June 8, 2018

I became a Ray Dalio fan earlier this year when I read his book Principles. I went on my Ray Dalio journey, read a bunch about him on the web, and watched some of his videos and interviews.

While on vacation last week, I watched his new 30-minute cartoon adventure Principles for Success. It’s spectacular and worth 30 minutes of your life to watch and ponder.

His Episode 4: The Abyss, made me think “just another one of those” (a Dalio construct that I’ve come to love.) My abyss happened between 2001 and 2003 and, while I wasn’t alone in my mistakes as Dalio was in his, I related deeply to it.

Also published on Medium.

EU competition commissioner Margrethe Vestager.Vimeo/Markus Böhnisch

Good morning! This is the tech news you need to know this Friday.

1. A Facebook bug accidentally set users' posts to be public by default, even if they were intended to be private. As many as 14 million users were affected by the bug and had their posts made public incorrectly.

2. Google has unveiled a set of ethical principles governing how it will use artificial intelligence (AI). The seven principles include making sure that AI is applied to applications that are socially beneficial, safe and won't create unfair bias.

3. Google's partnership with Chinese phone maker Huawei is under congressional scrutiny, according to The Washington Post. That comes after members criticised Facebook for its data-sharing agreements with Huawei and Chinese peer ZTE.

4. Google may face another billion-euro fine from the EU next month over an antitrust case involving Android. A fine would reflect growing political antipathy towards the big tech firms and could help trigger further regulatory action.

5. Facebook is looking to hire a pair of contractors as "news credibility specialists" to create a list of credible news organizations. That list could be used for various features on the site, from the newsfeed to its advertising system.

Advertisement

6. Google's pledge to quit doing military work involving its AI technology does not include its current job helping the Pentagon with drone surveillance. Diane Greene, the head of Google Cloud, wrote in a blog post on Thursday that Google will continue to honour the controversial Project Maven contract that it entered into last year.

7. Amazon became first tech company to win rights to show Premier League games in the UK. The US firm has acquired the rights to livestream 20 matches a season for three years, starting in August 2019.

8. Supposed images of the Google Pixel 3XL were leaked on Thursday. They support the rumors that the Pixel 3XL will have a notch, much like the iPhone X.

9. The world's largest producer of bitcoin mining chips, Bitmain Technologies, might be considering going public. Bitmain's founder, the cryptocurrency billionaire Jihan Wu, told Bloomberg that Bitmain was considering an initial public offering as it expands into producing hardware for artificial-intelligence computing.

10. Delivery robot startup Starship Technologies has raised $25 million and hired a new CEO, a former Airbnb executive. Lex Bayer headed up Airbnb for business but will take over the top job from Starship Technologies' cofounder and CTO Ahti Heinla.

Advertisement

Have an Amazon Alexa device? Now you can hear 10 Things in Tech each morning. Just search for "Business Insider" in your Alexa's flash briefing settings.

Pharmaceutical companies are slowly but surely turning their eyes toward Silicon Valley.

At a time when companies like Amazon and Apple are getting serious about entering the healthcare market, pharmaceutical firms in turn are considering ways they can learn from their tech counterparts.

But it has taken a long time to get there, Jay Bradner, the president of the Novartis Institutes for BioMedical Research, told Business Insider in an interview at a healthcare conference in Boston this past week.

"Healthcare is so technology forward, and biomedical discovery is so technology forward," he said. "Yet our field has been a late adopter to innovations that the computer-science world is sick of talking about." For example, he said, concepts like machine learning and artificial intelligence are standard for the tech world but new to biomedicine.

Advertisement

Novartis is taking a lot of that computing technology inherent to Silicon Valley and applying it to drug discovery.

"We like to think of ourselves as the lead turtle in the race of the turtles," Bradner said. About 4% of the 6,000 scientists working at Novartis' research engine are data scientists, and the company's biologists also have an understanding of computer science and coding.

"To interact with data today, the fundamental biologist needs to be comfortable 'in silico,'" i.e., through computers, he said. "It's too late for me to go back and learn Python, but every new graduate student from my lab will leave learning how to interact primarily with their data."

Ideally, the thinking goes, pharmaceutical researchers can use these skills and tools like machine learning to speed up drug development.

Virtual trials

One way Novartis put technology to use recently was in determining which trial the company wanted to pursue on an experimental drug.

Advertisement

Novartis was able to take data it had recorded from other clinical trials it had run and use that information to set up virtual trials. These trials looked at how experimental drugs might work. In the end, one of the trials seemed to show that its drug worked, while the other two didn't.

Running the trial virtually allowed Novartis to prioritize how to spend its time and resources.

"In the one case, we're more efficient because we didn't launch and resource two clinical trials, and in the other we can revisit that idea with a definitive trial," Bradner said.

Leaving Silicon Valley culture to the Valley

When it comes to Silicon Valley's culture, with its traditionally hyped-up ideas and "move fast and break things" pace, that's less likely to rub off on pharma.

"Silicon Valley seems to tolerate more hype than the biomedical research environments where I've worked," Bradner said. Still, it doesn't mean researchers within pharma don't have big visions. "It doesn't make this community that we're in any less ambitious or starry-eyed or dreamy," he said.

Advertisement

And while an idea for a new app in Silicon Valley can get off the ground and onto the market in months, the timeline to get a drug from discovery to approval can take upward of a decade and involves a lot of regulation.

As interest from technology companies to get into healthcare and biotech grows, pharmaceutical companies and tech giants are starting to become rivals. But Bradner said the two industries would ultimately be better off teaming up than finding themselves in competition.

"These are two communities that I find really fascinating," he said, "that stand to benefit more from working together than by pretending to have solved each other's problems."

Facebook is getting back into the business of having employees make decisions about news organizations and stories — including about their credibility — something company leaders including Mark Zuckerberg have previously renounced.

The company has open job listings for two "news credibility specialists," one of which would be required to be able to speak, read, and write fluently in Spanish. The jobs, which Facebook listed within the last few weeks, would be contract positions, rather than full-time, but would be based in Menlo Park, California, site of Facebook's headquarters.

"We're seeking individuals with a passion for journalism, who believe in Facebook's mission of making the world more connected," one of the two listings reads. It continues: "As a member of the team, you'll be tasked with developing a deep expertise in Facebook's News Credibility Program. You'll be conducting investigations against predefined policies."

Facebook would ask the specialists to help create a list of credible news organizations. That list could be used for various features on the site, from the newsfeed to its advertising system.

Advertisement

The move comes as Facebook tries to battle the spread of fake news and propaganda on its social network, and to shore up its damaged reputation. But in taking a bigger role appraising news organizations, Facebook risks opening itself up to charges of bias and censorship as it tries to make sense of an increasingly fragmented media landscape.

One example of how the credibility specialists might come into play involves Facebook's plans to ensure transparency with political ads. Facebook has said it will require political ads to disclose who paid for them, restricting ads that don't comply.

Since news organzations often pay to promote political articles on Facebook, the credibility specialist would be able to help Facebook identify those articles from standard political ads, to ensure that the news articles don't get caught in the net.

"We're working to effectively identify and differentiate news and news sources across our platform," company spokesman Adam Isserlis said.

Facebook had said it didn't want to judge news trustworthiness

At least on its face, the fact that the social networking giant is hiring "news credibility specialists" appears to contradict a policy it announced earlier this year. Facebook officials said they didn't want to be in the business of determining the credibility of news reports and outlets. Instead, the company said it would survey its users to figure out which sites and sources were trustworthy. It's also said that it would rely on third-party fact-checking organizations, including the Associated Press, to judge the credibility of articles.

Advertisement

"The hard question we've struggled with is how to decide what news sources are broadly trusted in a world with so much division," company CEO Mark Zuckerberg said in a post on his personal Facebook page in January. "We could try to make that decision ourselves, but that's not something we're comfortable with."

He continued: "We decided that having the community determine which sources are broadly trusted would be most objective."

The new news credibility specialists won't replace the third-party fact checkers or Facebook's surveys of users to determine the trustworthiness of sites, but will be an addition to them. But the lines between what they will do and what will be left to the survey of users could get blurry.

It makes sense for Facebook to have its own employees making decisions about the credibility of news organizations, said Daniel Kreiss, a professor of of political communications at the University of North Carolina at Chapel Hill. The fact is that Facebook makes editorial decisions all the time about what it will and won't allow on its site, and in keeping with the company's mission, it needs to have a role in determining what content on its site is legitimate, he said.

But there are plenty of questions to be raised about this effort, Kreiss said. It's unclear, for example, what kinds of qualifications Facebook will be looking for when trying to fill these roles; a journalist with 30 years of experience may have different ideas about what is a credible news organization than someone just out of college, he said.

Advertisement

It's also unclear what kind of criteria Facebook will use to determine what entities are news organizations and which are credible, he said. It's anyone guess whether Breitbart would qualify or whether a news organization run as a wing of a political organization would qualify.

"They're always playing this editorial role," Kreiss said. "My concern throughout has been that they haven't acknowledged they're playing this role, they aren't transparent about how they're making those decisions, and because of that, they aren't accountable for the decisions they do make."

Given Facebook's importance when it comes to distributing news and its role in the democratic process, it's going to be crucial when the company decides that entities are not a credible news organization that the social networking giant is transparent about its rationale and that it allows those entities to appeal the decisions, Kreiss said.

"There should be a way to ... hold Facebook accountable for the decisions they make," he said.

we might want to truncate and move up Weiland's comment. Maybe cut it down to just "They're giving us a lot of mixed signals" and put it right after tha part about the lines getting blurry

The company is sending "mixed signals"

Facebook has a responsibility to be transparent about how it plans to evaluate news organizations, said Morgan Weiland, an attorney and PhD candidate at Stanford whose research focuses on how the big tech platform companies are handling their role in distributing news.

Advertisement

"If they're going to build out a team like this, they need to be more explicit about how they understand their role or what kind of company they see themselves as," Weiland said.

She continued: "They're giving us a lot of mixed signals."

The process of determining the credibility and newsworthiness of articles and publications has long been a fraught issue for Facebook.

Two years ago, the company drew criticism after its employees reportedly downplayed stories about right-wing politicians and conservative topics in its trending news section. In response, the company fired its human editors, relying instead on algorithms to determine trending topics. But the automated editing system ended up promoting fake and offensive stories. Earlier this month, Facebook announced it would discontinue the trending news feature.

But the company's problems in figuring out the credibility of news go well beyond its trending news controversy.

Facebook has been trying to battle fake news

During the presidential election in 2016, Facebook's systems were hijacked by Russian-linked groups to spread false stories and other propaganda in an alleged attempt to influence the vote, a development the company was slow to acknowledge. Facebook has also been used by ultra nationalist groups in Myanmar to spread propaganda and hate speech that has helped incite violence against the country's Rohingya minority, according to the United Nations.

Advertisement

In an attempt to curtail the spread of fake news, Facebook early this year said it would de-emphasize posts from news outlets and other organizations in its news feed. Then, it said it would poll users to figure out which sites to trust. The algorithms underlying its news feed would use those determinations about trustworthiness when deciding which articles to highlight or to suppress.

The Netflix of China is continuing to rise in popularity among millennial investors.

Data from the stock trading app Robinhood — whose usership skews markedly younger than traditional brokerages — show iQiyi is now the 42nd most popular stock. It's gained 32 spots in the app's popularity ranking since last week, surpassing Dropbox, GM, and other more well-known companies.

The surge, which was first reported by Business Insider last week, mirrors a similar rise in share prices of the company, up 81% in the past month.

Ten-year-old iQiyi went public in March after being spun off by the Chinese search giant Baidu in an initial public offering that raised $2.25 billion for the video service. It currently boasts 20 million subscribers. For comparison, Netflix has more than 100 million worldwide, but Chinese regulations have hindered its expansion in the country.

Advertisement

IQiyi's post-IPO surge — a 90% gain in the roughly three months since — has led to several bullish analyst notes in recent weeks. Both Citic Securities, the investment-banking arm of the Chinese bank Citic Group, and First Shanghai upgraded shares to a "buy" rating. First Shanghai also raised its price target to $35 a share.

The FBI has warned that terrorists are working to acquire the cyber skills to carry out catastrophic attacks that could involve taking down America's 911 emergency response system.

Aristedes Mahairas, a special agent in charge of the New York FBI's Special Operations/Cyber Division, said terror groups like ISIS and Al-Qaeda are plotting attacks that could wipe out online networks to maximise casualties.

"The concern is that the cyber terrorists will one day be able to simultaneously execute an attack in the virtual world and in the physical world," Mahairas told Business Insider.

By way of an example, the FBI agent speculated on two horror scenarios:

"In such an environment, it would cause havoc," said the agent, who used to lead the New York FBI's counter-terror operations. "The truth is that such a scenario is only limited by your imagination."

Mahairas said terror groups may not currently have the capability to launch this kind of dual-pronged attack, but the FBI is working on the assumption that they one day could.

He continued: "It's really important to recognise that the terror groups are starting to leverage the technologies in such a way that it stands to reason that they are going to be able to recruit someone who is technically capable of spending time, effort, and energy to achieve that objective of eliminating the 911 emergency system...

"My working proposition is that they will eventually have the capability because they will be able to successfully recruit somebody."

Terrorists are also engaging hackers to help raise the money that actually carries out attacks, Mahairas said. "They need money, they need funds to pay their people, to feed them, to house them. If there's a lack of funding coming in then they may resolve to cybercriminal schemes," he added.

China appears to be censoring searches of words and phrases insulting to Kim Jong Un ahead of the US-North Korea summit on Tuesday.

Business Insider discovered the term "Fatty Kim the Third" appears to have been censored on both microblogging site Weibo and Chinese search engine Baidu. When the characters 金三胖 (pronounced Jin San Pang) are searched on these platforms, no results come up, despite dozens of articles being listed earlier in the year.

Empty results also occurred when Business Insider searched for other common nicknames for Kim, including "Prosperous Fat" (鑫胖). "Kim the Third Pig" (金三猪) turns up zero posts on Weibo and nothing from the last two years on Baidu, while "Fatty Kim Two Plus One" (金二加一胖) appears to be blocked on Baidu, but not on Weibo.

It's likely Beijing is trying to promote a positive perception of North Korea ahead of President Donald Trump's summit with Kim on Tuesday, and a potential China-Russia-North Korea meeting in the future.

Advertisement

Business Insider's findings come after Yonhap reported that China has "completely blocked internet searches for words and articles" that could be used to insult Kim Jong Un. But unlike in the past, Kim's name and other insults like "Fatty the Third" (三胖子) have not been blocked.

Earlier this year, China censored the term "Fatty on the Train" (胖坐火车) after people tried to avoid censorship of Kim's name and "North Korea" during the leader's surprise visit to Beijing by rail. Searching the term now on Weibo, Business Insider found no posts between January 2017 and May 2018.

Pyongyang has reportedly asked China to scrub unflattering nicknames of Kim in the past. But China maintains a strong bent toward censorship regardless of North Korea, which inadvertently pushes Chinese internet users to get creative with their search terms in order to avoid censorship algorithms and the thousands of human censors hired by large tech companies.

When Beijing ended presidential term limits earlier this year, dozens of words were banned on social media including just the letter 'n'. Censors also tried to block #metoo, but women instead used the term "Rice Bunny" (米兔) which is pronounced 'mi tu.'

Ride-hailing giant Uber is looking to get in on the scooter business, Axios reported Thursday.

The scooter companies say that these simple vehicles are a cheap, efficient way to provide so-called "last mile" transportation solutions, for trips that are too long to walk but too short to warrant a car ride. In that sense, this move makes sense for Uber — and will put it right up against its archrival Lyft in a burgeoning new market.

Uber confirmed to Axios the company is applying for a scooter permit in its hometown of San Francisco. The permit process was instituted after thousands of electric scooters descended on San Francisco in April with little-to-no-warning to city officials. However, despite the drama, these scooters have won plenty of fans.

With Uber joining the fray, at least seven companies are vying for a maximum of five scooter permits in San Francisco. These include the three companies that have already launched fleets in the city — Bird, LimeBike, and Spin — as well as newcomers, such as Lyft, Skip, and bike sharing company Ofo.

Advertisement

Uber did not immediately respond to a request for comment from Business Insider.

The permits allow a maximum of five companies to operate 500 each in San Francisco for a yearlong pilot program. The San Francisco Municipal Transportation Agency hopes to grant permits by the end of the month. Other cities across the country, including Seattle, Nashville, and Santa Monica, have all been targeted by the scooter companies, with worldwide expansion likely soon to follow.

If you're a criminal hacker you probably spend a lot of time on the "dark web."

That's the corner of the internet where hackers do things like sell stolen credit cards; buy "exploit kits," aka software products that help them hack; hire "botnets," or networks of hacked computers that can be programmed to do their misdeeds; or even hire a contract killer. The dark web is not accessible through a regular browser.

If you are a cyber security researcher, like Ziv Mador, you also spend much your time on the dark web, infiltrating the online criminal networks, studying them, and using your knowledge to help businesses defend against them.

Mador has spent two decades doing that. He's currently a lead researcher for security company Trustwave and previously spent 14 years working in computer security at Microsoft.

One of the most surprising findings in his years on the dark web is that these criminal organizations operate with a code of ethics much like the same ethics used by legit businesses, he told Business Insider.

"These are vivid communications, very functional. These are communities where cyber criminals exchange a lot of information and are very helpful [to each other ] if they are looking for a piece of information," he described.

So just like a programmer can get advice from fellow programmers on Stack Overflow or an IT pro can get product recommendations from others in IT on Spiceworks, cyber criminals will freely help one another solve problems or find products to do their own dark sites on their community websites.

A trustworthy reputation

There's a good reason for this: their street cred is their most important asset.

Thomson Reuters "Their reputation is very important to them. Much like it is in the business world," Mador says. "Even though they are involved with criminal or shady activities, they have their own rules of engagement and it's very similar to what people in the legit world do."

For instance, they don't share another's contact info without that person's permission. Spamming each other is a no-no and, above all else, they can't cheat or con one another.

Should they violate these ethics of behavior they would face any number of repercussions.

For one, they would lose customers to their competitors. "They are very competitive," Mador says.

Or, if they've really angered their fellow hackers, they could be "doxed," Mador says, meaning everything about their their real-life identity would be published for the other hackers to see. Unmasking a hacker's identity, especially in relation to a pissed off customer, is dangerous for them on all sorts of levels.

They are also price competitive with the products they sell. For instance, a handful of gangs sell exploit kits and compete aggressively on price and features. These kits must always be up-to-date on the latest security holes that can be used for hacking.

For the criminals that specialize in running botnets, networks of hacked computers for hire, they offer sophisticated realtime customer data analytics tools, similar to what any app developer wants from a cloud provider

"They have administration panels where their customers can login and see live data on infections," he describes.

Violence for sale

Even services that advertise a terrifying list of violent services are often run with similar code-of-conduct considerations.

For instance, these criminals often post a price list for the explicit acts of violence they will do from burning the car of an enemy to breaking bones.

When offering a hitman for hire, some outfits advertise the money saving option of hiring a novice who might fail. That could cost $5,000 compared to $200,000 to hire the most experienced killer on the roster.

Being a researcher on the dark web takes a level of courage but also patience, Mador says. It can take years to infiltrate such websites, getting criminals to trust that the fake identity of the researcher is indeed a fellow criminal and not the account of a researcher or law enforcement officer.

But once there, security researchers use their fake identities to monitor the underground, learning about things like stolen passwords, new types of malware and "what the next attacks are going to look like," Mador says.

He also shared with Business Insider a few examples of actual ads pulled from the dark web.

Google Cloud chief Diane Greene Business Insider

Google's pledge to quit doing military work involving its AI technology does not include its current job helping the Pentagon with drone surveillance.

Diane Greene, the head of Google Cloud, wrote in a blog post on Thursday that Google will continue to honor the controversial Project Maven contract that it entered into last year.

"I would like to be unequivocal that Google Cloud honors its contracts," Greene said in the blog post, adding that Google would fulfill the contract in a way that's "consistent" with the company's AI principles.

The comments came shortly after Google CEO Sundar Pichai published a list of governing principles on how the company plans to work with AI technology in the future. Those principles include not building applications that can be used as weapons or cause harm to humans.

Advertisement

The principles follow a conflict inside Google, pitting thousands of employees against management. These workers protested the company's involvement in Project Maven — the controversial collaboration between Google and the US Department of Defense. In March, news leaked that Google had quietly supplied AI technology to the Pentagon to help analyze drone video footage.

Greene described Maven on Thursday as a "limited contract" that Google entered into in September 2017 that "involved drone video footage and low-res object identification using AI" in which "saving lives was the overarching intent."

"We will not be pursuing follow on contracts for the Maven project," Greene said, but noted that the company would see the current contract through.

"While this means that we will not pursue certain types of government contracts," Greene wrote, "we want to assure our customers and partners that we are still doing everything we can within these guidelines to support our government, the military and our veterans. For example, we will continue to work with government organizations on cybersecurity, productivity tools, healthcare, and other forms of cloud initiatives."

Greene's comments underscore the tricky balance facing Google, particularly its cloud business, as it seeks to compete with Amazon, Oracle and Microsoft for lucrative government contracts while seeking to quell concerns among employees who feel that the company has veered from its Don't Be Evil corporate credo.

Read Greene's memo here.

For many in the cryptocurrency community, the hashtag #whenlambo signifies the Lamborghini as the embodiment of the aspirational wealth they someday hope to possess.

But for Elizabeth White, #whenlambo conveys a personal challenge.

When the hashtag started trending in cryptocurrency forums last year, White says she thought to herself, "You know what, I can get these people a Lambo."

Affluence may abound in the cryptocurrency community, but it isn't always easy to transform digital wealth in for real-world assets. Punishing transactional fees, low daily exchange limits on trading platforms, and wildly fluctuating values can make trading digital currencies in for fiat a difficult endeavor.

Advertisement

White, who has long worked in the racecar industry, decided to leverage her connections to hedge funds and luxury car dealerships to provide cryptocurrency holders access to the Italian supercar of their dreams.

"There's so much new wealth in the cryptocurrency community," says White. "Cars seem to be a big seller because it establishes you as a cryptocurrency holder."

Elizabeth White pictured with one of her cars, a Porsche 918 Spyder Weissach Edition. Elizabeth White

Typically, White's initial exchanges take place over the secure messaging app Telegram. White says she can facilitate a deal for a luxury car in exchange for cryptocurrencies in a matter of days.

"We had a very large sale to a buyer in China from a seller in California for a $4 million car," says White. "The negotiation was very quick. It took less than a week and the settlement took about 30 minutes."

White says her company, suitably called "the White Company," can handle such fast-paced deals in a mix of fiat and digital currencies because of the liquidity of the hedge fund, Apis Capital Management, with which she works.

"We're able to quickly convert someone's holdings at any moment," says White. "I can take these large amounts of money and purchase the items for my client, and then re-ingest their cryptocurrencies back into the fund."

Among the transactions she's handled are deals for Super Bowl suites, yachts, honeymoons, luxury fashion items, and engagement rings.

The ability to cash in digital assets for real-world goods is an important and validating function for cryptocurrencies, says White.

Advertisement

"It all goes back to offering cryptocurrency holders something they need," she says. "They need the ability to purchase something in the real world with their digital wealth."

Elizabeth White with a McLaren P1 Carbon Series. Elizabeth White

White's experience cashing in cryptocurrency assets for tangible goods has inspired her in an entirely new direction. Now, along with her finance partner Edgar Radjabli, she's launching her own cryptocurrency called "the White Standard" that she hopes will someday be used to handle the majority of online transactions.

Unlike ether and bitcoin, which have both been historically traded as a method of speculation, the "White Standard" is a "stable coin" — a digital token with the chief goal of serving as a means of digital commerce. White says that each White Standard, which are built on the blockchain application Stellar, will be backed by an American dollar, ensuring that the coins maintain a real-world value.

"As they gain adoption, White Standards will be trade-able on other exchanges," says White's finance partner, Edgar Radjabli. "You'll always be able to trade ether and bitcoin in and out of the White Standard."

Advertisement

Radjabli said that both he and White started considering the possibility of their own dollar-backed digital currency when rumors began swirling that Tether, another dollar-backed stable token, wasn't accurately representing the fiat backing for their coins.

The White Standard, which will be backed by Radjabli's hedge fund, plans on building a foundation of payment networks for cryptocurrency trade worldwide.

"We want to build a coin that will let you instantly buy a cup of coffee or a Lambo with cryptocurrencies anywhere in the world," says Radjabli.

Microsoft's Xbox group is in a weird place.

It has sold an estimated 30 to 50 million Xbox One consoles, putting Microsoft in a distant second place in the console race behind Sony's PlayStation 4 with more than 75 million.

And Nintendo's Switch console? It's a runaway success.

"Super Mario Odyssey" is available only on the Nintendo Switch. Over 9.7 million copies were sold between its launch in October and the end of 2017. Nintendo

In just over a year, Nintendo sold more than 17 million Switch consoles; it's the fastest-selling console in US history. Nintendo attributes this success primarily to one thing: a lot of really good games you can play only on the Switch.

The Xbox One, by comparison, isn't doing so great — though on paper it's competitive with or outright better than the competition from Sony and Nintendo.

Starting at $200, the Xbox One is cheap and jammed with great games to boot. Even the lowest-end model of the Xbox One supports HDR, a high-end video technology that makes games look better on TVs that support it.

It does everything a set-top box like the Apple TV does, like let you watch Netflix. And it plays blockbuster games.

The Xbox One S, the entry-level model, costs a cool $199. Florence Fu / Business Insider

In reality, though, it's the console I'm least likely to suggest.

The PlayStation 4 has many of the same games, like the latest "Call of Duty," and a bunch of great exclusive games, including the critically acclaimed "God of War."

Though the Nintendo Switch doesn't have third-party blockbusters like "Call of Duty," it has a big edge in terms of exclusives — there's no other way to play the latest "Super Mario" games. That's a pretty big advantage.

And if you already own a PlayStation 4 or PC? There simply aren't many major Xbox exclusive games that make the Xbox One worth owning. And this year's big Xbox One game, "Crackdown 3," just got delayed to February 2019.

"Crackdown 3" has been shown off by Microsoft for several years, but has yet to launch. Microsoft

So what's Microsoft going to do? That's the big question.

Here are some ways it could go:

Google CEO Sundar Pichai published a set of "ethical principles" on Thursday about how the company will work with artificial intelligence.

Pichai said that while AI technology provides consumers and businesses with many benefits, Google realizes the tech "will have a significant impact on society for many years to come" and that the managers "feel a deep responsibility to get this right."

Pichai said the AI applications will be screened to make sure that they are socially beneficial, won't create unfair bias, are safe, accountable to people, incorporate design principles, consistent with scientific excellence, and be made available for uses that maintain the previous principles.

The principles follow a conflict inside Google, pitting thousands of employees against management. These workers protested the company's involvement in Project Maven — the controversial collaboration between Google and the US Department of Defense. In March, news leaked that Google had quietly supplied AI technology to the Pentagon to help analyze drone video footage.

Advertisement

In April, more than 4,000 workers signed a petition demanding that Google's management cease work on Project Maven and promise to never again "build warfare technology." Soon after that, Gizmodo reported that a dozen or so Google employees had resigned in protest.

Pichai also made it clear on what sorts of applications that Google will not develop. Those include weapons or technologies that ccause overall harm, as well as anything that can be used for surveillance that violates "internationally accepted norms" or anything that conflicts with "widely accepted priciples of international law and human rights."

Read Pichai's full memo here.

Story developing...

Tesla CEO Elon Musk said the company will update its semi-autonomous Autopilot feature and start letting Tesla owners try it for free.

Speaking during the company's annual shareholder meeting on Tuesday, Musk said he thought the update would arrive this week and mark a "significant improvement" over Autopilot's current iteration. Musk didn't elaborate on what the update will include, and a Tesla representative declined to offer details.

Musk also said Tesla will give free trials of Autopilot to Tesla owners who didn't purchase the feature (it costs between $5,000-$6,000) and have the necessary hardware to run it "hopefully next month." Tesla last offered free trials of the feature in 2016.

In the next year, Musk said he expects Autopilot to make exponential improvements. When asked by a shareholder about when the full "Enhanced Autopilot" suite — which includes the ability to change lanes without driver input, switch from one freeway to another, and exit a freeway when close to the driver's destination — would arrive, Musk said at least one new feature could be rolled out in the next few months.

Advertisement

In its current iteration, Autopilot can keep a car in its lane and adjust its speed based on surrounding traffic, among other features. Recent accidents involving the feature have raised questions about whether drivers place too much trust in it and fail to pay attention to the road. Tesla has repeatedly said Autopilot is meant to be used with an attentive driver whose hands are on the wheel, but the most visible accidents involving Autopilot have included reports of distracted drivers.

Tesla has received criticism for how it has promoted the feature. In May, Consumer Watchdog and the Center for Auto Safety sent a letter to the Federal Trade Commission asking the agency to investigate the strategies the company has used to sell Autopilot.