All posts by MTechnology

YouTube to Display Wikipedia Blurbs Alongside Conspiracy Videos

YouTube will begin displaying text from Wikipedia articles and other websites alongside some videos in a couple of weeks as the unit of Alphabet Inc’s Google attempts to combat hoaxes and conspiracy theories on the service, its chief executive said on Tuesday.

Susan Wojcicki, speaking on stage at the South by Southwest Conference in Austin, Texas, displayed a mock-up of the new feature, which are called information cues.

YouTube intends to present an alternative viewpoint to videos questioning science or describing conspiracies about events such as the U.S. moon landing. She said information cues would first roll out to topics for which there are a significant number of YouTube videos.

“People can still watch the videos but then they actually have access to additional information, can click off and go and see that,” Wojcicki said.

Lawmakers and media advocacy groups have called on YouTube to help stop the spread of hoaxes and false news stories. Last year, the company adjusted its algorithms to promote what it described as authoritative sources.

Though music and gaming videos are far more popular on YouTube, the company has made addressing the criticism around news and science videos a top priority this year. 

Google Brings Free WiFi to Mexico, First Stop in Latin America




Alphabet’s Google said on Tuesday that it will launch a network of free Wi-Fi hotspots across Mexico, part of the search giant’s effort to improve connectivity in emerging markets and put its products in the hands of more users.

Google Station, an ad-supported network of Wi-Fi hotspots in high-traffic locations, is launching in Mexico with 56 hotspots and others planned, the company said.

Mexico will be Google Station’s third market following India and Indonesia, and the first in Latin America.

Mexico has made great strides in connectivity since a 2013-14 telecom reform intended to loosen the grip of billionaire Carlos Slim’s America Movil, which has long dominated the market.

From 2013 to 2016, the number of people accessing the Internet in Mexico rose by 20 million, according to a report last fall by the Organization for Economic Co-operation and Development. Still, the country lags behind other OECD nations in terms of internet access, the report said.

“We are finding that public Wi-Fi remains still a very important way to get online,” Anjali Joshi, a vice president for product management at Google, told reporters.

She added that Google saw Mexico as a good entrypoint for the product in Latin America. Mexico-based SitWifi provided equipment for the hotspots.

Google’s initial batch of Wi-Fi zones is scattered across the country, from the Ciudad Juarez airport at the U.S. border to posh shopping centers in Mexico City.

Google Station now counts roughly 8 million users a month in India, where the program began in 2016.

A New Method for Extracting CO2 from Seawater

Scientists are always on the lookout for affordable and efficient methods for capturing carbon dioxide, responsible for global warming and the rising acidity of seawater. A new procedure, developed at the University of York in Britain, promises to extract large amounts of CO2 from seawater and store it safely, and recycle millions of tons of aluminum waste at the same time. VOA’s George Putic has more.

UN Investigators Cite Facebook Role in Myanmar Crisis

U.N. human rights experts investigating a possible genocide in Myanmar said Monday that Facebook had played a role in spreading hate speech there.

Facebook had no immediate comment on the criticism Monday, although in the past the company has said that it was working to remove hate speech in Myanmar and kick off people who shared such content consistently.

More than 650,000 Rohingya Muslims have fled Myanmar’s Rakhine state into Bangladesh since insurgent attacks sparked a security crackdown last August. Many have provided harrowing testimonies of executions and rapes by Myanmar security forces.

The U.N. human rights chief said last week he strongly suspected acts of genocide had taken place. Myanmar’s national security adviser demanded “clear evidence.”

Marzuki Darusman, chairman of the U.N. Independent International Fact-Finding Mission on Myanmar, told reporters that social media had played a “determining role” in Myanmar.

“It has … substantively contributed to the level of acrimony and dissention and conflict, if you will, within the public. Hate speech is certainly, of course, a part of that. As far as the Myanmar situation is concerned, social media is Facebook, and Facebook is social media,” he said.

U.N. Myanmar investigator Yanghee Lee said Facebook was a huge part of public, civil and private life, and the government used it to disseminate information to the public.

“Everything is done through Facebook in Myanmar,” she told reporters, adding that Facebook had helped the impoverished country but had also been used to spread hate speech.

“It was used to convey public messages, but we know that the ultra-nationalist Buddhists have their own Facebooks and are really inciting a lot of violence and a lot of hatred against the Rohingya or other ethnic minorities,” she said. “I’m afraid that Facebook has now turned into a beast, and not what it originally intended.”

The most prominent of Myanmar’s hardline nationalist monks, Wirathu, emerged from a one-year preaching ban Saturday and said his anti-Muslim rhetoric had nothing to do with violence in Rakhine state.

Facebook suspends and sometimes removes anyone that “consistently shares content promoting hate,” the company said last month in response to a question about Wirathu’s account.

“If a person consistently shares content promoting hate, we may take a range of actions such as temporarily suspending their ability to post and, ultimately, removal of their account.”

World Wide Web Inventor Says Big Tech Must Be Regulated

The inventor of the worldwide web, Tim Berners-Lee, called on Monday for powerful internet platforms and social media companies to be regulated to prevent the internet from being “weaponized at scale.”

The British computer scientist, in an open letter published on the 29th anniversary of the creation of the web, said a “new set of gatekeepers” was now dominant, controlling the spread of ideas and opinions.

“The fact that power is concentrated among so few companies has made it possible to weaponize the web at scale,” he wrote.

“In recent years, we’ve seen conspiracy theories trend on social media platforms, fake Twitter and Facebook accounts stoke social tensions, external actors interfere in elections and criminals steal troves of personal data.”

The intervention by the 62-year-old MIT professor comes as some European governments turn to legislation to curb “fake” news and hate speech that they fear is undermining the basis of their democracies.

In Germany, a law entered force on January 1 that foresees fines of up to 50 million euros ($62 million) on internet platforms that fail to remove hate speech — which is illegal — within 24 hours.

French President Emmanuel Macron meanwhile plans legislation that would empower judges to order the removal of fake news during election campaigns.

And in Brussels, the European Commission has served notice to internet platforms that they must find a way to remove extremist content within one hour of being notified, or face legislation compelling them to do so.

Berners-Lee, whose Web Foundation campaigns for a more open and inclusive internet, doubted that companies that have been built to maximize profits can adequately address the problem on a voluntary basis.

“A legal or regulatory framework that accounts for social objectives may help ease those problems,” he said.

Expressing concern over how big internet platforms handle users’ data in targeting advertising, Berners-Lee said a balance needed to be found between the interests of companies and online citizens.

“This means thinking about how we align the incentives of the tech sector with those of users and society at large, and consulting a diverse cross-section of society in the process.”

What Happens at SXSW?

What originally started as a music festival in the 1980s has evolved into an event that is much bigger and harder to define. Imagine networking and partying for more than a week. That is what is happening in Austin, Texas. Musicians, film promoters and tech companies from around the world are gathering for the South by Southwest (SXSW) conference and festival. VOA’s Elizabeth Lee has the details from Austin.

Scientists Hope to Clean Space Junk

Space scientists say the satellites and other spacecraft orbiting the Earth, including the International Space Station, are in increasing danger of collision with pieces of junk. Engineers are working hard to solve the problem of removing the trash that threatens functioning satellites worth millions of dollars. VOA’s George Putic reports.

India, France Call for Affordable Solar Technology to Address Climate Change

French President Emmanuel Macron pledged over $850 million for solar projects in emerging economies, as both India and France called for affordable solar technology for emerging nations at the first conference of the International Solar Alliance (ISA) held in New Delhi.

 

The alliance was co-founded by both countries two years ago on the sidelines of the Paris climate summit to boost the use of solar power, countering the impact of climate change.

 

Dozens of country leaders, including many from Africa, attended the meeting in the Indian capital and emphasized the need for access to solar technology and concessional financing to address massive energy shortages in many of their sun-drenched nations.

 

Promising more loans and donations for solar projects by 2022, Macron stressed the need to remove obstacles in scaling up clean energy.

“We only have one planet, and we are sharing it,” he said.

 

Pointing to African women called “solar mamas” who are trained in India to use solar technology to light up homes and villages, Macron said they had continued their mission, even after “some countries decided just to leave the floor and leave the Paris agreement” — apparently alluding to U.S. President Donald Trump’s decision to quit the Paris climate accord.

 

“Because they decided it was good for them, for their children, their grandchildren. They decided to act and keep acting, and that’s why we are here, in order to act very concretely,” Macron said amid applause.

 

One hundred and twenty-one countries, situated between the tropics, have signed on to the ISA. Backed by the World Bank and other multilateral agencies, it aims to raise $1 trillion for projects by 2030 for a massive deployment of solar energy.

 

Rwandan President Paul Kagame, who is chairman of the African Union, pointed out that half the members of the ISA are African countries.

“The sunniest countries in the world should not lack for energy,” he said. “The fact that they do is an unacceptable irony.”

 

The solar alliance initiative is seen as a bid by India to be at the forefront of countries addressing the challenge of climate change — a departure from its stand some years ago that developed economies should cut their emissions more drastically, rather than pressure developing countries.  

 

After the U.S. walked out of the Paris accord, Indian Prime Minister Narendra Modi pledged to abide by it. India, which is the world’s third largest polluter, is ramping up solar energy rapidly in a bid to reduce its carbon footprint. The country plans to source at least 40 percent of its energy from renewables by 2030.

 

“If you want all of humanity to benefit, then I am confident that we all will come together and think like one family, so that we are able to bring unity in our objectives and efforts,” said Modi, advocating a solar revolution worldwide.

United Nations environment chief Erik Solheim, who attended the meeting in New Delhi, called the ISA a “milestone” in the fight against climate change and pollution.

 

 

 

 

 

The Rising Problem of Old Batteries

Technology increasingly relies on rechargeable batteries as a source of energy. Today’s batteries are better and last longer, but when their capacity drops under a certain level they have to be replaced. Some experts say that, even with a half of their capacity, batteries can be used for less critical purposes. VOA’s George Putic has more.

Facebook Exclusive Deal: Streaming 25 MLB Games

Facebook is getting deeper into the professional sports streaming game, partnering with Major League Baseball to air 25 weekday afternoon games in an exclusive deal.

The games will be available to Facebook users in the U.S. on Facebook Watch, the company’s video feature announced last August, via the MLB Live show page. Facebook said Friday that recorded broadcasts will also be available globally, excluding select international markets.

The package, MLB’s first digital-only national broadcast agreement, precludes teams from televising those games on their regional sports networks. The concept is similar to the exclusive package of Sunday night games on ESPN.

Facebook, Twitter and Amazon and other tech companies are in a race to acquire sports streaming rights, which can be lucrative and potentially boost user loyalty. The deal comes at a time when leagues are worrying about cord-cutters causing a decrease in viewers among cable television networks.

Verizon signed a deal with the NBA to stream eight basketball games on Yahoo, and Amazon paid $50 million to stream NFL games to Prime members last season.

The games will be produced by the MLB Network for Facebook Watch, with interactive and social elements that differentiate them from live streaming.

Facebook’s first-month schedule includes Philadelphia-New York Mets on April 4, Milwaukee-St. Louis on April 11, Kansas City-Toronto on April 18 and Arizona-Philadelphia on April 26.

Facebook had a package of 20 non-exclusive Friday night games last year that began in mid-May and used broadcast feeds from the participating teams.

Watchdog: Western Tech Used for Hacking in Turkey, Syria

A Canadian company’s hardware is being used to hack internet users along Turkey’s border with Syria, researchers said Friday, adding that there were signs that Kurdish forces aligned with the United States might have been targeted.

The revelation comes as Turkey presses its offensive against the Kurds dug in along the country’s frontier with northwestern Syria, a conflict that threatens to disrupt the American-led effort to extinguish the Islamic State group. The apparent use of Canadian technology to target a U.S. ally was an irony underlined by Ron Deibert, the director of the internet watchdog group Citizen Lab, which published a report on the spying.

“These companies are not closely regulated, and that can lead to a lot of unintended consequences, including consequences that harm our foreign policy interests and human rights interest as well,” Deibert said. “It’s a strong argument for government control over this kind of technology.”

Canadian tech

 

Citizen Lab identified the hardware behind the hacking as PacketLogic devices produced by Procera, a Fremont, California-based company that was recently folded into Canada-based network management firm Sandvine, which is owned by American private equity group Francisco Partners. 

 

In a statement issued before the report’s release, Sandvine said it investigates all allegations of abuse but said it had been unable to complete its inquiry because Citizen Lab refused to provide the company with its findings in full. 

“Once we have the necessary data, we will conduct a full investigation and take appropriate action,” Sandvine said.

The statement also said Citizen Lab’s allegations were “technically inaccurate and intentionally misleading,” but a representative for the company has yet to supply an example of a misleading or inaccurate claim.

Government spying

Citizen Lab said it discovered the hacking after a European cybersecurity company reported that network service providers in two unidentified countries were trying to compromise their users using a powerful hacking technique known as network injection. Citizen Lab scoured the internet for signs of the spying and eventually traced the activity to the Turkish provinces of Adana, Hatay, Gaziantep, Diyarbakir and to the Turkish capital, Ankara, as well as parts of northern Syria and Egypt. 

 

Network injection — so-called because malicious software is injected into everyday internet traffic by whoever controls the network — has long been feared as a particularly powerful form of government spying.

“This can potentially be used to target anyone in the country with the click of the button,” said Bill Marczak, the lead author of the report.

 

Although the identities of those being spied on in Turkey and Egypt aren’t clear, Marczak said that the devices appeared to be installed on the network belonging to Turk Telekom, a leading phone and internet provider in Turkey as well as parts of northern Syria. He said there were hints suggesting some of the targets are affiliated with the YPG, the Kurdish Marxist rebel group which is fighting Turkish forces for control of the northwestern Syrian province of Afrin. Although Turkey considers the YPG a terrorist organization, the group provides the backbone of the U.S.-backed operations against the Islamic State in eastern Syria.

 

American officials acknowledged Monday that ground operations against the jihadist group’s remnants in eastern Syria were on hold because Kurdish fighters were being diverted to the battle against Turkey. 

Turk Telekom statement

 

Turk Telekom said in a statement that it complies with Turkish law and doesn’t interfere with internet users’ access. It added that the company “does not redirect any internet user to receive malicious downloads of popular software applications.” A representative for the company did not immediately respond to follow-up questions.

 

Sandvine’s ties to the Turkey government have been the subject of previous reporting. In 2016, Forbes reported that engineers at Procera were so troubled at the prospect of supplying surveillance hardware for use by Turk Telekom that six of them quit in protest. 

 

“I do not wish to spend the rest of my life with the regret of having been a part of (Turkish President Recep Tayyip) Erdogan’s insanity, so I’m out,” one the engineers said in a letter of resignation quoted by Forbes.

 

LinkedIn shows at least 16 Procera-Sandvine employees listed as working in Egypt or Turkey. One Sandvine engineer based in Cairo listed “lawful interception” — a commonly used euphemism for state-sanctioned surveillance — as one of his interests.

Students Learn Real Skills By Creating Virtual Businesses

Students around the United States are creating virtual businesses that produce simulated products, which are marketed and sold for virtual money. Mike O’Sullivan reports that hundreds of student entrepreneurs came to Pasadena, California, to promote the virtual companies they have created.

Judge to Weigh Whether Trump’s Twitter Blocks Violate Free Speech

A federal judge is expected to hear arguments on Thursday about whether President Donald Trump violated Twitter users’ free speech rights under the U.S. Constitution by blocking them from his account.

The arguments before U.S. District Judge Naomi Reice Buchwald in Manhattan are part of a lawsuit brought last July by the Knight First Amendment Institute at Columbia University and several individual Twitter users.

Trump and the plaintiffs are seeking summary judgment, asking Buchwald to decide the case in their favor without a trial.

Twitter lets users post short snippets of text, called tweets. Other users may respond to those tweets. When one user blocks another, the blocked user cannot respond to the blocker’s tweets.

The plaintiffs have accused Trump of blocking a number of accounts whose owners criticized, mocked or disagreed with him in replies to his tweets.

They argued that Trump’s Twitter account, @realDonaldTrump, is a public forum, and that denying them access based on their views violates the First Amendment.

Trump in court papers countered that his use of Twitter is personal, not a “state action.”

Even if it were a state action, he said, his use of Twitter was a form of “government speech,” not a public forum.

Trump’s Twitter use draws intense interest for his unvarnished commentary, including attacks on critics. His tweets often shape news and are retweeted tens of thousands of times.

Ocean Discovery XPRIZE Competition Finalists Announced

Nine teams with members from 25 countries entered the final phase of the $7 million Ocean Discovery XPRIZE competition for technologies that could cheaply and efficiently map the floors of the world’s oceans.  Using current technology this feat would take up to 600 years. VOA’s George Putic reports.

FBI Chief: Corporate Hack Victims Can Trust We Won’t Share Info

The FBI views companies hit by cyberattacks as victims and will not rush to share their information with other agencies investigating whether they failed to protect customer data, its chief said Wednesday. Christopher Wray, director of the Federal Bureau of Investigation, encouraged companies to promptly report when they are hacked to help the FBI investigate and prevent future data breaches.

He contrasted the FBI’s approach to that of other regulators and state authorities. Without naming other agencies, Wray referred to “less-enlightened enforcement agencies,” some of which he said take a more adversarial approach.

“We don’t view it as our responsibility when companies share information with us to turn around and share that information with some of those other agencies,” Wray said in response to an audience question at a cybersecurity conference at Boston College.

Amid a wave of high-profile data breaches at major corporations, the Federal Trade Commission (FTC) and state attorneys general are investigating how many of them secured consumer data before they were hacked.

Equifax Inc, which suffered a breach in 2017 that compromised the data of more than 147 million consumers, is fighting a lawsuit by Massachusetts Attorney General Maura Healey and faces probes by over 40 other states and the FTC.

Ride-sharing company Uber Technologies Inc is also facing investigations by state attorneys general after a data breach of 57 million accounts. Uber has been sued by the states of Washington and Pennsylvania, and like Equifax faces private class action lawsuits over the breach.

Speaking at the conference, Wray said the FBI needed to partner with the private sector to combat an evolving threat that has “turned into full-blown economic espionage and extremely lucrative cybercrime.”

Wray, who took over as director in August, said in order to prevent cyber threats, companies should approach the FBI as soon as they see signs of unauthorized access to their computer systems or malware infesting them.

“At the FBI, we treat victim companies as victims,” he said.

 

FOMO at SXSW: How to Conquer Fear of Missing Out in Austin

The South by Southwest festival in Austin, Texas, starts Friday. It’s grown from a grassroots event to a phenomenon that attracts 400,000 people.

For attendees, it can feel overwhelming. What’s worth your time? Where’s the buzz?

 

The latest AP Travel “Get Outta Here” podcast offers strategies for conquering FOMO (fear of missing out) at SXSW.

 

One approach is to let the nostalgia acts go – the former big-name bands promoting comebacks. Instead, pack your schedule with artists that have their best years ahead of them.

 

And you need a plan. You can’t just wing it. Be ready for long lines. But have some backups. Consider less-crowded venues outside downtown. Film screenings take place at theaters all over, and up-and-coming bands play a lot of shows.

Facebook, Twitter Urged to Do More to Police Hate on Sites

Tech giants Facebook, Twitter and Google are taking steps to police terrorists and hate groups on their sites, but more work needs to be done, the Simon Wiesenthal Center said Tuesday.

The organization released its annual digital terrorism and hate report card and gave a B-plus to Facebook, a B-minus to Twitter and a C-plus to Google.

Facebook spokeswoman Christine Chen said the company had no comment on the report. Representatives for Google and Twitter did not immediately return emails seeking comment.

Rabbi Abraham Cooper, the Wiesenthal Center’s associate dean, said Facebook in particular built “a recognition that bad folks might try to use their platform” into its business model. “There is plenty of material they haven’t dealt with to our satisfaction, but overall, especially in terms of hate, there’s zero tolerance,” Cooper said at a New York City news conference.

Rick Eaton, a senior researcher at the Wiesenthal Center, said hateful and violent posts on Instagram, which is part of Facebook, are quickly removed, but not before they can be widely shared.

He pointed to Instagram posts threatening terror attacks at the upcoming World Cup in Moscow. Another post promoted suicide attacks with the message, “You only die once. Why not make it martyrdom.”

Cooper said Twitter used to merit an F rating before it started cracking down on Islamic State tweets in 2016. He said the move came after testimony before a congressional committee revealed that “ISIS was delivering 200,000 tweets a day.”

Cooper and Eaton said that as the big tech companies have gotten more aggressive in shutting down accounts that promote terrorism, racism and anti-Semitism, promoters of terrorism and hate have migrated to other sites such as VK.com, a Facebook lookalike that’s based in Russia.

There also are “alt-tech” sites like GoyFundMe, an alternative to GoFundMe, and BitChute, an alternative to Google-owned YouTube, Cooper said.

“If there’s an existing company that will give them a platform without looking too much at the content, they’ll use it,” he said. “But if not, they are attracted to those platforms that have basically no rules.”

The Los Angeles-based Wiesenthal Center is dedicated to fighting anti-Semitism, hate and terrorism.

Porsche Says Flying Cab Technology Could Be Ready Within Decade

Porsche is studying flying passenger vehicles but expects it could take up to a decade to finalize technology before they can launch in real traffic, its head of development said Tuesday.

Volkswagen’s sports car division is in the early stages of drawing up a blueprint of a flying taxi as it ponders new mobility solutions for congested urban areas, Porsche R&D chief Michael Steiner said at the Geneva auto show.

The maker of the 911 sports car would join a raft of companies working on designs for flying cars in anticipation of a shift in the transport market toward self-driving vehicles and on-demand digital mobility services.

“We are looking into how individual mobility can take place in congested areas where today and in the future it is unlikely that everyone can drive the way he wants,” Steiner said in an interview.

VW’s auto designer Italdesign and Airbus exhibited an evolved version of the two-seater flying car called Pop.Up at the Geneva show. It is designed to avoid gridlock on city roads and premiered at the annual industry gathering a year ago.

Separately, Porsche expects the cross-utility variant of its all-electric Mission E sports car to attract at least 20,000 buyers if it gets approved for production, Steiner said.

Porsche will decide later this year whether to build the Mission E Cross Turismo concept, which surges to 100 kph (62 mph) in less than 3.5 seconds, he said.

Preventing Unauthorized Use of Face Recognition

Every day, billions of photos uploaded to the Internet contain faces. Experts say sophisticated algorithms can collect these images, compare and glean information – some for law enforcement agencies and some for hackers, intent on stealing and misusing that data. An Israeli company says there’s a way to prevent that. VOA’s George Putic has more.

AI Has a Dirty Little Secret: It’s Powered by People

There’s a dirty little secret about artificial intelligence: It’s powered by an army of real people.

From makeup artists in Venezuela to women in conservative parts of India, people around the world are doing the digital equivalent of needlework -drawing boxes around cars in street photos, tagging images, and transcribing snatches of speech that computers can’t quite make out.

Such data feeds directly into “machine learning” algorithms that help self-driving cars wind through traffic and let Alexa figure out that you want the lights on. Many such technologies wouldn’t work without massive quantities of this human-labeled data.

These repetitive tasks pay pennies apiece. But in bulk, this work can offer a decent wage in many parts of the world – even in the U.S. And it underpins a technology that could change humanity forever: AI that will drive us around, execute verbal commands without flaw, and – possibly – one day think on its own.

For more than a decade, Google has used people to rate the accuracy of its search results. More recently, investors have poured tens of millions of dollars into startups like Mighty AI and CrowdFlower, which are developing software that makes it easier to label photos and other data, even on smartphones.

Venture capitalist S. “Soma” Somasegar says he sees “billions of dollars of opportunity” in servicing the needs of machine learning algorithms. His firm, Madrona Venture Group, invested in Mighty AI. Humans will be in the loop “for a long, long, long time to come,” he says.

Accurate labeling could make the difference between a self-driving car distinguishing between the sky and the side of a truck – a distinction Tesla’s Model S failed in the first known fatality involving self-driving systems in 2016.

“We’re not building a system to play a game, we’re building a system to save lives,” says Mighty AI CEO Daryn Nakhuda.

Marjorie Aguilar, a 31-year-old freelance makeup artist in Maracaibo, Venezuela, spends four to six hours a day drawing boxes around traffic objects to help train self-driving systems for Mighty AI.

She earns about 50 cents an hour, but in a crisis-wracked country with runaway inflation, just a few hours’ work can pay a month’s rent in bolivars.

“It doesn’t sound like a lot of money, but for me it’s pretty decent,” she says. “You can imagine how important it is for me getting paid in U.S. dollars.”

Aria Khrisna, a 36-year-old father of three in Tegal, Indonesia, says that adding word tags to clothing pictures on websites such as eBay and Amazon pays him about $100 a month, roughly half his income.

And for 25-year-old Shamima Khatoon, her job annotating cars, lane markers and traffic lights at an all-female outpost of data-labeling company iMerit in Metiabruz, India, represents the only chance she has to work outside the home in her conservative Muslim community.

“It’s a good platform to increase your skills and support your family,” she says.

The benefits of greater accuracy can be immediate. At InterContinental Hotels Group, every call that its digital assistant Amelia can take from a human saves $5 to $10, says information technology director Scot Whigham.

When Amelia fails, the program listens while a call is rerouted to one of about 60 service desk workers. It learns from their response and tries the technique out on the next call, freeing up human employees to do other things.

When a computer can’t make out a customer call to the Hyatt Hotels chain, an audio snippet is sent to AI-powered call center Interactions in an old brick building in Franklin, Massachusetts. There, while the customer waits on the phone, one of a roomful of headphone-wearing “intent analysts” transcribes everything from misheard numbers to profanity and quickly directs the computer how to respond.

That information feeds back into the system. “Next time through, we’ve got a better chance of being successful,” says Robert Nagle, Interactions’ chief technology officer.

Researchers have tried to find workarounds to human-labeled data, often without success.

In a project that used Google Street View images of parked cars to estimate the demographic makeup of neighborhoods, then-Stanford researcher Timnit Gebru tried to train her AI by scraping Craigslist photos of cars for sale that were labeled by their owners.

But the product shots didn’t look anything like the car images in Street View, and the program couldn’t recognize them. In the end, she says, she spent $35,000 to hire auto dealer experts to label her data.

Trevor Darrell, a machine learning expert at the University of California Berkeley, says he expects it will be five to 10 years before computer algorithms can learn to perform without the need for human labeling. His group alone spends hundreds of thousands of dollars a year paying people to annotate images.