have tasty menu offerings of its own. “The lessons we’re learning and problems we’re solving in Prime Air are definitely of interest to other parts of Amazon,” he says.
In fact, it already seems to be happening. “If somebody’s looking at an image in one part of the company, like Prime Air or Amazon Go, and they learn something and create an algorithm, they talk about it with other people in the company,” says Beth Marcus, a principal scientist at Amazon robotics. “And so someone in my team could use it to, say, figure out what’s in an image of a product moving through the fulfillment center.”
Is it possible for a company with a product-centered approach to eclipse the efforts of competitors staffed with the superstars of deep learning? Amazon’s making a case for it. “Despite the fact they’re playing catchup, their product releases have been incredibly impressive,” says Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence. “They’re a world-class company and they’ve created world-class AI products.”
The flywheel keeps spinning, and we haven’t seen the impact of a lot of six-pager proposals still in the pipeline. More data. More customers. Better platforms. More talent.
In Daniele Abate’s Sicilian home town, many people don’t even have running water, and he blames the politicians. So the former cook will be voting for Five Star on March 4.
At the other end of the country, across the economic divide that runs through Italy, a third of small company owners in Vicenza plan to do the same, according to Luigino Bari, who runs a local business association. They want tax cuts and deregulation, he says.
As an uncertain country gears up for a crucial election, the anti-establishment Five Star Movement is demonstrating a rare ability to appeal to disaffected voters across geography and social strata. Its eclectic mix of environmentalism, euro-skepticism and widely questioned promises on taxes and benefits offers something for anyone with an ax to grind about the way Italy has been run.
“It’s a catch-all party,” said Piergiorgio Corbetta, a political science professor at the University of Bologna. “There are many reasons to vote for Five Star.”
With four weeks to go, polls show Five Star may have provided enough reasons to secure one of the biggest victories yet for populists in western Europe. With an outright majority still a distant prospect and few natural allies in parliament, the party is still likely to be kept out of office by an alliance of establishment groups. But their success highlights the challenge facing the next administration.
“Whatever color of government Italy ends up with, they will weigh heavily on the debate,” said Marc Lazar, a professor at Sciences Po in Paris. “When you take almost 30 percent of the vote, you are a reality that must be dealt with.”
Since starting as an internet-based campaign group in 2009, Five Star’s rise has been driven by support in places like Abate’s home region of Trapani, which was found to have the lowest quality of life among Italy’s 110 provinces by La Sapienza University last year.
Abate has been living off a 280-euro ($350) disability pension each month since his knee gave out a few years ago, forcing him to give up kitchen work. He’s 53, but looks older and struggles to stand. For Abate, the appeal of Five Star is its pledge to take on the privileges of lawmakers and civil servants in Rome.
“We work for many years and barely get a thing,” he said, sitting in the main square of his hometown of Alcamo near a 17th century church. “They serve for a few months and can retire.’’
The key to electoral success for Five Star leader Luigi Di Maio will be pushing into Italy’s wealthier north. While the party won 40 percent of the vote in Trapani in the last national elections 2013, it got 25 percent in the manufacturing center of Vicenza near Venice.
Vicenza’s entrepreneurs are also frustrated with the status quo, regardless of the recent pickup in growth. They are demanding cuts to business taxes and regulations, and investment in the single-lane roads crowded with trucks carrying products from the region’s factories.
“It’s clear that the traditional parties have made promises that they haven’t kept,” said Bari, 64, who wouldn’t say who he’ll be voting for.
Just down the road, the 7,000 inhabitants of Sarego elected the first Five Star mayor in the northeastern Italy in 2012. Roberto Castiglion, a 37-year-old IT manager, was re-elected last year with an increased vote.
Most of Castiglion’s work as mayor has involved the environment, installing solar panels and increasing recycling, but he says the party is very keen to help local businesses which ship factory machinery, adult diapers and leather goods around the world.
“In this country, we are drowning in norms and regulations,” he said.
“Five Star is saying the right things to small businesses, but there is some hesitancy,” said Remigio Bisognin, the 63-year-old founder of a 14-employee Sarego firm that stamps plastic parts. “We don’t really know these people.’’
One source of concern for business leaders has been Five Star’s past threats to pull Italy out of the euro. Bisognin says mistakes were made introducing the single currency but it’s too late to go back now, and Di Maio has walked back his comments. It’s a move that broadens the party’s appeal in the north without hurting its base in the south.
“The euro is not something we worry about,” said Gaetano Milazzo, a 40-year-old tax collector as he talked to friends where the warren of narrow streets opens out into Alcamo’s square. “Some houses here get water one day a week and there’s hardly any public transport.”
Indeed, parts of the sprawling town of 45,000 aren’t even connected to the water mains and Domenico Surdi, the 34-year-old lawyer Five Star mayor since in 2016, says the existing pipes hadn’t been maintained for decades when he took office.
With no budget for repairs, Surdi has had to improvise. He’s aiming to raise the amount of garbage that’s recycled to 70 percent from about 60 percent to save about 1 million euros a year on trash hauling.
“We’ve been mismanaged for so long,” said Abate. “The problems won’t go away overnight.”