Inc. Uncensored: Here's What Happened When an Economist Walked Into a Brothel

First up is a segment with economist and author Allison Schrager, whose new book, An Economist Walks Into a Brothel, examines economic risk of all stripes. As a retirement planner Schrager works in a financial field largely averse to great risk, so she set out to study the other extreme. For example, why would sex workers in Nevada gladly forfeit 50 percent of their earnings for added security? Why do some criminals continue to repeat committing offenses after being caught? Schrager explains it all–including her fascinating journey to report on the riskiest economic choices.

Next, editor-at-large Kimberly Weisul explains that ever since raising her first fund in 2015, Arlan Hamilton has been unusually high-profile for a venture capitalist. She came onto the scene with almost missionary zeal, advocating for the potential of women, people of color, and the LGBTQ community as entrepreneurs. Recently however, she took a step back from leading her venture fund, Backstage Capital. We discuss how the firm’s recently reported $36 million round didn’t come together as expected.

Finally, I talk about a story from our Private Titans series, on Specialized Bicycles. I explain how surprising it was to speak to the company’s founder and chief executive, Mike Sinyard, a man who has infused his company with what I dubbed a “highly functional breed of late-stage hippie capitalism.” Specialized is now a 1,600-person, $1 billion global operation–and one run on a particular balance of intuition and analytics.

This week’s Like Buttons included:

Check out the full episode in the player below, in iTunes, or wherever you get your podcasts.

Published on: Apr 5, 2019

This Former WeWork Executive Shares the 6 Steps to Leadership He Used to Climb the Fortune 500 Ladder

Francis Lobo had to learn about leadership in record time. He grew up in a family business in India, but after attending graduate school in the US, he decided to stay stateside. Within 5 years of graduating and joining the company, Lobo had already become the President of AOL Services and was running products like MapQuest, AIM, and search. He then became CEO of United Online, a public tech company he brought through an impressive turnaround. Next, he became Head of Services and Chief Revenue Officer of WeWork. Today, Lobo is the Chief Product & Business Officer and President of Labs at Nationstar Mortgage, recently renamed Mr. Cooper. Lobo credits his ability to determine and refine his 6-step method with his rapid rise.

1.     People

The very first thing Lobo mentions every time he discusses a new company is the people he worked with there. It’s no surprise, then, that “People” is his first step to leadership. He explains that companies need “the right people in the right positions in the company. You need the best people.” Lobo admits that it’s easier to hire the best when your company has a strong culture and brand. But even with that, “You really have to have the right motivation. If you don’t have a motivation of love, or you don’t have a mission-oriented motivation in anything you do, you won’t be as happy. You won’t be as successful,” Lobo asserts. He keeps this at the forefront when he interviews candidates. He says, “Anyone I tried to recruit, it’s very, very critical for me to realize that I’m recruiting them to be part of a team, part of a journey, part of a mission. There’s interdependence – we work together. And if that happens, then 1 plus 1 equals 11 every single time.”

2.     Culture

It’s clear that to Lobo, People and Culture are highly steps in his process. Several times in our discussion, Lobo brings up the importance of culture. To Lobo, good culture is creating an environment for people to be as successful as they can be and achieve their maximum potential.” He also repeatedly brings up the importance of community, and particularly points to AOL and WeWork as places where the community in the company wanted to impact the community outside. Lobo says, “It’s all about community and it’s all about people, and creating an environment where people can excel as best as possible.” That feeling helps recruit the best talent and keep the company running on all cylinders.

3.     Data

Lobo is a strong believer in data. “You need to know what you’re working with!” he enthuses. “I believe for any business to be truly successful, you need to have product that makes people’s lives better, and then you take data from that product,” he says. He goes on, “And if you can have a business plan from that data and monetize it, you’ve got the beginnings of a really good business.” Lobo experienced this firsthand in his time at AOL. He explains, “I was able to look at data and find opportunity and growth in that data.” It’s these insights that allowed him to turn around multiple businesses that were declining at doubt digit pace.

4.     Strategy

This step is highly correlated to the prior step of data. Once you know what you’re working with, From that data you have strategy,” Lobo explains. But there’s another element Lobo believes must accompany strategy – “intense focus,” he says. Once your strategy is established, “You give them the resources they need, and you’re able to channel or funnel their focus in the right way, against the specific goal,” he says. “Everything I do is goal-oriented,” Lobo states. He goes on, Every single meeting I go to, every single discussion I have, either I do it verbally or to myself, I say, “What’s my goal in this session?’ And when you have purpose, when you have a goal, and when you have focus attached to that, you will come out in a better place.”

5.     Maniacal Execution

Once you have the right people and the right strategy, you need to stick to the plan. Lobo calls it maniacal execution,” and you can hear the fire in his voice when he says it. “Every day you come to your job, and every day you sit down and…have a goal-oriented approach. You take these 6 steps and say, ‘I’m gonna accomplish this today, and this week, and this month, and this year, and achieve these goals.’ That’s how you can do whatever you want to do,” he describes. His own meticulous exercise of this habit helped him manage his ever-growing responsibilities. “That 6-step approach is something I executed on every day, and over time I just kept managing more and more businesses,” he recalls.

6.     Continuous Innovation

Lobo believes that the company and the employees need to constantly look for ways to improve. But improvement doesn’t have to be complex – sometimes you have to go back to the basics. “I’ll tell you something very silly which I often joke about,” Lobo shares. “I didn’t know what Sarbanes-Oxley was until I was a public company CEO!” Of course, there are more complex issues, too. For Lobo, his goal was always to change the world. “I wanted to take my skills and what I believed I was capable of, and take a company that’s changing the world and help them change the world a little faster,” he says. Indeed, the companies he’s worked for have been changing the world. AOL brought about a revolution in internet access. More recently, WeWork is changing how people conduct business. This passion and energy keep Lobo inspired and his companies always seeking more.

On Fridays, Kevin explores industry trends, professional development, best practices, and other leadership topics with CEOs from around the world.

3 of the Best iPads Are on Sale Right Now

We’re big fans of Apple’s iPad line. We can debate about iPhone versus Android phones all day, but iPads rule the tablet world. They’re speedy, they have higher-quality apps than competitors, and the hardware lasts for years. Today, they’re also on sale.

Apple doesn’t run discounted sales on its own website, but since it began listing more products on Amazon last autumn, it has used (or allowed) the retailer to toy around with larger deals more frequently than usual. Best Buy and Walmart have also been discounting Apple products at what feels like an accelerated rate lately. This week, a few different iPad deals are floating around.

128 Gigabytes for the Price of 32

A 128-GB standard iPad is $329, $100 off its normal price at Amazon. Walmart also has the deal. Normally, you only get 32 gigabytes of storage for $329. If you download a lot of apps, offline Netflix movies, or use your iPad as a camera (I hope not), the storage may really come in handy. I tend to use my iPad as a receptacle for tons of offline downloads on Amazon, Netflix, and Spotify when I travel. If you’re the same, the extra storage is definitely worth it on long trips.

An Old Pro That’s Worth a Fresh Look

Amazon

If you want a nicer iPad, the 2017 10.5-inch iPad Pro is also on sale for $499 ($150 off) at Amazon and Best Buy. It’s even cheaper at Walmart ($479). It’s faster and has a laminated display—there’s less of an “air gap” between the glass and the pixels beneath it. This is especially preferable if you’re using it with the Apple Pencil ($95), since a laminated display more closely mimics the visual experience of seeing ink flow onto the page from the tip of a pen. The 12-megapixel camera and quad speakers are also upgrades. (Normally the 64 GB storage option would also be an upgrade, but not today.)

The $499 price is equal to the new iPad Air, and if you want that tablet, they’re nearly equals. The Air has a camera that’s flush with the body (no little bump) and the tablet performs slightly better on speed tests because it’s newer, but Pro has a better camera, speakers, and screen refresh rate.

The iPad We All Want, but Can’t Afford

Apple

The fancy new 2018 iPad Pro is $749, or $50 off its normal price. It really wowed us with Face ID, powerful internals, and support for the new magnetic Apple Pencil 2. It’s no runaway bargain at this ever-so-mild discount, but it’s still a better deal than you’ll find at an Apple store. The larger 12.9-inch iPad Pro is also $50 off.

One More Thing …

For what it’s worth, the new MacBook Air is also $1,115, or $84 off, and some iMacs are on sale at Amazon. I wasn’t going to point them out, but since these computers are the only other major Apple products with decent discounts, I’ve included them. The Air is currently the Best MacBook for most buyers.

When you buy something using the retail links in our stories, we may earn a small affiliate commission. Read more about how this works.


More Great WIRED Stories

Ethiopia to issue first Boeing investigation report

ADDIS ABABA/SEATTLE/PARIS (Reuters) – Investigators will release a keenly awaited report on the deadly crash of an Ethiopian Airlines jet on Thursday, Ethiopia’s transport ministry said, giving the first official clues to the second crash of a new Boeing 737 MAX in five months.

Some 35 nationalities were among the 157 passengers and crew who died when the nearly full plane crashed six minutes after take-off from the capital Addis Ababa in clear conditions.

The March 10 disaster prompted the worldwide grounding of Boeing’s best-selling plane and scrutiny of its certification process.

“The 10:30 a.m. (0730 GMT) press conference is to present the preliminary report,” Ethiopian transport ministry spokesman Musie Yehyies said.

The report may shed light on how a piece of cockpit software came back to life after pilots initially switched it off as they tried to save the doomed jet, people familiar with the matter said, placing both technology and crew in the spotlight.

The Ethiopian-led investigation has begun piecing together details of flight 302, starting with faulty sensor data on take-off from Addis Ababa, questions over the Boeing 737 MAX’s high speed and a nosedive coinciding with the software re-activation.

The aircraft’s high speed and initial climb suggests the engines were running at a higher than usual thrust, experts say.

The so-called MCAS anti-stall software is at the center of accident probes in both the Ethiopian crash and October’s Lion Air accident in Indonesia that have together killed 346 people.

MCAS was designed to help prevent an aerodynamic stall by issuing commands to push the plane’s nose lower. However, in both cases it is suspected of firing up in response to faulty airflow data from a single sensor designed to measure the ‘angle of attack,’ a parameter needed to avoid stalling or losing lift.

SENSOR PROBLEMS

Echoing the fate of the Lion Air jet, initial evidence suggests the Ethiopian Airlines jet experienced sensor problems shortly after take-off, causing the MCAS software to begin lowering the nose to grab air under the wings.

Unlike the Lion Air crew, who were flying at a time when pilots had been told little about the MCAS software, the Ethiopian crew used switches to turn the automatic system off but it later re-engaged, people familiar with the matter said.

Although aircraft experts say MCAS cannot turn back on by itself, the report is expected to shed light on whether and why the crew chose to restore electrical power to the system at the risk of setting off more automated nose-down movements.

Aerospace analyst Bjorn Fehrm said in a blog post for Leeham News that pilots may have deliberately re-activated the system in order to make it easier to trim or control the aircraft only to be overwhelmed by rapid counter-moves from MCAS.

Investigators will also look at whether the crew carried out all necessary procedures, including a recommendation to stabilize the plane using the trim system before turning the crucial software off.

The pilots maneuvered the plane upwards at least two times before hitting the stabilizer cut-out switches to disable the system, one person familiar with the matter said.

However, initial flight data indicates the aircraft was flying nose-heavy and not in a “neutral” attitude when pilots hit the cut-out switches, the person added, making the situation harder to manage.

Cockpit procedures call for pilots to leave the MCAS system off for the rest of the flight once it has been disengaged.

GLOBAL GROUNDING

Safety experts stress the investigation is far from complete and most aviation disasters are caused by a unique combination of human and technical factors.

In a statement, Boeing said: “We urge caution against speculating and drawing conclusions on the findings prior to the release of the flight data and the preliminary report.”

The 737 MAX is Boeing’s top-selling jet with almost 5,000 on order. Ethiopian Airlines is also in the midst of an expansion drive, while other 737 MAX customers and victims’ families want answers, and potentially compensation.

Boeing shares were down 1.5 percent at 1800 GMT. They have lost more than 8.5 percent since the Ethiopian crash.

Getting the planes flying again depends partly on the role that Boeing design features are found to have played in the crash, though investigators are also paying attention to airline operations, crew actions and regulatory measures.

FILE PHOTO: Airplane engine parts are seen at the scene of the Ethiopian Airlines Flight ET 302 plane crash, near the town of Bishoftu, southeast of Addis Ababa, Ethiopia March 11, 2019. REUTERS/Tiksa Negeri/File Photo

Boeing is upgrading the MCAS software and training while stressing that existing cockpit procedures enable safe flight. It however faces mounting lawsuits alleging poor system design.

The Wall Street Journal reported earlier that the pilots had initially followed Boeing’s emergency procedures but later deviated from them as they tried to regain control of the plane.

Reporting by Eric M. Johnson, Tim Hepher, Jamie Freed, Jason Neely, David Shepardson; Editing by Georgina Prodhan, Mark Potter and Kirsten Donovan

What Boston Dynamics' Rolling 'Handle' Robot Really Means

For internet-goers, Boston Dynamics is that company that uploads insane videos of the humanoid Atlas robot doing backflips, of four-legged SpotMini opening doors and fighting off stick-wielding men, and as of last week, of a Segway-on-mescaline called Handle jetting around picking up and stacking boxes with a vacuum as its arm. For journalists and industry watchers, however, Boston Dynamics is that company that almost never talks about where all of this work is ultimately headed.

That’s beginning to change. The company is now teasing its ambitions as the four-legged SpotMini nears its commercial release. Today, Boston Dynamics is getting even more explicit about its vision with an announcement that it’s acquired a Silicon Valley startup called Kinema Systems, which builds vision software that helps industrial robot arms manipulate boxes. This acquisition is giving the Handle robot the gray matter it needs to follow SpotMini to market. What for years has been fodder for internet video gold is now taking shape as a unified vision of the robotic future.

One of the biggest obstacles holding robots back has been their limited perception. We humans enjoy a rich constellation of senses that help us navigate our surroundings. Robots need the same, lest they destroy themselves. Go to pick up a box, for example, and you as a human probably don’t think deeply about the lighting, and how it may cast shadows that throw off your hand placement.

Kinema’s software—which is robot-agnostic, meaning it already works on a range of robots beyond Handle—helps the machine through all these challenges. “Their system is able to look at a stack of boxes,” says Michael Perry, vice president of business development at Boston Dynamics, “and despite how ordered or disordered the boxes are, or the markings on top, or the lighting conditions, they’re able to figure out which boxes are discrete from each other and to plan a path for grabbing the box.”

That’s a huge part of what Handle, a robot designed to work in warehouses, needs to do. But the robot also relies on its overall shape to do its (soon-to-be) job. This is where BD’s larger strategy gets even more interesting: Although Handle, Atlas, and SpotMini look almost nothing alike, they are in fact intimately connected.

“Handle isn’t entirely different from Atlas,” says Boston Dynamics boss Marc Raibert. Indeed, a video of Atlas three years ago showed the robot picking up boxes with two arms that ended in stubs, arms that Handle wielded in its own video a year later. The challenges of bipedal locomotion are largely the same, namely the balance problems that a four-legged robot like SpotMini doesn’t share, as are the challenges of manipulation with two arms, which SpotMini (being the dog to Atlas’ human form) also doesn’t share.

LEARN MORE

The WIRED Guide to Robots

But this is the beauty of robots. You can iterate on their shapes to tailor them to different tasks and environments. Atlas walks on two legs and Handle rolls on two wheels, but either way, that bipedal locomotion cuts down on the robots’ footprint. “If it was a four-wheeled robot, it would have to be much larger in order to get that level of reach and lift boxes,” says Perry. “So this is a robot that’s designed to go into human-purposed environments and still be able to complete a task.”

The reason BD is able to riff on its robot shapes with relative ease boils down to one big thing: repurposed software.

When you think Boston Dynamics, you probably don’t first marvel at the code that’s running these machines—BD is famous for its hardware. But Raibert takes issue with that characterization. “I think it’s a misconception that we’re a hardware company,” he says. “The only reason any of our machines do what they do is because of the controls and perception and the systems that coordinate with the hardware. It’s just that our hardware is so strong, that’s what makes us look like that.”

Someone, after all, has to program Atlas to do those backflips. SpotMini needs software to autonomously navigate its world. And two-wheeled Handle needs finely tuned control algorithms to keep from falling on its face. BD works out these algorithms across its platforms. “There’s a lot of stuff that flows,” says Raibert. “The next group uses a lot and then creates their own stuff, and then that flows back.”

With a cognitive core that’s developed over time and shared across platforms, BD has been able to devote energy to honing each of its robots’ specialties. In SpotMini’s case, it’s about becoming an expert at navigating challenging terrain. “When we’ve been looking at applications for Spot,” Perry says, “we’re very careful to screen out tasks we think a wheeled or tracked robot could do even better.” SpotMini is a good match for environments that transition from one terrain to another. “So street to curb, stairs, lips between rooms,” he says.

A relatively structured environment like a warehouse, on the other hand, tends to be a great place for a wheeled robot. Clutter can make such places chaotic, sure, but in general the robot can rely on a flat, smooth surface to glide across. In such an environment, wheels are often more efficient than legs: Handle can manage four hours of operation on a charge, whereas with SpotMini it’s more like an hour and a half. And Handle could potentially go even longer. Swinging around Handle’s backside is a counterweight that could hold even more batteries, Raibert says.

The previous iteration of Handle had stump arms instead of a single vacuum arm. Also notice that the bulk of the weight is in the torso, whereas the new version has a swinging counterweight on its rear end to balance.

Boston Dynamics

Plus, a human worker can wield Handle as a unique kind of tool. “It’s also got a mode where it can squat down and you can manually wheel it around,” says Raibert. To be clear, BD doesn’t intend Handle to be a particularly collaborative robot—it’ll likely work in isolation from humans, unloading pallets autonomously while humans take care of other tasks in the warehouse. At least, that’s the plan.

Accordingly, Handle is a bit simpler as far as perception is concerned. It’s got one camera to localize itself in space, another for obstacle avoidance, and another looking for the best place to grab a box. SpotMini, on the other hand, “is trying to be a little more general purpose,” says Raibert. “So we have cameras looking in all directions.”

With Handle stacking boxes and SpotMini wandering more widely, perhaps inspecting oil and gas operations, Atlas’ destiny might lie somewhere in between. Its legs allow it to stomp over difficult terrain, but its humanoid form might make it better suited to navigating indoor spaces designed for humans. It could one day, for instance, climb ladders, which would befuddle Handle and SpotMini.

But all that hardware we’ve been marveling at over the years has been a kind of illusion—sophisticated machinery, to be sure, that obscures equally sophisticated software. With the acquisition of Kinema Systems, BD not only bolsters the software side of things, it can now sell that system for use in warehouse robots it doesn’t manufacture itself.

Oh, and it means Boston’s most famous robotics company now has a base of operations on the West Coast. “We’ll have machines out there, but they’ll be for the development of the applications and perception and software,” says Raibert. “Our current plan is to keep the core of the hardware engineering here. We’ll see how that evolves.”


More Great WIRED Stories

Apple Card vs. American Express Platinum: A High-End Credit Card Face-Off

Apple Card is coming this summer. And when it’s released, don’t be surprised if some industry watchers compare it to a variety of high-end credit cards including the American Express Platinum.

The tech giant unveiled the credit card during its otherwise star-studded Hollywood upfront-like event last week. While the Apple Card will allow users to make purchases anywhere MasterCard is accepted, it’s also quite unlike many other cards available today.

Apple Card doesn’t have late fees, annual fees, international fees, or over-limit fees. And although users will be encouraged to pay their monthly statements, technically, if you want to skip a month, you can. Just be ready to pay extra interest.

For its part, Apple has called its credit card revolutionary and suggested it could upend the industry. But it’s also entering a market that’s dominated by companies including Visa, Chase, American Express, and others, that have the credit card game down to a science.

“Apple is starting from the bottom of base camp in what looks to be a climb of Mt. Everest to gain 5% to 7% market share over the next few years,” Wedbush analyst Dan Ives told Fortune in an interview. “The stage is set, now it’s about will consumers sign onto the platform and use the Apple Card over the coming years.”

Ives went on to say that Apple’s clear opportunity is “tapping into its 1.4 billion active iOS devices and monetizing the financial vertical.” He believes the company could attract 20 million Apple Card customers in its first three years.

But before it can get there, Apple Card needs to intrigue users. As a metal card with plenty of incentives for using it, I can’t help but compare it to the American Express Platinum—a similarly high-end metal card with its own slate of perks that some may find compelling.

So, to shed some light on where Apple Card might fit in the market, I’ve compared the Apple Card and American Express Platinum below.

From design to interest, here’s a look at how the two cards compare:

Design

Apple Card and American Express Platinum are clearly vying for metal credit card dominance.

Apple Card has a titanium design and laser-etched Apple logo and cardholder name. There’s a security chip in the card, but no credit card number, CVV, or any other information.

The American Express Platinum is a metal card that features American Express’ logo and border and uses the same security chip in the Apple Card. Unlike the Apple Card, the American Express Platinum displays credit card number, expiration date, and CVV numbers in addition to the cardholder’s name.

Fees

Apple created some buzz when it said that cardholders won’t be subject to annual fees, international usage fees, over-limit charges, and even late fees.

American Express Platinum, on the other hand, is quite different. Cardholders are charged a $550 annual fee for having the card and up to $38 for both late payments and returned payments. And if cardholders want to add cards to their account for family members, American Express charges $175. There’s no charge for foreign transactions.

Interest

American Express Platinum cardholders pay off their balances each month to avoid interest on purchases. However, the company has launched a program for customers to carry a balance on the American Express Platinum and pay over time.

That program, aptly named Pay Over Time, will charge interest on the card’s outstanding balance. American Express doesn’t say on its dedicated Pay Over Time page how much interest it charges.

Apple Card balances come with an interest charge on the outstanding balance. As of this writing, Apple is predicting a variable annual percentage rate (APR) on Apple Card balances of 13.24% to 24.24%.

Perks

The iPhone maker’s Apple Card perks are based on a simple, 3-2-1 system.

Whenever cardholders buy products at any Apple marketplace, including its brick-and-mortar stores, its App Store, or iTunes, Apple will automatically give them 3% of the purchase price back.

Apple Card can be linked to Apple’s mobile payment service Apple Pay, which lets users hold their iPhones to purchase terminals and buy goods. When they use Apple Pay linked to the Apple Card at any store, they will get 2% of the purchase price back.

Using the actual Apple Card at a store will earn cardholders 1% cash back.

American Express has no shortage of perks and offers available to Platinum cardholders.

The card comes with up to $200 in free Uber rides each year. American Express also uses a membership points system that adds points to cardholder accounts based on their purchase. Those points can be redeemed for cash, travel, or retail gift cards.

American Express Platinum cardholders can also get up to $200 in airline fee credit to cover extra baggage costs, in-flight refreshments, and more. And with American Express’ partnership with hotels around the world, Platinum cardholders can get discounted rates when using their cards to book a room.

Add that to access to lounges at airports around the world and other cashback offers, American Express Platinum is a clear leader in perks.

Business

So far, Apple has not said whether the Apple Card will be available to businesses. On the company’s website promoting Apple Card, the focus is on the consumer.

Meanwhile, American Express offers its Platinum card to both consumers and businesses.

Apple Has A New iPad Problem

With the amount of press surrounding Apple’s (NASDAQ: AAPL) services business, some analysts seem to be forgetting that today, right now, the company gets 87% of its revenue from products. In the future, services may outstrip products, but products will still be a gateway. One product that doesn’t seem to get much press is the company’s iPad business. With the iPhone business on shaky ground, Apple really can’t afford for its iPad or Mac business to struggle as well. Last quarter, iPad and Mac made up just over 19% of Apple’s total revenue. Unfortunately, Apple’s most recent iPad announcements could spell trouble for both businesses.

The tablet market is shrinking… wait its growing… actually we don’t know

The tablet business seems to have reached a point where analysts can’t figure out if customers want these devices or not. A study from Statista suggests the following bleak picture:

Year

2018

2019

2022

2023

Tablets Sold

150.3m

136.8m*

134.1m*

122.1m*

(Source: Statista tablet market *projections)

Users are being told that tablet demand will decline as users choose to buy laptops, Chromebooks, and hybrid devices. On the one hand, Apple has a stranglehold on tablet operating systems. On the other hand, it doesn’t do much good to own a market that is falling apart.

There is just one problem… the market may not be falling apart after all. As quickly as you can find a study that says tablet sales will decline, you can find another that suggests tablets will return to growth. According to Frost & Sullivan, the tablet market will grow at a compound annual growth rate of 1.6% through the year 2024. Though this doesn’t sound like fast growth, by 2024, it would mean 187 million units worldwide.

iOS worldwide tablet market share

(Source: iOS Worldwide Tablet market share)

This second study suggests tablets will gain new capabilities and cannibalize laptop sales. Assuming that iOS continues to hold its 74% worldwide market share, by 2024, the company would sell about 138 million of these devices. To give investors an idea of what this growth would look like, during 2018, the company sold roughly 43 million tablets.

In the last four quarters, the average revenue per iPad has hovered around $430. At 138 million devices, this would generate about $15 billion in quarterly revenue. Last quarter, Apple generated $6.7 billion from the sale of iPads. More than doubling its iPad revenue in the next five years, is something investors would be more than happy to accept.

If this were the end of the story

It seems like every time Apple begins to solve a problem it ends up creating another. The first example is the iPhone X. It seems obvious in hindsight, that users who bought the iPhone X were those who waited through the iPhone 6, 7, and 8, for a different form factor. Priced at $999, this premium device seemed to answer a question of demand for something different. Apple mistakenly assumed that everyone would pay $1,000 for a phone, and rolled out the iPhone XS, and iPhone XS Max. The company eventually made the iPhone XR available at $749. In the meantime, Apple witnessed slower iPhone demand than expected and is still trying to make up lost ground.

A second mistake was the timing of the introduction of the iPad Mini. The original iPad was released in April of 2010. Over the next two plus years, others introduced smaller tablets while Apple introduced the iPad 2 and even the iPad 3. With users salivating over the Retina display on the iPad 3, in November 2012, Apple introduced the iPad Mini… without the Retina display. In fact, it took until a year later when the Mini finally gained a Retina display of its own.

Unfortunately, Apple seems to have again created a problem just as it began to solve another. Over the course of 2018, Apple’s iPad sales momentum deteriorated significantly.

Quarter

Q1 2018

Q2 2018

Q3 2018

Q4 2018

iPad Revenue Growth

+7.3%

+5.1%

-6%

-14.6%

(Source: AAPL quarterly results Q1 ’18Q2 ’18Q3 ’18Q4 ’18)

It’s really no surprise this slowdown occurred, given Apple’s iPad release schedule. The iPad from March of 2017, was a step up in performance, but not transformative. June 2017’s introduction of slightly different iPad Pro devices didn’t give users a reason to rush out and upgrade. Users had to wait for the iPad Pro 3rd generation, and the elimination of the home button, to get a change worth noticing. Since the devices weren’t released until the fourth quarter was already one-third over, it’s not surprising that results suffered.

Buyers seemed impressed by the advance of the iPad Pro lineup, because iPad revenue jumped more than 13% in the first quarter of 2019. If Apple left its lineup alone for another quarter or two, iPad results might have continued to improve, but unfortunately that’s not the end of the story.

Letting the Air out of iPad results

The introduction of the newest iPad Air and the revised iPad Mini is a confusing choice on multiple fronts. First, the iPad Mini hasn’t seen a refresh since 2015. Second, the iPad Air hasn’t been a name used since 2014. It’s almost like Apple forgot it sold these devices in the first place.

Device

Price

Screen

Weight

Storage

Processor

Other

iPad Air 2019

$499

10.5”

1 lbs

64 GB

A12 Bionic – M12 coprocessor

Touch ID – Lightning connector

iPad Mini 2019

$399

7.9”

0.66 lbs

64 GB

A12 Bionic – M12 coprocessor

Tough ID – Lightning connector

(Source: Apple iPad Air and iPad Mini pages)

If these devices were standalone products, the company’s results would likely be amazing. The problem is Apple now gives users a lower-priced choice, that may cut into the average selling price that just started to recover.

Quarter

Q1 2018

Q2 2018

Q3 2018

Q4 2018

Q1 2019

iPad Average Selling Price

$446.97

$450.55

$405.17

$422.68

*$507.57 to $511.45

(Source: AAPL quarterly earnings – *ARPU for Q1 is derived from Apple selling 13.1 to 13.2 million iPads in the first quarter the last two years. Taking the $6.7 billion in iPad sales and dividing by 13.1 or 13.2 million yields the range of $507.57 to $511.45)

iPad Air 2019

(Source: iPad Air page)

On the iPad Air side of things, buyers who are faced with either the iPad Air or one of the iPad Pro models have a relatively simple choice. For $300 to $500 more, they can get a Pro device with a faster processor and a somewhat better screen to form factor, or they can just get the iPad Air. In short, the iPad Air is such a good device, with such impressive specs, that it will likely make buyers wonder why they would waste $300 or more on an iPad Pro.

In addition, if users are considering a MacBook 12”, the pricing between the iPad Air becomes astoundingly challenging. The MacBook is $1,299 to start and has a screen with 226 pixels per inch (PPI). The device weighs 2 pounds and comes with 256 GB of storage. Instead of spending $1,299, a buyer could opt for an iPad Air with the Smart Keyboard. A 256 GB iPad Air costs $649, the Smart Keyboard runs $159 for a total cost of $808.

Though the iPad Air runs a different operating system, the Files app moves the device a little closer to the Mac. What’s more, the iPad Air is half the weight without the keyboard, supports Apple Pencil, and has a higher ppi at 264. This is a case where spending almost $500 less gives the user a little less screen, but is that enough to justify the huge cost difference? Probably not.

Where the iPad Mini is concerned, it’s possible that some users will opt for the Mini at $399 instead of the slower processor and bigger screen iPad at $329. However, it’s equally possible that some buyers who would have bought the Air at $499, will decide to save the $100 and cut weight by going with the Mini. The bottom line is Apple revived two products from multiple years ago, just in time to potentially cut the gains in iPad ARPU that just started.

Today versus tomorrow

A significant number of articles are being written about what Apple will look like tomorrow. The company’s services business is being talked about as though its growth is guaranteed. It’s possible services will overtake products, but that is a long road indeed. To be clear, I own Apple stock, I plan on holding onto my shares. I do think that services will be a growth driver for a long time.

However, Apple has shown it’s not infallible in releasing major products. The iPhone XR was released after the XS, yet based on demand, the situation should have been reversed. Apple made a strong move redesigning the iPad Pro lineup with their beautifully stretched screens, USB-C charging, and keyboard and pencil support. Users had a choice between spending $329 on the cheaper iPad or $799 or more for an iPad Pro. The ARPU seems to have moved up significantly on the strength of Pro sales.

Though the iPad Air and iPad Mini are designed to drive demand for iPad sales, the pricing may put a squeeze on ARPU again. In addition, Apple made the new iPad Air so capable, that some buyers may choose to opt for the Air over certain MacBook models. The company has been successfully pushing the price point for the MacBook up to drive higher revenue usually on relatively tepid unit growth. If the redesigned Air with its keyboard support steals MacBook sales, Apple’s overall revenue could suffer significantly from the massive difference in pricing.

Apple is constantly attempting to improve its hardware options to appeal to many users. At the present time, the lineup between the iPad Mini, iPad Air, iPad Pro, and MacBook has gotten a bit convoluted. Apple is hoping that users will stay in their lane and buy what it wants them to. Unfortunately, Apple may have just caused its own iPad problem.

Disclosure: I am/we are long AAPL. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

4 Differences Between an ICO and a Penny Stock

The coins sold by small companies in Initial Coin Offerings are often compared to penny stocks. Like penny stocks, they’re cheap. Penny stocks cost less than five bucks; a new coin released at an ICO can literally cost a penny or less. They also have the potential for huge returns. Monster Beverage, a drinks company, was selling at around 60 cents a share at the start of 2005. It’s now worth nearly $60 a share. If you had bought $100 of those shares fourteen years ago, you’d now be sitting on nearly $10,000. That’s not as high as the returns earned by early Bitcoin investors but it’s still worth having. There are some important differences between penny stocks and cheap coins from ICOs though. Here are four of them:

  1. An ICO Doesn’t Give You a Company

Penny stocks might be cheap but they’re still stocks. They give you a share of a company, possibly with voting rights. An ICO only releases a product whose value you hope will rise. It’s like a new casino raising funds by selling its unique poker chips cheaply. If the casino is popular those chips could be worth a lot of money. But if the casino is never built, you’ll be left with a pile of useless discs.

  1. You Can Research the People Behind the ICO

One reason that a penny stock is such a high risk is that there’s often very little information about the company or the people behind it. You might not know who the managers are, what they did before they launched the company or whether they’re serious. You might know no more than the price of the stock and the name of the business. The rest is a shot in the dark.

Before launching an ICO, cryptocurrency firms release white papers. Those white papers will explain the background of the people launching the firm. You can often contact them on Telegram and ask them questions. That doesn’t mean that you can find all the information you want, or always get the answers you need. There will always be gaps and risks. But ICOs can provide details about the people behind them.

  1. You Can Research the Business Idea

The white paper should also explain what the company is doing and how it plans to do it. Again, that doesn’t mean that the company will actually do what it says. It doesn’t mean that the managers have the skill or the competence to do what they intend. But you should be able to assess their idea and decide for yourself whether or not you think it has legs. A bet on an ICO is a bet on a business idea.

  1. Coins Are Easier to Buy and Sell than Penny Stocks

Penny stocks are usually bought and sold through brokers. The markets are illiquid, the commissions are high and the process isn’t straightforward. The products of ICOs aren’t always sold on major cryptocurrency exchanges but you can usually buy them directly from the companies and if the coin is a success, you can expect it to be listed in the future.

“Easier” isn’t the same as “easy” though. Trading volumes will still be small. Not all coins will be listed on an exchange and those that are listed, often find themselves on small exchanges.

Like penny stocks, buying a small coin at an ICO is a high risk venture. But you can keep your losses low, and who knows, you might just strike it big!

Published on: Mar 31, 2019

The Art of the Bracket Smack Talk, Perfected With Purpose Inside Your Company

Host a company-wide bracket challenge – they’re already doing it anyway, even if you don’t know about it. Offer a grand prize that doesn’t cost cash, like an extra day off, or convert an office lunch you were already planning into a bracket-themed bonanza. Then have a bunch of smaller inexpensive gifts for top contenders. Encourage a number of smaller pools by team to increase everyone’s chances of winning something.

To do March Madness right, and to have more fun, you must introduce smack talk. It doesn’t have to be mean or dirty. Sports have provided a long line of terrific trash talkers, but rappers, celebrities, and soldiers have all contributed greatly to the Western cannon of smack.

2.     “Don’t wish it were easier, wish you were better.” – Jim Rohn

3.     “I don’t care what you think about me. I don’t think about you at all.” – Coco Chanel

4.     “I have heard there are troubles of more than one kind. Some come from ahead and some come from behind. But I’ve bought a big bat. I’m all ready you see. Now my troubles are going to have troubles with me!” – Dr. Seuss

7.     “May God have mercy upon my enemies, because I won’t.” – General George S. Patton

9.     “Ladies and Gentlemen, I don’t know whether you fully understand that I have just been shot, but it takes more than that to kill a Bull Moose.” – Theodore Roosevelt

10.  “I don’t talk trash often, but when I do, I go for the jugular.” – Kobe Bryant

11.  “I’m just looking around to see who’s gonna finish second.” – Larry Bird

14.  “I got my own back.” – Maya Angelou

15.  “Get your popcorn ready, ’cause I’m gonna put on a show.” – Terrell Owens

18.  “The best thing I like about human beings is that they stack so neatly.” – Frank Underwood

19.  “If he calls that number, I’ll be sure to pick up after the fifth ring.” – Kobe Bryant

20.  “All right. They’re on our left, they’re on our right, they’re in front of us, they’re behind us … They can’t get away this time.” – Lt. Gen. Lewis B. “Chesty” Puller

23.  “I wish people would love everybody else the way they love me. It would be a better world.” – Muhammad Ali

Leaky Databases Are a Scourge. MongoDB Is Doing Something About It

MongoDB, a database software provider whose stock has been on a tear recently, just hired its first-ever chief information security officer. The appointment, which came Friday, signals that the company plans to take security more seriously even as it faces stiffened competition from the likes of Amazon and other tech giants.

The new boss is Lena Smart, a Glaswegian cybersecurity professional. Smart formerly held the same title at IPO-bound Tradeweb, a financial services firm that supplies the technology behind certain electronic trading markets. Prior to Tradeweb, she headed security at the New York Power Authority, where she worked for more than a decade. A cellist in her spare time, Smart told me in her Scottish brogue that her priority in the new job will be “knowing what the crown jewels are—that’s our customer data—and making sure that’s always protected.”

People leaving MongoDB and other databases unsecured on the web has been a persistent source of data-leaks over the years. Just this month, a security researcher discovered one such sieve that exposed to public view a trove of sensitive information, including location data, on millions of people in China. The misconfigured repository appears to have originated from SenseNets, a Shenzhen-based company that is likely providing the Chinese government with crowd-surveilling, facial recognition technology to track the country’s muslim Uyghur population. This is just the latest leak example; there are innumerable others.

Despite the frequency of these leaks, the situation seems to be improving. Most of these inadvertent leaks have sprung, in fairness, from people using outdated instances of the company’s so-called community edition software, a free, barer-bones version of the database product. Mark Wheeler, a MongoDB spokesperson, conceded that the 12-year-old company “struggled in its early years to find the right balance with security.” But he avers that updates to the default settings of MongoDB’s software over the past few years, plus key security team hires—including guardians Davi Ottenheimer, Kenn White, and now Smart—are changing the equation.

As Smart’s scope involves securing the totality of MongoDB’s business, the data-spillage issue ultimately falls to her. She says she’ll continue educating customers in best practices when it comes to security. She says she will also aim to imbue the company’s product development process with security, quality assurance, and testing from the earliest stages. “If we can get in at the very start” of the software development lifecycle, Smart says, it will “save us time and money and make our products more reliable and secure.”

The leaky database issue is one that extends well beyond MongoDB. It’s also a problem for rivals such as Amazon, particularly its S3 buckets, Elastic, and others. Like so many companies, these database-makers are looking now to shore up their software in the hopes of turning a historical weakness—cybersecurity—into a competitive strength. As Dev Ittycheria, MongoDB’s president and CEO, tells Fortune: making the company’s products as secure as possible “is critical to our business.”

Indeed, it’s critical to MongoDB and, increasingly, every business.

A version of this article first appeared in Cyber Saturday, the weekend edition of Fortune’s tech newsletter Data Sheet. Sign up here.

My Company Asked More Than 1,200 CEOs About the Most Important Decision They'll Make in 2019. Here's What They Said

Talent issues are top-of-mind for business leaders in 2019. In fact, according to a new report from my organization, CEOs of small and midsize businesses rank decisions about talent higher in importance than decisions about customers and financials.

Despite concerns for the economy, 65 percent of the 1,257 CEOs included in the Q4 2018 Vistage CEO Confidence Index report said they planned to increase hiring this year. This was a shade lower than the recent 15-year peak of 71 percent, but current recruitment intentions are still greater than at any other time recorded by our survey since 2003.

With the United States almost at full employment and wages rising, hiring won’t be easy during the next 12 months. CEOs are employing a variety of strategies to cope with this challenge, such as boosting wages (64.6 percent), adding employee benefits (36.1 percent), investing in equipment to automate tasks (35.2 percent) and allowing employees to work remotely (25.1 percent).

However, there are other strategies to consider. Here are two that I recommend.

1. Work smarter, not harder, on talent sourcing  

One of the most effective ways to source new talent is through employee referrals. Engage your existing workforce in the recruitment process by starting a referral program that provides incentives to employees to help bring the best people on board. Start an open conversation with all employees about how you reward effort to address any questions about compensation.  

Professional networks can also be a valuable source of top talent. Use them to target both people who are actively seeking new roles and those who are happy working elsewhere, even if that’s with your competitors. 

2. Get creative with professional development

Nearly three-quarters (71 percent) of the CEOs we surveyed recognized employee development will be key to their talent-management strategy in 2019. People are a business’s number-one asset. Giving them room to grow in the organization is one of the most effective ways of making them feel more valued, which can increase their productivity and make them stick around longer.

But don’t stop at training workers to be better in their current roles. Give them opportunities to develop communication, collaboration and leadership skills, and recognize their achievements with managed career progression. Connect senior leaders to junior team members through a mentorship scheme, too, and you’ll soon have a strong talent pipeline. 

Burley Encore X Review: A Fun but Flawed Bike Trailer

“There’s no reason to be afraid,” my spouse scolded, as my 1-year-old and 4-year-old shrieked at the top of their lungs. You would’ve thought they were being roasted alive, instead of merely strapped into the Burley Encore X as their parents gingerly hauled it down a small, steep hill to the beach.

For a minute, the stroller was poised over a three-foot drop. I held the roll bar from the top and lowered it to my spouse as I braced my feet on a tree root and thought, “Hey, I might start shrieking, too.” You can’t blame toddlers for tantruming when the tantrum makes perfect sense.

Our kids are used to this. Ever since my son has been big enough to hold his head up on his own, we’ve been hauling them around in the active parents’ bike trailer of choice, a Thule Chariot. The Chariot has different iterations at different price points, but each iteration can be modified for jogging, biking, or cross-country skiing.

This year, Burley released a series of new, rugged child bike trailers. While the the Eugene, Oregon-based company is known for super-safe designs, it’s hoping that the new Cub X, D’Lite X, and Encore X will get more Burley trailers off the streets and onto the sand, snow, and dirt.

I opted to test the Encore X performance sport stroller-trailer. It has suspension, in comparison to the more affordable Encore, but fewer of the luxury features of the D’Lite model. After a few weeks of testing, I still prefer our Chariot. But Burley’s many fans will find plenty of reasons to love the Encore X.

And It Was All Yellow

Burley

The Encore X is easy to assemble and use. Like Burley’s jogging stroller, the Solstice, the manipulable parts are set off in bright yellow plastic, so you know exactly which parts you are supposed to wrestle with and which ones you should leave alone.

At 31 inches across, it’s narrow enough to fit through our front door—just barely—and at 24.7 pounds, it’s lighter than our Chariot Cheetah, which weighs 26.5 pounds. It comfortably fits my two kids, but it’s worth noting that its total capacity is only 100 pounds. I’m probably only going to be able to carry both children in it for another year or so.

I might be able to use it for a little longer if I can resist packing it full of stuff. The Encore X has an awe-inspiring cargo capacity. It’s hard not to start tossing random things into the 60-liter cargo bin, like picnic blankets, tennis rackets, or dog food. You can also remove the seats to convert it to a cargo trailer.

It also comes with a one-wheel stroller conversion kit. To use it, screw the Burley hitch on your rear axle. When you want to bike, hook up the trailer hitch with by sliding in the pin and locking it; flip small front wheel up and you’re ready to go. When you want to convert it to a stroller, unhook the pin and flip the front wheel down. The transition is quick and easy, and unlike the Chariot, you don’t have to worry about finding a way to carry or store the hitch bar. Some convertible strollers, like the Thule Chariot, do have a sturdier ball-and-socket attachment in addition to a pin.

Finally, the Encore X comes with all the standard features that help make the company’s trailers so beloved among biking baby-havers: it comes with a skid guard to protect the bottom of the trailer, and the wheels have guards and are easy to switch out with the pop of a big, yellow button.

And the suspension works! I biked two kids and all their stuff on everything from dirt trails, to sand and gravel paths, and no one protested or cried (except for that one time).

Not so Burly

Burley

As a bike trailer, the Encore X is nearly perfect. For two weeks, I towed my children to and from school. A sunshade and UV-protective panels protected my kids from the sun, and the big storage container meant that I didn’t have to attach panniers to my bike rack to carry all their backpacks and jackets. I could throw in a friend’s skateboard in the back when he wanted to walk with us, or a basketball to play at the park.

When I took it on more adventurous excursions, cracks began to show. The Encore X meets ASTM F1975-09 safety standards and survived extensive drop- and crush-testing thanks to its heat-treated aluminum roll frame, but I have some concerns with its durability.

The first flaw is that the trailer’s handlebar doesn’t lock into place. When I picked up the bike trailer an inch or two to pull it around a gate or over a curb, the handlebar popped out, rotated, and plonked my children on the ground. When we had to lift the trailer over a log on the trail, my spouse and I picked the stroller up by its frame and ignored the handlebar altogether; it was just easier.

Burley assured me that you can tighten the clamp to lock the handlebar in place. However, in order to do so, you need to pop out the barrel nut that holds the handlebar in place. And if you tighten it too much, you might snap the handlebar’s cinch lever. As I pondered this conundrum, I couldn’t help but think that a sport trailer should be a little hardier than this.

I also wonder how long the Encore X will hold together. The fabric is made from tough 600-dernier polyester, but after a mere two weeks of being folded up and shoved in the back of my car, it has already started to wear through. The damage isn’t covered by the three-year warranty. Burley suggests a little Tenacious Tape might do the trick, but I’ve owned the Thule Chariot for three years and put it through similar paces, and its only signs of wear are fading from the sun.

The Thule Chariot’s accessories also just make more sense. For example, the Chariot’s two-wheel stroller kit is included in the base price, whereas with the Burley, the two wheel stroller kit is an add-on. The one-wheel stroller conversion kit might be more convenient in some ways, but I missed having two wheels. They make the stroller smaller and easier to maneuver, and I wouldn’t want to pay extra for them.

I was excited to test Burley’s sand- and gravel-riding kit, but I found that the big, fat, 16-inch tires were unnecessary. If you want to bike to the beach and push the stroller through sand, you have to buy the $149 jogger kit on top of the $199 fat tires. Without the jogger kit, the puny front tire sunk into the sand, tipping the stroller forward.

If you pick the Encore X, my advice is to skip the sand kit and stick with the ski kit for snow. Opt for the jogger kit if you want to go on sand or trails, or the two-wheel kit if you live in a city.

Encore Ready

If you want a one-and-done bike trailer that you can also hoist over a tree root without your children screaming, my vote would still be for one of the Thule Chariots like the one I recommended in our Best Strollers guide. Still, I found it to be a surprisingly difficult decision.

The Encore X has many admirable qualities, especially if you don’t go off-roading very much. It’s lighter and narrower, with much better storage options. With a few refinements to improve its durability, and a little Tenacious Tape, I might see a lot more of these on the roads and trails this summer.

Auto1 may consider IPO in future but no need for cash now: CEO

BERLIN (Reuters) – German used-car dealing platform Auto1 said it could seek a public offering in future but a 2018 cash infusion from Japan’s Softbank means it has no immediate need for extra funding of its European growth plans.

FILE PHOTO: A worker loads a second hand car on a car transporter truck at the Auto1.com company grounds in Zoerbig, Germany January 28, 2017.REUTERS/Fabrizio Bensch /File Photo

Last year’s Softbank’s deal valued Berlin-based Auto1 at 2.9 billion euros ($3.27 billion), making it one of Germany’s top so-called tech unicorns.

It is virtually unknown to consumers except through its used car buying arm Wir Kaufen dein Auto (We Buy Your Car) in Germany and similar names elsewhere. It operates from Finland to Romania to Portugal, 30 countries in all.

Revenues rose by 32 percent to 2.9 billion euros last year, and although it is profitable in Germany, investments in other markets have led to a loss on group level.

“Currently, an initial public offering is not a topic for us,” Auto1 co-founder Christian Bertermann told Reuters, adding this could change in future.

Auto1 buys cars using its vehicle pricing database to calculate an offer within minutes and then sells the vehicles on to one of its roughly 35,000 dealerships for a commission.

Its platforms helped 540,000 vehicles change hands in 2018.

The company will now also start a retail platform to compete with Scout24’s Autoscout unit or Ebay’s Mobile.de offering, Bertermann said.

He confirmed a Reuters report about Auto1’s talks with Scout24 about an acquisition of Autoscout, adding that these would not lead to a takeover.

Scout24 in February agreed to be acquired by buyout groups Hellman & Friedman and Blackstone.

Auto1 was set up in Berlin by entrepreneur Christian Bertermann after having trouble selling two old cars owned by his grandmother, along with Koc, who previously worked at Rocket Internet-backed firms Zalando and Home24.

Reporting by Nadine Schimroszik,; Writing by Arno Schuetze; Editing by Alexandra Hudson

Netflix pilots $4 mobile plan in India to woo users

MUMBAI (Reuters) – Netflix Inc is testing a 250 rupee ($3.63) monthly subscription for mobile devices in India, the video streaming giant said, aiming to boost its presence in a price-sensitive market where data consumption on smartphones is surging.

FILE PHOTO: The Netflix logo is seen on their office in Hollywood, Los Angeles, California, U.S. July 16, 2018. REUTERS/Lucy Nicholson/File Photo

California-based Netflix currently offers three monthly plans in India, ranging from 500 rupees to 800 rupees, but those are still expensive compared with similar offerings from rivals.

Amazon’s Prime service, which offers video streaming, music and faster shipping of purchases, is priced at 999 rupees a year while local rival Hotstar has a free service as well paid plans starting at 365 rupees a year.

Netflix’s test plan at 250 rupees a month gives users access to standard definition video on smartphones and tablets, a company spokesman said.

“We will be testing different options in select countries where members can, for example, watch Netflix on their mobile device for a lower price and subscribe in shorter increments of time,” he added.

Netflix’s Indian roster includes blockbuster originals such as “Sacred Games”, global superhits such as “Narcos” as well as Indian cinema. However, its premium pricing is seen by critics as a stumbling block to bulking up its Indian user base.

Chief Executive Reed Hastings told Reuters late last year that Netflix had no plans for cheaper prices in India, where it aims to win its next 100 million subscribers.

The company emphasized on Tuesday that the new plan is a test and the company might not roll out these specific plans beyond the tests.

Netflix’s strategy to launch the test for mobile devices in India comes against the backdrop of rising demand for smartphones in the world’s second-biggest mobile phone market with more than 1.1 billion wireless connections.

Aspirational buyers looking for bigger screens and better user experience are likely to spend more on their second or third smartphones, pushing up the average selling price by 18 percent from last year to $190, said Tarun Pathak of technology researcher Counterpoint.

Reporting by Sankalp Phartiyal; Editing by David Goodman

Bahrain to use Huawei in 5G rollout despite U.S. warnings

DUBAI (Reuters) – Bahrain, headquarters of the U.S. Navy’s Fifth Fleet, plans to roll out a commercial 5G mobile network by June, partly using Huawei technology despite the United States’ concerns the Chinese telecom giant’s equipment could be used for spying.

FILE PHOTO: Logos of Huawei are pictured outside its shop in Beijing, China, February 28, 2019. REUTERS/Jason Lee/File Photo

Washington has warned countries against using Chinese technology, saying Huawei could be used by Beijing to spy on the West. China has rejected the accusations.

VIVA Bahrain, a subsidiary of Saudi Arabian state-controlled telecom STC, last month signed an agreement to use Huawei products in its 5G network, one of several Gulf telecoms firms working with the Chinese company.

“We have no concern at this stage as long as this technology is meeting our standards,” Bahrain’s Telecommunications Minister Kamal bin Ahmed Mohammed told Reuters on Tuesday when asked about U.S. concerns over Huawei technology.

The U.S. embassy in Bahrain did not immediately respond to a request for comment.

The U.S. Fifth Fleet uses its base in Bahrain, a Western-allied island state off the Saudi coast, to patrol several important shipping lanes, including near Iran.

Bahrain expects to be one of the first countries to make 5G available nationwide, Mohammed said, although he cautioned it would depend on handset and equipment availability.

Early movers like the United States, China, Japan and South Korea are just starting to roll out their 5G networks, but other regions, such as Europe, still years away and the first 5G phones are only likely to be released in the second half of this year.

Bahrain’s state controlled operator Batelco is working with Sweden’s Ericsson on its 5G network, while the country’s third telecom Zain Bahrain is yet to announce a technology provider.

No foreign company is restricted by the government from providing equipment for Bahrain’s 5G network, Mohammed said, adding that the mobile operators chose who they worked with.

Australia and New Zealand have stopped operators using Huawei equipment in their networks but the European Union is expected to ignore U.S. calls to ban the Chinese company, instead urging countries to share more data to tackle cybersecurity risks related to 5G networks.

Mohammed said the rollout of the 5G network was an “important milestone” for Bahrain, which is hoping investments in technology will help spur the economy which was hit hard by the drop in oil prices.

“It is something we are proud to have,” he said.

Reporting by Alexander Cornwell; Editing by Kirsten Donovan

The Very Mathematical History of a Perfect Color Combination

A couple of years ago, I fell in love with a color scheme: off-white text accented with a buttery yellow-orange and a neutral blue against a deep gray, the “color of television, tuned to a dead channel,” to borrow a phrase from Neuromancer author William Gibson. The colors were part of a theme called “Solarized Dark” for the popular MacOS code editor TextMate. To be honest, I didn’t think much of Solarized at first. But I soon found that I couldn’t work with any other color scheme. Staring at screens all day can make you particular about fonts and colors.

It turns out I’m not alone. I’m not a coder by trade, but I like to use code editors for writing and organizing notes. While hunting for tools after switching from a Mac to Windows, I started to see Solarized Dark and its sibling Solarized Light, which uses the same 16-color palette, practically everywhere I looked. It’s hard to say how many programmers use it. The design is free and open source, so there’s no tally of purchases. It’s available for every major code editor, and many other programming tools. Microsoft even bundled it with its popular code editor VS Code. Solarized has a loyal following.

“If I bring up a terminal window that doesn’t have Solarized, I feel out of place, I don’t feel at home,” says Zachery Bir, a Richmond, Virginia, programmer and artist who has been using Solarized since shortly after it was released in 2011. Bir likes Solarized so much he uses it as the color scheme for his computer-generated art. “I didn’t trust myself to come up with a palette that was balanced and looked good both in a dark and light medium,” he says.

The Solarized color scheme is no accident. It reflects the obsessive attention to detail of its creator, Ethan Schoonover. “I didn’t release it until I was 1,000 percent sure I loved all the colors and they were all dialed in mathematically,” Schoonover says. “I had multiple monitors, some were color calibrated, others were deliberately messed up. Sometimes I showed my wife, who thought I was a little nuts.”

Too Much Contrast

Schoonover was working as a designer and programmer in Seattle when he started work on Solarized in 2010. He’d recently switched operating systems was disappointed in the color schemes available for the tools he used. Many applications offered only a simple white-on-black scheme that harkens back to old-school text-based computer terminals. But Schoonover found these throwback color schemes much harsher than the retro displays they tried to emulate. That’s because the backgrounds displayed on old 1980s monitors weren’t truly black, Schoonover says. “They had less contrast.” Today’s LCD’s, on the other hand, are capable of displaying much darker, and much brighter, colors.

The optimal amount of contrast for text on a screen is controversial; many people prefer high-contrast themes. But contrast wasn’t Schoonover’s only concern. He found most low-contrast color schemes lacking as well. Even the best-designed themes tended to use at least one color that appeared distractingly brighter than others. That’s because the apparent brightness of a color varies depending on its background. In other words, a specific shade of blue will appear more or less bright, depending on the surrounding colors.

This phenomenon, known as the Helmholtz–Kohlrausch effect, is particularly aggravating for programmers because coding tools use color to distinguish different parts of code. In the code for a web page in a typical text editor using the Solarized Dark theme, for example, web links appear in green; the syntax for formatting, such as adding italics, is blue, and comments that developers write for themselves are gray. Ideally, the colors should help tell these elements apart, but no single element should stand out more than others.

Schoonover set out to find a set of colors that would not only look good together, but have the same apparent brightness. That task was made more difficult because he wanted to use the same palette in both a light and a dark theme. Hence the need for all the monitors and testing.

Examples of the Solarized Dark (left) and Solarized Light (right) themes displaying HTML code in the code editor Vim.

Ethan Schoonover

Ethan Schoonover

Schoonover talks a lot about the mathematical nature of his color selections, but he picked the starting colors, a blue and a yellow, for very personal reasons. The blue reminds him of his long standing thalassophobia, the fear of very deep water. And though he says he doesn’t otherwise experience synesthesia—such as hearing colors or tasting words—the yellow invokes tastes and smells he associates with his childhood. “My parents are artists, I’m comfortable picking things for obscure reasons,” he says.

With those starting points, Schoonover sought out other colors that provided just enough—but not too much—contrast between elements, and that maintained the same level of contrast in the light and dark versions. The result is a palette of just 16 colors that retain the same relationships even when inverted. “I suppose it’s a little like composing music with only a limited number of notes,” Schoonover says. “There can be something sparse and beautiful about it.”

An Open Source Program Takes Off

Schoonover released Solarized for free in April 2011 on GitHub, a code hosting platform and collaboration service. He says he never intended to commercialize it. “It would kill something special about it, taint it,” he says. “I believe in open source software, I believe in giving something special to the world that anyone can use.”

Although he’d tested the color scheme in a variety of applications, Schoonover initially only released themes for a few tools he used in his own work, like the code editor Vim and the text-based email client Mutt. He announced the release of Solarized on the Vim mailing list; soon after, the project hit the front page of the online community Hacker News. It was an immediate hit with programmers, who soon went to work adapting it to other programming tools beyond those Schoonover initially supported. In 2013, Solarized Dark appeared on the monitors of developers in a Facebook commercial—watch for those dark rectangles on the screens and notice the faintly colored lines that cross them.

Solarized is slowly starting to find its way into applications for non-geeks. Ulysses, a writing application for MacOS, includes Solarized themes as an option. The color scheme was used for many of the graphics in the video game N++ in 2014. The note-taking app MicroPad even advertises Solarized as a feature on its website. “Solarized Dark for MicroPad is especially useful for late-night studying, which I do more often than I would like to admit,” says MicroPad creator Nick Webster, a computer science student at Victoria University in Wellington, New Zealand.

But it still hasn’t really crossed over to the mainstream as a color scheme for, say, a major web application or software suite. “When Apple introduced dark mode for MacOS, I thought it was cool,” says Bir, the Virginia programmer and artist. “But I wish it was Solarized.”

With more applications, like Google Chrome, Facebook Messenger, and Slack, releasing dark mode themes, though, Solarized just might have its day in the sun.


More Great WIRED Stories

AI cloudops is coming, whether you like it or not

As someone who’s worked with AI for the last 30 years (yes, it was a thing 30 years ago), I’ve often thought of its capabilities were overrated and used for the wrong things in many cases. Now that it’s cheap thanks to cloud computing, and much more effective thanks to the pace of innovation, AI as a solution is coming up again, including the use in cloud operations.

The idea is to replace people with AI to be both proactive and reactive to cloud operational issues such as outages, resource governance, security attacks, and performance. Cloudops involves largely repeatable problems, right?

There are of course some upsides and some downsides to this. Moreover, although the use of AI in cloud operations maybe a foregone conclusion, there will still be a learning curve that is required. As long as you understand that and know what to expect in terms of ROI for both the short term and long term, I’m okay with anything that that makes cloud operations more effective.

So, let’s look at the pros and cons.

The pros of AI for cloudops

The pros are that you can have a 7/24/365 monitoring and management program on the cheap. If you believe operational staff is expensive, try hiring them for shift work. AI-based monitoring and management systems never sleep, never take time off, and never ask for a raise. Once they are up and running, they cost almost nothing beyond their license fees and infrastructure costs. And they are self-learning at the same time; in other words, the more they run, the better that they get at the job.

Another pro is that these systems get smarter every day and share a common brain. People get smarter with experience as well, but they don’t do a good job sharing their experiences with others. People also retire and quit, with the knowledge and experience walking out the door with them.

The cons of AI for cloudops

One con is that the cost of rolling out these systems is high, even in the cloud. Vendors that have married AI and operational tools are going to charge a premium to get them up and running and in production. While the prices are all over the place, count on paying 50 percent more than for traditional tools, including consulting services for the first year or so to get the tools learning correctly.

Another con is that operations people don’t seem to like them no matter how well they perform. The number of passive-aggressive actions that I’ve seen over the years from people pushing back on AI-enabled operations tools has been huge.

They view this technology as not to be trusted, plus the fact that AI some day may displace their jobs does not make things better. Organizations that implement these tools need to have change agents, plus an understanding about the human factors with this technology.

Is the future AI-enabled cloud-operations tools? I don’t see how it won’t be. The pros will get better, and the cons will begin to diminish, like any other rollout of new technology. Hopefully, our new AI operations overlords will have mercy on us in a few years.

A Cab’s-Eye View of How Peloton’s Trucks ‘Talk’ to Each Other

Techno-optimist prognosticators will tell you that driverless trucks are just around the corner. They will also gently tell you—always gently—that yes, truck driving, a job that nearly 3.7 million Americans perform today is perhaps on the brink of extinction. At the very least, on the brink of uncomfortable change.

A startup called Peloton Technology sees the future a bit differently. Based in Mountain View, California, the eight-year-old company has a plan to broadly commercialize a partially automated truck technology called platooning. It would still depend on drivers sitting in front of a steering wheel but it would be more fuel efficient and, hopefully, safer than truck-based transportation today.

The company employs ten professional truck drivers to help refine its tech, and I’m about to meet two of them out on Peloton’s test track in California’s Central Valley. Michael Perkins is tall, thin, and has been driving very big trucks for about 20 years. Jake Gregory is shorter and picked up truck driving in college, before taking a detour to the FBI.

We hit the highway first, because the rain has suddenly cleared. (Here’s an unfortunate reality about Peloton’s driver assistance tech: It doesn’t work great in the rain. Or snow. It’s a safety issue. More on that later.) Out on Interstate-5, Perkins’ long, white semitrailer cruises along in front of me. I’m on board the second, identical truck behind it, with Gregory behind the wheel. A small screen mounted on Gregory’s dashboard shows a camera view of what’s happening in front of Perkins’ rig. It’s like their trucks are connected. Which, in fact, they are about to be.

Peloton

Perkins radios in that he’s ready to go; Gregory says he is too. Inside the two truck cabs, each driver hits a button. Three ascending tones—la, la, la—means Peloton’s automated system has authorized the trucks to platoon on this stretch of highway. A dedicated short range communications (DSRC) connection is now established between the two vehicles. It’s like WiFi but faster and easier to secure. Now, whatever the front truck does, the back truck in will near-simultaneously “know”—and react accordingly.

Then Gregory speeds up, pulling his truck up so it’s tailgating about 70 feet from the leader. Sounds risky! But right now, the two trucks are platooning. Ours is on a kind of hopped-up cruise control, which means Gregory’s feet aren’t actually controlling the brakes or accelerator. At the same time, Gregory maintains control of his steering wheel. If Perkins were to brake, hard, Gregory’s truck would, too, faster than a human could. The robots have taken over. Kind of? Not really? More like, they’re collaborating, with some human oversight.

Peloton’s name, a reference to bicycle racing, helps explain how this platooning works. Just as the riders in the peloton, or main group of racing cyclists, preserve energy by drafting off of those around them, the following trucks in the truck platoon reduce their aerodynamic drag by drafting off the ones in front of it. The lead truck, meanwhile, get a little push. This saves fuel, according to Pelton—up to 10 percent for the following car and 4.5 percent for the first one, depending on the road and weather conditions and the following distance. It might also prevent crashes, since this tech has much faster reaction times (about 30 milliseconds) than puny humans (about 1 to 1.5 seconds).

Other companies in Europe, China, Japan, and Singapore are seriously experimenting with truck platooning. The American military has hosted platooning demonstrations. Just this week, the US Department of Transportation gave out $1.5 million in grants to universities studying the tech. And Peloton has tested in a bunch of US states: Arizona, California, Michigan, Florida, and Texas, where Peloton has immediate plans to run the majority of its routes.

Right now, the company says it does have paying customers, though it won’t reveal their names until later this year. According to Josh Switkes, the company’s CEO, some pair of US truck drivers are running a route while platooning on a Peloton-enabled truck every day.

And testing continues, on the software in its office, on its test track, and on actual highways, where it confirms the technology’s reliability. “The highway or field is not for testing,” Switkes says. “The goal of testing is to find failures, and you don’t want those failures to be on public roads.” In a report released today, the company lays out this approach to safety for regulators and interested industry parties alike. It borrow from automotive processes more than Silicon Valley-style software ones, amounting to something like, easy does it.

It turns out, the linking-up move Perkins and Gregory just performed on I-5 is one of the most safety-critical parts of truck platooning, says Switkes. The moment when the following truck has to move faster than the one in front of it is the most dangerous part.

To make sure drivers like Perkins and Gregory don’t crash into each other, or anyone else, Peloton needs to make sure that the platooning drivers know how the tech works. (Right now, the company’s driver training process takes about a half a day.) It also needs to understand exactly how heavy the trucks are when they start platooning, how their brakes are working, and how their tires function. For this reason, the company says, it has carved out partnerships with its suppliers, which means its trucks are built from the ground up with platooning in mind.

This is also why Peloton doesn’t platoon in the rain right now, or in the snow: The company can’t yet gauge exactly how tires deteriorate over time, which means it can’t quite predict how they’ll react in a hard-braking situation. Worn tires might slide in the moisture, leading to a domino chain of truck crashes. So no platooning in the Midwest in the winter, or anywhere during a rainy spring. “On certain routes, it’s a significant limitation,” says Switkes. “But we’re erring on the side of safety.”

And if that seems a little dull, Switkes would tell you that’s the point. His favorite word is “pragmatic,” and he doesn’t believe driverless trucks will prowl the highways any time soon. The technology is too complicated, he argues, and developers will have to go through years of safety testing before they’re ready for the roads—and before the public feels safe riding in their own bitty cars around 50,000-pound robot trucks. So Peloton is going all in on making human-based driving both safer and more efficient. With a bit of tech boost.

Not all manufacturers agree: In January, Daimler announced it would stop its platooning development to focus on autonomous trucking. Tests showed that “fuel savings, even in perfect platooning conditions, are less than expected,” the German company wrote in a press release. “At least for U.S. long-distance applications, analysis currently shows no business case for customers driving platoons with new, highly aerodynamic trucks.”

Platooning advocates disagree, but even the most supportive believe finding a market for this trucker assistance isn’t simple. Steven Shladover is researcher with the California Partners for Advanced Transportation Technology program at UC Berkeley. He has studied platooning for two decades, and points out that the truck industry would need to execute a fair bit of choreography to pull off platooning. Fleet operators would have to coordinate deliveries, matching up trucks heading in the same direction at the same time. “Does the truck industry see enough of a benefit in platooning to fit it into their operational strategies?” he says.

While everyone in trucking waits to find out, Perkins and Gregory head back to Peloton’s test track and proceed to show off a few, freakier moves: some hard braking, some driving side-by-side to prove that the trucks can still “talk” to each other in that position. At one point, another company employee in a white Toyota Tundra cuts into the 55-foot space between the two trucks, and they smoothly part to make room for him. Maybe platooning will improve life for truckers—too bad it can’t fix the problem of everyday reckless drivers, too.


More Great WIRED Stories

Instagram back up after several hours; Facebook still down for some

(Reuters) – Instagram is back up after suffering a partial outage for over several hours, the photo-sharing social network platform said in a tweet, but its parent Facebook Inc’s app still seemed to be down for some users across the globe.

FILE PHOTO: Silhouettes of mobile users are seen next to a screen projection of Facebook logo in this picture illustration taken March 28, 2018. REUTERS/Dado Ruvic/File Photo

Certain users around the world were facing trouble in accessing widely used Instagram, Whatsapp and Facebook apps earlier on Wednesday, in one of the longest outages faced by the company in the recent past.

“Anddddd… we’re back,” Instagram tweeted here along with GIF image of Oprah Winfrey screaming in excitement. Facebook did not provide an update.

Social media users in parts of United States, Japan and some parts Europe were affected by the outage, according to DownDetector’s live outage map here

Facebook users, including brand marketers, expressed their outrage on Twitter with the #facebookdown hashtag.

“Ya’ll, I haven’t gotten my daily dosage of dank memes and I think that’s why I’m cranky. #FacebookDown,” a user Mayra Mesina tweeted. bit.ly/2TDCYDK

The Menlo Park, California-based company, which gets a vast majority of its revenue from advertising, told Bloomberg that it was still investigating the overall impact “including the possibility of refunds for advertisers.”

A Facebook spokesman confirmed the partial outage, but did not provide an update. The social networking site is having issues since over 12 hours, according to its developer’s page.

Facebook took to Twitter to inform users that it was working to resolve the issue as soon as possible and confirmed that the matter was not related to a distributed denial of service (DDoS)

attack.

In a DDoS attack, hackers use computer networks they control to send such a large number of requests for information from websites that servers that host them can no longer handle the traffic and the sites become unreachable.

Reporting by Mekhla Raina in Bengaluru; Editing by Gopakumar Warrier and Rashmi Aich

How the FAA Decides When to Ground a Jet Like Boeing’s 737 MAX 8

When an Ethiopian Airlines Boeing 737 MAX 8 jet crashed shortly after takeoff from Addis Ababa on Sunday morning, killing all 157 people aboard, observers quickly noted that the circumstances resembled those of another flight. In October, Lion Air Flight 610 crashed into the Java Sea, killing all 181 passengers and eight crew. Both flights plummeted a few minutes after takeoff, in good weather. And both were on 737 MAX 8 jets, the plane Boeing started delivering in 2017 to replace the outgoing 737 as the workhorse of the skies. Since 2017, Boeing has delivered 387 MAX 8s and 9s. It has taken orders for 4,400 more, from more than 100 customers.

As of Tuesday evening, various foreign aviation regulators and airlines had decided that after the two crashes, the plane shouldn’t be in the air. Officials in the European Union, China, Indonesia, Singapore, Australia, and the United Arab Emirates have all grounded the planes. Of the 59 operators that fly the new 737, at least 30 have parked it.

In the US, though, Boeing’s plane is free to fly. American Airlines, Southwest Airlines, and United Airlines are still putting their 737 MAX jets—74 in total—in the air. (So is Air Canada.) And the Federal Aviation Administration—the agency that oversees American airspace—says that’s just fine.

Which might seem strange, since the FAA is notoriously safety-conscious. Planes in search of an airworthiness certificate must meet stringent standards; the certification process usually takes years. And it gets results: Just one person has died in American airspace on a commercial airplane since 2009. But, it seems, the agency has not yet found reason to ground the new 737.

In a statement Tuesday, acting FAA administrator Daniel Elwell said the agency is looking at all the available data from 737 operators around the world, and that the review “thus far shows no systematic performance issues and provides no basis to order grounding aircraft.” Elwell said the FAA “would take immediate appropriate action” should such problems be identified. The FAA and the National Transportation Safety Board both have teams at the crash site outside Addis Ababa to investigate and collect data.

The agency did note in a directive published Monday that it would probably mandate flight control system enhancements that Boeing is already working on, come April. And after the Lion Air crash, the FAA made a Boeing safety warning mandatory for US airlines.

“We have full confidence in the safety of the 737 MAX,” Boeing said in its own statement Tuesday. “Based on the information currently available, we do not have any basis to issue new guidance to operators.”

A number of senators, including Ted Cruz of Texas, Elizabeth Warren of Massachusetts, and Dianne Feinstein of California, have called for the US to ground the aircraft. But it’s the FAA chief who has final say. (Elwell has been the acting administrator since January 2018, though Politico reports that the Trump Administration is close to nominating Delta Air Lines executive Steve Dickson as administrator.) He doesn’t make that decision alone, says Clint Balog, a flight test pilot and human factors expert with the College of Aeronautics at Embry-Riddle University. Any grounding goes through a “semi-formal” process, full of discussions with experts on the specific aircraft and crash situation, both in- and outside the federal government.

“The FAA looks at all of this information and decides, ‘OK, if it’s just likely that there’s a significant problem here, it doesn’t matter what the cost to the traveling public is—we have to put safety first and ground this aircraft,’” Balog says. “However, if they look and say, ‘Well, jeez, grounding this aircraft is going to be a monumental cost to the world and we simply don’t have enough information to know what the risk really is with this aircraft, do we really want to ground it at this point in time?’”

The FAA has grounded aircraft before. In 1979, the FAA grounded all McDonnell Douglas DC-10s (and forbid the aircraft from US airspace) after a crash in Chicago killed 273 people. An investigation found the problem was maintenance issues, not the aircraft design, the FAA lifted the prohibition just over a month later.

In early 2013, the FAA grounded Boeing’s 787 Dreamliner, after two lithium ion-battery related fires in the aircraft. “We are issuing this [directive] because we evaluated all the relevant information and determined the unsafe condition described previously is likely to exist or develop in other products of the same type design,” the FAA wrote in its emergency airworthiness directive. It didn’t let the jet take to the sky again until Boeing found and corrected its design issues. (That happened in April.)

So far, though, we have little concrete information on whatever might be going on with the 737 MAX. The investigation into the Ethiopia crash is in its earliest stages. Indonesia’s civil aviation authority has released a preliminary report on the Lion Air crash, but has not issued any findings on what caused it.

Based on its directives, the FAA hasn’t “seen any red flags that are significant enough” to ground the aircraft, Balog says. So he’d have no problem getting on a 737 MAX-8. “More importantly, I would have no problem having my family get on a 737 MAX-8 at this point.”


More Great WIRED Stories

Elon Musk Says Tweeting Is Free Speech in His SEC Battle

Elon Musk will not go quietly. On Monday night, lawyers representing the Tesla CEO submitted a filing to a federal judge in New York arguing that she should deny the Securities and Exchange Commission’s request to hold Musk in contempt of court for—what else?—a tweet. Musk’s legal team argued the SEC overreached in its request, and claimed the agency is trying to violate his First Amendment right to free speech.

If the judge, Alison Nathan of the Southern District Court of New York, does hold Musk in contempt of court, she would decide the penalty. “If the SEC prevails, there is a good likelihood that the District Court will fine Mr. Musk and that it will put him on a short leash, with a strong warning that further violations could result in Mr. Musk being banned for some period of time as an officer or director of a public company,” Peter Haveles, a trial lawyer with the law firm Pepper Hamilton, told WIRED last month.

This latest chapter in Musk’s ongoing legal spat with the SEC dates back to the evening of February 19, 7:15 pm Eastern Time to be exact, when Musk wrote on Twitter, “Tesla made 0 cars in 2011, but will make around 500k in 2019.” About four and a half hours later—at 11:41 pm ET—Musk corrected himself, tweeting, “Meant to say annualized production rate at the end of 2019 probably around 500k, i.e. 10k cars/week. Deliveries for the year still estimated to be around 400k.”

Musk is the head of a publicly traded company, so making a mistake about his business on Twitter—which investors treat as a valid source of news like any other—is already less than ideal. But Musk and Tesla also reached a settlement with the SEC in September over another tweet containing misinformation about the electric carmarker’s operations. That was after Musk tweeted that he planned on taking Tesla private, and that he had the “funding secured.” He soon revealed he did not have that funding secured, and Tesla announced it would stay public.

In the ensuing deal with the SEC, Musk gave up his role as Tesla’s chairman for at least three years. He and Tesla each paid a $20 million fine. And Musk and Tesla agreed that the CEO’s tweets about the carmaker would be truthful, and reviewed by a team of Tesla lawyers before sending. According to the filing, Tesla’s general counsel and an assigned “disclosure counsel” are in charge of approving Musk’s Tesla tweets. The lawyers write that “the disclosure counsel and other members of Tesla’s legal department have reviewed the updated controls and procedures with Musk on multiple occasions.”

In December, Musk said on CBS’s 60 Minutes that he does not respect the SEC, and that the only tweets of his that require pre-approval are those that can affect Tesla’s stock price. Asked how Tesla could know which tweets would do that, Musk said, “Well, I guess we might make some mistakes. Who knows?” The SEC cited that interview in its motion for a contempt of court charge, writing that “Musk has not made a diligent or good faith effort to comply” with the terms of his settlement.

Now, though, Musk and the SEC are debating what that “pre-approval” actually means. Tesla’s lawyers say nobody pre-approved the tweet in question, but that it shouldn’t matter, because it had already made public the information about those production numbers: in an earnings call, in end-of-year financial results, and in an SEC filing submitted on the day Musk sent out the tweets in question. Musk did not receive pre-approval before sending that tweet because it “was simply Musk’s shorthand gloss on and entirely consistent with prior public disclosures detailing Tesla’s anticipated production volume,” according to the filing.

Moreover, the Musk team argues, the SEC’s attempt to limit Musk’s tweeting is a violation of his First Amendment rights to free speech.

The Musk legal team also argues that the CEO has really worked very hard since the SEC settlement to be careful about his tweeting behavior. It wrote that Musk’s less frequent tweeting about Tesla “is a reflection of his commitment to adhering the Order and avoiding unnecessary disputes with the SEC.” In fact, it says the correction tweet, the one sent four-and-a-half hours later, “is precisely the kind of diligence that one would expect from someone who is endeavoring to comply with the Order.”


More Great WIRED Stories

23andMe’s New Diabetes Test Has Experts Asking Who It’s For

On Sunday, the DNA testing company 23andMe revealed a new genetic analysis that it says will tell its customers if they have an elevated risk of developing the most common, and preventable, form of diabetes. The report—which has not been cleared by the FDA and is not intended to diagnose type 2 diabetes—arrives as the disease is becoming an intractable public health crisis in the US. One in four healthcare dollars goes to treating diabetes and its related complications. The situation is especially dire for African Americans and Native American populations, where obesity is rampant and one out of every seven or eight people has the disease.

Which is why it’s a little disappointing that 23andMe’s test is tuned to be most useful for skinny white people.

Unlike 23andMe’s other health reports, which inspect one or two genes for mutations with big, well-understood effects, its new diabetes test uses something called a polygenic risk score. It’s calculated by summing together each of the small risk contributions made at thousands of locations across a person’s genome. Alone, each one might increase your chances of getting a disease about as much as walking through a body scanner at the airport—but taken together, or in certain combinations, the risk can start to add up.

The algorithms that calculate polygenic risk scores aren’t new—they were largely pioneered in the mid-2000s. What’s new is the ability to derive them using huge genetic databases like 23andMe’s. Those databases, however, are overwhelmingly white. Because polygenic scores perform best for people with the same ethnic background as the DNA data used to train the algorithms, 23andMe’s new diabetes test isn’t as accurate for folks of non-European ancestry. It performs especially poorly for black Americans—barely better than a coin flip.

It’s a problem the company’s own vice president of research (a Chinese woman married to a Mexican man) recently wrote about, in a column for Stat. Through research collaborations with academics, the company is trying to fill out those sparse parts of its database, but the process is slow.

On its new diabetes test, the company used a machine learning trick called Platt scaling to recalibrate the polygenic score it had created in a European cohort for its other ancestry groups. “It’s a common shortcut, but it smashes down the level of risk you can convey to people,” says Ali Torkamani, a geneticist at the Scripps Research Translational Institute who studies polygenic risk scores and health outcomes. For the Hispanic, east and south Asian groups, he says it shouldn’t make too much of a difference. “For users of African American descent that score is not at all relevant.” Would such a shortcut be safe for a clinical test—one that people use to make decisions about their health? “No, not for an African American individual,” says Torkamani.

Because of these limitations, many of the polygenic risk tests already on the market are race-restricted. A test for Alzheimer’s developed by scientists at UC San Diego, for example, carries language advising you to only use it if you’re of European ancestry. In 2017, Salt Lake City-based Myriad Genetics added a polygenic “riskScore” to its physician-ordered breast cancer test, which more than 200,000 women have since taken, all of them of European ancestry.

Jerry Lanchbury, Myriad’s chief medical officer, says that adding the polygenic risk score sharpens the test’s performance for high-risk women. “If you’re in that high-risk category you may qualify for enhanced screenings or even surgical preventative offerings, so it’s a big deal,” says Lanchbury. To make it available to women of other ethnicities, the company has recruited 14,000 women of Hispanic descent and shown it can generate a risk score for that group, which it hopes to offer later this year. Recruiting is underway for an African American cohort as well.

The Bay Area’s Color Genomics also recently announced a plan to enroll 100,000 volunteers from historically underrepresented groups to better assess the risk of heart attack.

Starting Monday, qualified 23andMe customers will be able to access their polygenic risk score for diabetes—a single number that represents their chances of getting diabetes based on age, ethnicity, and DNA. It also determines if a user gets labeled as having a ‘typical’ or ‘increased’ likelihood of getting diabetes as compared to other users in the database. 23andMe drew that line where the amount of risk carried in a person’s DNA exceeded the diabetes risk associated with being overweight—the single biggest risk factor for the disease. The company expects that about 22 percent of 23andMe research participants—roughly one million current customers—will learn they have an increased likelihood for developing diabetes at some point in their lives. (The new number will only be available to customers who have compatible DNA chips—roughly those who joined in the last two years.)

If 23andMe’s customer base reflected the US population, two-thirds of them would be overweight or obese and already at an increased risk for diabetes. It should come as no surprise to them that they need to avoid sugary, processed foods and get regular exercise. That’s why Torkamani says the new genetic test will be most useful for the small subset of people who don’t have any clinical risk factors—they look lean and healthy, they exercise, they’re under the age of 45. “If they don’t know anything about how their blood glucose levels are responding to food, they may have no idea of an underlying issue that could devolve into type 2 diabetes later on,” says Torkamani.

The trouble is, if you already have a healthy lifestyle, there’s not much you can do with that information. It’s not like heart disease, where you can start taking cholesterol-lowering drugs. Torkamani says one thing would be to start getting regular blood tests. Among other things, 23andMe suggests you sign up for a digital coaching program through the company’s newest partner, Lark. In January, 23andMe customers got the option to integrate their genetic information with Lark’s AI-powered chatbots for an additional fee. One of those bots delivers a CDC-recognized diabetes prevention program, which some insurance companies cover.

Unlike its other genetic health risk tests, 23andMe developed its diabetes report under the FDA’s guidelines for low-risk general wellness devices, products that promote a healthy lifestyle—which can be beneficial for all people regardless of their genetic disposition. ”Like all of our reports, we hold this report to high scientific standards,” said a 23andMe spokesperson, who also noted that a type 2 diabetes test was one of the reports most often requested by customers. “It’s built using data from 2.5 million individuals, which we believe to be the largest cohort ever used to develop a genetic model for type 2 diabetes.”

Other experts express skepticism of such actions. “There’s so little value in these scores that I have no idea what people should do with it,” says Cecile Janssens, an epidemiologist at Emory University who studies how genomics enters health practice. “You might as well just look in a mirror, that’s as good a predictor for diabetes as all your genes put together.”

She started looking at polygenic risk scores in the early 2000s, right after the Human Genome Project wrapped up and before 23andMe was even a “Google for DNA” in Anne Wojcicki’s eye. But Janssens says she got bored with the field because there hadn’t been any real progress in close to a decade. Then all of a sudden, in the last two years polygenic risk scores started popping up again. This time, they included millions of variants with effects so tiny that earlier studies on only a few thousand people couldn’t detect them. “The only trouble is they don’t add anything to the predictions,” she says. “Diseases like type 2 diabetes don’t become more heritable just because we’ve got better technology.”

The resurgence of polygenic risk scores has reignited old party lines in the research community and drawn some new ones. Depending on one’s side, polygenic risk scores are either going to revolutionize complex disease prevention by creating more precise pools of risk (the Torkamani camp), or they’re mostly rubbish (Janssen). The scientific validity of polygenic risk scores is still up for debate, but with 23andMe bringing it back to the mainstream, the discussion is becoming more urgent. Because if there’s one thing researchers know, it’s that no amount of tiny type at the bottom of the page will dissuade people from making decisions on risk scores in their grasp, however uncertain they might be.


More Great WIRED Stories

Facebook Will Crack Down on Anti-Vaccine Content

As Clark County, Washington, combats an ongoing measles outbreak, Facebook announced Thursday that it’s diminishing the reach of anti-vaccine information on its platform. It will no longer allow it to be promoted through ads or recommendations, and will make it less prominent in search results. The social network will not take down anti-vaccine posts entirely, however. The company also said it was exploring ways to give users more context about vaccines from “expert organizations.”

The decision was widely anticipated: Facebook, along with YouTube and Amazon, has faced criticism from journalists and lawmakers in recent weeks for allowing vaccine misinformation to flourish on their sites. Facebook also told media outlets in February that it was looking into how it should address anti-vaccination content.

Last month, Adam Schiff, a Democratic representative from California, sent letters to the CEOs of YouTube and Facebook demanding they answer questions about the spread of anti-vaccine information on their company’s platforms. He followed up with a similar letter to Amazon CEO Jeff Bezos last week. On Wednesday, an 18-year-old from Ohio testified before the Senate that his mother primarily read misinformation about vaccines on Facebook and opted not to inoculate him. (A major study released Monday found no link between the MMR vaccine—which protects against measles, mumps, and rubella—and autism.)

In a blog post written by Monika Bickert, Facebook’s vice president of global policy management, Facebook said it will begin rejecting ads that include false information about vaccinations. The company also removed targeting categories such as “vaccine controversies” from its advertising tools. Last month, the Daily Beast reported that more than 150 anti-vaccine ads had been bought on Facebook, which often targeted women over 25. Some of the ads were shown to users “interested in pregnancy.” In total, they were viewed at least 1.6 million times. YouTube similarly announced last month that it would begin preventing ads from running on videos featuring anti-vaccine content.

Facebook will also reduce the ranking of pages and groups that spread misinformation about vaccines in search results and in its News Feed. In February, The Guardian found that anti-vaccination propaganda often ranked higher and outperformed accurate information from more reliable sources on Facebook.

The social network’s effort to fight vaccine disinformation extends to Instagram, where the company says it will stop recommending content that includes vaccine misinformation on the app’s Explore page. Instagram will also stop displaying vaccination misinformation in hashtag search results. It’s not clear how long these new controls will take to roll out: An Instagram search for #vaccine Thursday afternoon surfaced the hashtag #vaccineskill as the number one result, for instance. Last month, Pinterest received praise for its decision to stop displaying search results for vaccines entirely, even if they are medically accurate. (In 2017, Pinterest previously banned “anti-vaccination advice” from its platform.)

As The Atlantic has pointed out, the majority of anti-vaccination content on Facebook appears to originate from only a handful of fringe sources. It likely won’t require a herculean effort for Facebook to tackle this strain of misinformation. The question is why the company waited until it became the subject of media reports and criticism from lawmakers to finally act.

Facebook increased its efforts to fight false information more broadly on the platform in the wake of the 2016 presidential election, including with initiatives like third-party fact-checking. The company admits it won’t catch everything, and demonstrably fake stories still do go viral. While there is little public data about user behavior on Facebook, researchers have found signs that the reach of fake news declined between 2016 and 2018 midterm elections. (Though they also say there remains plenty to be concerned about when it comes to misinformation.)

It’s not yet clear whether the proliferation of anti-vaccination content online has led to a significant decrease in vaccination rates in the United States. Unscientific information about vaccines has been circulating on- and offline for well over a decade. But as Slate has pointed out, the number of children under 3 who have received their first dose of the MMR vaccination has remained steady for years, according to data from the Centers for Disease Control and Prevention. The World Health Organization named vaccine hesitancy one of its “ten threats to global health in 2019,” but cites “complacency and inconvenience in accessing vaccines” as two of the key reasons why people choose not to vaccinate, in addition to “lack of confidence.”

There’s still little doubt that social media platforms like Facebook, but also YouTube and Amazon, have indeed made anti-vaccination talking points more accessible to wider audiences. Its proponents were aided by recommendation and search ranking algorithms, which often promoted anti-vax content to the top of the pile. Facebook’s announcement today is further acknowledgment of its role in that ecosystem, and the idea that free speech is not the same as free reach.


More Great WIRED Stories

An Email Marketing Company Left 809 Million Records Exposed Online

By this point, you’ve hopefully gotten the message that your personal data can end up exposed in all sorts of unexpected internet backwaters. But increased awareness hasn’t slowed the problem. In fact, it’s only grown bigger—and more confounding.

Last week, security researchers Bob Diachenko and Vinny Troia discovered an unprotected, publicly accessible MongoDB database containing 150 gigabytes-worth of detailed, plaintext marketing data—including 763 million unique email addresses. The pair are going public with their findings today. The trove is not only massive but also unusual; it contains data about individual consumers as well as what appears to be “business intelligence data,” like employee and revenue figures from various companies. This diversity may stem from the information’s source. The database, owned by the “email validation” firm Verifications.io, was taken offline the same day Diachenko reported it to the company.

While you’ve likely never heard of them, validators play a crucial role in the email marketing industry. They don’t send out marketing emails on their own behalf, or facilitate automated mass email campaigns. Instead, they vet a customer’s mailing list to ensure that the email addresses in it are valid and won’t bounce back. Some email marketing firms offer this mechanism in-house. But fully verifying that an email address works involves sending a message to the address and confirming that it was delivered—essentially spamming people. That means evading protections of internet service providers and platforms like Gmail. (There are less invasive ways to validate email addresses, but they have a tradeoff of false positives.) Mainstream email marketing firms often outsource this work rather than take on the risk of having their infrastructure blacklisted by spam filters, or lowering their online reputation scores.

“Companies have email lists and want to start emailing them, but they’re not sure how valid they are,” says Troia, who founded the firm Night Lion Security. “So they go to a company that will essentially send out spam.” Troia speculates, but has not confirmed, that the database may be so large and varied because it comprises all of Verification.io’s customers’ data. WIRED was unable over the course of several days to contact the company or CEO Vlad Strelkov. On Monday, the entire Verifications.io website went offline and has not been restored since.

Record Setter

In general, the 809 million total records in the Verifications.io trove include standard information like names, email addresses, phone numbers, and physical addresses. But many also include things like gender, date of birth, personal mortgage amount, interest rate, Facebook, LinkedIn, and Instagram accounts associated with email addresses, and characterizations of people’s credit scores (like average, above average, and so on). Meanwhile, other records in the collection seem related to generating sales leads at businesses, including company names, annual revenue figures, fax numbers, company websites, and industry identifiers for categorizing companies called “SIC” and “NAIC” codes.

The data doesn’t contain Social Security numbers or credit card numbers, and the only passwords in the database are for Verifications.io’s own infrastructure. Overall, most of the data is publicly available from various sources, but when criminals can get their hands on troves of aggregated data, it makes it much easier for them to run new social engineering scams, or expand their target pool.

In the exposed database, the researchers also found some of what appear to be Verifications.io’s own internal tools like test email accounts, hundreds of SMTP (email sending) servers, the text of emails, anti-spam evasion infrastructure, keywords to avoid, and IP addresses to blacklist. Diachenko suggests that in the Verifications.io work flow, customers would upload an Excel spreadsheet listing the email addresses to validate, and then Verifications.io would run their tests and return lists of clean addresses and ones that bounced back. It’s possible, given the piecemeal nature of the data and evidence that it was imported from numerous different Excel files, that Verifications.io also retained some or all of the data it received from customers after concluding its email address checks.

The researchers validated samples of the data with companies listed as Verifications.io customers. Troia says his own information appears in the database. WIRED spoke to the proprietor of an email marketing firm who confirmed the validity of a segment of the data. WIRED also checked for four individuals, but did not find them listed. Diachenko and Troia also note that they have no way to know whether anyone discovered and downloaded the Verifications.io data while it was publicly available and fully exposed.

“I have no idea if anyone else accessed this besides us,” Troia says. “But it was definitely out there for anyone to grab.”

‘Another Day on the Internet’

Much remains unknown about the database and Verifications.io, because the company is difficult to track. When the researchers initially contacted the company through a messaging portal on its site to disclose the database exposure, someone responded with an unsigned note. “Thank you for reporting the issue. We appreciate you reaching out and informing us,” the reply said. “This is our company database built with public information, not client data. We were able to quickly secure the database. Goes to show, even with 12 years of experience you can’t let your guard down.”

Much of the data in the database is publicly available, though it’s not clear that all of it is. When the researchers asked in the portal for the name of the owner of the company and the legal name of the company, someone wrote back declining to answer.

It is also unclear where Verifications.io is based. Most of its materials list Boca Raton, Florida, but some of its web assets are registered in California and Delaware. The Verifications.io website lists addresses in Estonia, but some of those matched up with what appear to be a museum and a government building.

Security researcher Troy Hunt is adding the Verifications.io data to his service HaveIBeenPwned, which helps people check whether their data has been compromised in data exposures and breaches. He says that 35 percent of the trove’s 763 million email addresses are new to the HaveIBeenPwned database. The Verifications.io data dump is also the second-largest ever added to HaveIBeenPwned in terms of number of email addresses, after the 773 million in the repository known as Collection 1, which was added earlier this year. Hunt says some of his own information is included in the Verifications.io exposure.

“The main takeaway for me is that this is just another case where someone has my data, and hundreds of millions of other people’s data, and I’ve absolutely no idea how they got it,” Hunt says. “I’d never heard of the company until now and I certainly can’t ever recall consenting to their use of my data. Of course, it’s entirely possible that buried in some other service’s terms and conditions it says they’re allowed to pass my data around in this fashion, but that’s not really consistent with my expectations of how my data should be used.”

As with recent data exposures from the business data aggregator Apollo and the marketing firm Exactis, there’s not a lot you can do to individually protect yourself when vast repositories of data compiled from both public and private sources leak. Check HaveIBeenPwned to see if your data was in the Verifications.io exposure, and continue your general vigilance about using strong, unique passwords, monitoring your financial statements, and giving out your Social Security number as infrequently as possible. But also know that none of those measures provide a full solution to this society-scale problem.

The disjointed nature of the exposed Verifications.io data speaks to the chaotic state of the data industry overall. People’s personal information is shared by massive companies like Facebook, bought and sold by shady marketers, or stolen from data giants and doomed to circulate endlessly in the purgatory of criminal forums. The churn makes it difficult for consumers to control who has their data and where it ends up. As Hunt puts it, “Sadly, it’s just another day on the internet.”


More Great WIRED Stories

Amazon to close U.S. pop-up stores, focus on opening more book stores

FILE PHOTO: The logo of Amazon is seen at the company logistics centre in Boves, France, August 8, 2018. REUTERS/Pascal Rossignol

(Reuters) – Amazon.com Inc will close all of its U.S. pop-up stores and focus instead on opening more book stores, a company spokesperson said on Wednesday.

The company’s shares closed down 1.4 percent, while shares of bookseller Barnes & Noble Inc ended 8.9 percent lower.

Amazon’s 87 pop-up stores in the United States are expected to close by the end of April, the Wall Street Journal reported earlier on Wednesday, citing some of the employees at the stores.

The news underscores how the online retailer is still working out its brick-and-mortar strategy.

Pop-up stores for years helped Amazon showcase novel products like its voice-controlled Echo speakers, but the company is now able to market those products and more at its larger chain of Whole Foods stores, acquired in 2017, and cashierless Amazon Go stores, which opened to the public last year.

The online retail giant will also open more “4-star stores” – stores that sell items rated 4-stars or higher by Amazon customers, the spokesperson added.

“After much review, we came to the decision to discontinue our pop-up kiosk program, and are instead expanding Amazon Books and Amazon 4-star, where we provide a more comprehensive customer experience and broader selection.”

Reporting by Uday Sampath in Bengaluru; Editing by Maju Samuel

Britain's Hunt promises 'doctrine of deterrence' against cyberattacks on democracy

LONDON (Reuters) – British foreign minister Jeremy Hunt will set out on Thursday a “doctrine of deterrence”, including economic and diplomatic counter-measures, to prevent cyberattacks that threaten to turn elections into “tainted exercises”.

Britain’s Foreign Secretary Jeremy Hunt is seen outside of Downing Street in London, Britain, March 5, 2019. REUTERS/Peter Nicholls

Britain will try to prosecute those responsible for cyber crimes, part of a growing response by the West against countries that hope to influence elections through disinformation and voter manipulation, he will say in a speech in Glasgow.

“We will always seek to discover which state or other actor was behind any malign cyber activity, overcoming any efforts to conceal their tracks,” Hunt will say, according to pre-released extracts of his speech.

Western countries issued coordinated denunciations of Russia in October for running what they described as a global hacking campaign. Russia has denied the allegations.

In the United States, a federal special counsel is investigating Russian interference in the 2016 presidential election and possible collusion with Donald Trump’s campaign. Moscow has denied any meddling and the U.S. president has said there was no collusion.

Hunt will say there has been no evidence that foreign states have interfered with British votes but that unnamed hostile states are intent on using cyberspace to undermine Western democracies.

“Events have demonstrated how our adversaries regard free elections – and the very openness of a democratic system – as key vulnerabilities to be exploited … authoritarian regimes possess ways of undermining free societies that yesterday’s dictators would have envied,” he will say.

The British response could include the public naming and shaming of any perpetrator together with allies, exposing how the action was carried out and prosecuting those responsible to show they are not above the law.

Hunt will also say that Britain, as part of the European Union, agreed last year to impose sanctions to stiffen its response to cyberattacks and to rush through new curbs on online campaigning by political parties.

“After Brexit, the UK will be able to impose cyber-related sanctions on a national basis,” he will say.

Reporting by Elizabeth Piper; Editing by Frances Kerry

Uber not criminally liable in fatal 2018 Arizona self-driving crash: prosecutors

(Reuters) – Uber Technologies Inc is not criminally liable in a March 2018 crash in Tempe, Arizona, in which one of the company’s self-driving cars struck and killed a pedestrian, prosecutors said on Tuesday.

FILE PHOTO: National Transportation Safety Board (NTSB) investigators examine a self-driving Uber vehicle involved in a fatal accident in Tempe, Arizona, U.S., March 20, 2018. National Transportation Safety Board/Handout via REUTERS

The Yavapai County Attorney said in a letter made public that there was “no basis” for criminal liability for Uber, but that the conduct of the back-up driver, Rafael Vasquez, should be referred to the Tempe police for additional investigation.

Police said last year that Vasquez was streaming a television show on a phone until about the time of the crash and called the incident “entirely avoidable.”

An Uber spokeswoman declined to comment on the letter.

Vasquez could face charges of vehicular manslaughter, according to a police report last June. Vasquez has not previously commented and could not immediately be reached on Tuesday.

The Maricopa County Attorney, whose jurisdiction includes Tempe, referred the case last year to another office because of a conflict. In Tuesday’s letter Yavapai County Attorney Sheila Sullivan Polk said its investigation concluded that “the collision video, as it displays, likely does not accurately depict the events that occurred.”

The letter said an “expert analysis” is needed to “closely match what (and when) the person sitting in the driver’s seat of the vehicle would or should have seen that night given the vehicle’s speed, lighting conditions, and other relevant factors.”

The National Transportation Safety Board and National Highway Traffic Safety Administration are still investigating the fatal crash.

The Uber car was in autonomous mode at the time of the crash, but the company, like other self-driving car developers, requires a back-up driver inside to intervene when the autonomous system fails or a tricky driving situation occurs.

The Tempe police report said Vasquez repeatedly looked down and not at the road, glancing up a half second before the car hit Elaine Herzberg, 49, who was crossing the street at night.

Police obtained records from Hulu, an online service for streaming TV shows and movies, which showed Vasquez’s account was playing the TV talent show “The Voice” for about 42 minutes on the night of the crash, ending at 9:59 p.m., which “coincides with the approximate time of the collision,” the report said.

The Maricopa County Attorney’s Office did not immediately comment on Tuesday.

In December, Uber resumed limited self-driving car testing on public roads in Pittsburgh, nine months after it suspended the program following the Arizona crash.

The company is now testing with two employees in the front seat and more strictly monitor safety employees. The company also said last year it made improvements to the vehicles’ self-driving software.

In March 2018, authorities in Arizona suspended Uber’s ability to test its self-driving cars. Uber also voluntarily halted its entire autonomous car testing program.

The NTSB has said Uber had disabled a manufacturer-installed automatic emergency braking system in the 2017 Volvo XC90 while the car was under computer control in order to “reduce the potential for erratic vehicle behavior.”

Reporting by David Shepardson; Editing by Bill Berkrot and Grant McCool