MIT Researchers are using neural networks and haptic feedback to make conversations easier to navigate.

Wearable electronic devices today are, for the most part, toys for narcissists.

Think about it: Fitbit is great at gathering data about your own body, Snap Spectacles captures the world through your own eyes, and even the Apple Watch offers an inwardly focused information stream. What if, instead, wearables were made to listen and help communicate to the world around you?

That's the goal behind a wave of interesting projects, including one announced this week by a group of scientists from the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory, and an ongoing project from a Houston startup. Both have created wearables that not only gather sound but communicate with the wearer using tactile or haptic feedback that doesn't require a visual display. This could open a world of new use cases for wearable devices.

A pair of graduate students from from the MIT lab are slated to present a paper at next week's Association for the Advancement of Artificial Intelligence conference in San Francisco highlighting how machine learning can be used to detect emotion in real-time speech and then communicate that to another person.

Their system takes spoken-word transcripts, sound samples, and data from a Samsung Simband, including a speaker's electrocardiogram readings and skin temperature measurements, and feeds it into a neural network. The network is trained based on human-labeled speech snippets to use all that information to detect the speakers' emotional states, which the researchers report it does at a rate 18% better than chance and 7.5% better than previous approaches.

The resulting emotion, once detected, can then be transmitted to another person. This technique could be applied, for example, to people with Asperger's syndrome or other conditions that can make it difficult to perceive human emotions. Now they could receive real-time feedback on how their conversational partners are feeling, says Mohammad Mahdi Ghassemi, one of the coauthors of the paper.

"If you want to navigate from one point in San Francisco to another or one point in Boston to another, you can pull out a piece of technology—Apple Maps, Google Maps or whatever—and it will navigate you from Point A to Point B," he says. "Navigating a Thanksgiving conversation can be just as complicated, and there's no guidance on it."

Future research is likely to streamline the approach. That means potentially limiting the input to just spoken audio rather than biometric readings, which could make the tool more practical for real-world scenarios where conversation partners won't freely give you a feed of their biometric data. And it would likely use real-time notifications, which could discreetly notify users through a wearable device about when a conversation starts going sideways.

"If we're assessing in an interaction that a conversation is getting, let's say, awkward, or negative, the wristwatches they're wearing will buzz twice," says Ghassemi. "It can sort of cue you in that you might want to change gears."

Smartwatches are already being hacked by people with difficulty hearing and seeing to serve as an alternative input channel. Accessibility consultant Molly Watt, who has severely impaired vision and hearing, wrote a blog post in 2015 about her success with the Apple Watch. The device allows her to exchange simple messages from friends and family with just a few taps, even in environments where she can't otherwise easily see or hear, she wrote.

"Mum has certainly found benefit in the 'tap' for getting my attention when I am in my bedroom without my hearing aids on," she wrote. "I feel the nudge to get a move on or she wants my attention for something."

A more specialized device called The VEST, developed by Houston startup NeoSensory, goes further: It translates sounds into vibrations from 32 motors distributed across a user's torso. NeoSensory cofounders David Eagleman and Scott Novich, both neuroscientists, say the VEST can potentially help deaf and hearing-impaired people learn to recognize sounds and even spoken words by training themselves to recognize the different vibrations from the device.

Research shows that people can learn to perceive signals normally received through one sense when it's delivered through another, meaning deaf people can learn to feel sounds or blind people can hear images, says Eagleman, the company's chief science officer and the host of PBS series The Brain with David Eagleman. In a TED talk filmed in 2015, Eagleman screened footage of a profoundly deaf man successfully using the VEST to recognize spoken words, explaining the devices could be up to 40 times cheaper than the cochlear implants now typically used to enable deaf patients to hear.

"We have people wear the vest all day long, and they play this whole suite of games that we have on the phone," Eagleman tells Fast Company. One game triggers a vibration corresponding to a particular spoken word, then asks the user to choose between two options, he says.

"At first, you're completely guessing, because you've never felt this before," Eagleman says, but users get better with practice. "The younger the subject, the better the learning is—the better and faster the learning is."

Since then, the NeoSensory team has been refining the algorithms that determine how the motor vibration patterns shift in response to different sound. The company's also been streamlining prototypes of the VEST with an eye toward manufacturing the devices later this year, says Novich, the chief technology officer.

"You can wear them underneath your clothing," he says. "Nobody knows you really have them on, they're really silent, last all day on a single charge, and they're totally robust."

And once the VESTs are released to the public, even people with no hearing impairments may be able to benefit, since an API will let developers experiment with sending arbitrary signals to the vibrating motors from a linked smartphone. That could let the devices be used to transmit anything from infrared images to financial data to video game-enhancing feedback.

"We can think of like 25 things that we would be awesome data streams to feed in, but there are going to be 25,000 things to try that we haven't even thought of," says Eagleman.

Travelers looking to keep device data private at border crossings need to rely on technology, not the law.

Anyone who has traveled internationally knows the drill: body scans, bag searches, and lots of questions. But there's something else U.S. Customs and Border Protection officers increasingly want to scour: your cell phone.

Custom officials are taking advantage of legal uncertainties that civil liberties advocates say provide limited options for travelers to keep their devices and data private.

That means everything on your phone—from financial info to potentially embarrassing photos—can land in the government's hands. The policy pre-dates the controversial travel restrictions President Donald Trump ordered Friday, and applies to U.S. citizens and international visitors alike. In fact, the Florida chapter of the Council on Islamic-American Relations filed a complaint the week before Trump's inauguration after reports of cellphone searches and demands for social media handles. And in July, a reporter for The Wall Street Journalsaid that federal officials demanded to inspect phones she was carrying into the country, before relenting when she said they belonged to the newspaper.

For travelers looking to keep their personal or business data private, that leaves a few options:

• Travel with "burner" phones and laptops only loaded with files you absolutely need for your trip.
• Keep private data in the cloud, not on your devices.
• Use a secure deletion tool to wipe sensitive data from your phone and laptop before passing through customs.
• Lock and encrypt your phone before crossing the border, though officers may urge you to unlock it or reveal your passwords.

Digital inspections are unlikely to go away any time soon, experts say.

"This issue of device searches at the border and asking for social media information was starting to happen under Obama," says Sophia Cope, a staff attorney at the Electronic Freedom Foundation. She predicts the searches may intensify under Trump.

Customs officers have long had wide leeway to search people's possessions after they enter the U.S.; usually in a hunt for contraband, including guns and drugs, says Nathan Wessler, a staff attorney with the American Civil Liberties Union's Speech, Privacy, and Technology Project. U.S. border officials have generally interpreted those rules to also allow warrantless searches of digital devices.

"Under DHS authorities to conduct border searches, travelers' electronic devices are as subject to search as any other belongings, because the information contained in them may be relevant to DHS's customs and immigration inspection processes and decisions," wrote Department of Homeland Security officials in a 2009 report. And in an email to Fast Company on Wednesday, CBP spokesman Roland Filiault reiterated the agency's authority to inspect electronics coming into the U.S.

"Keeping America safe and enforcing our nation's laws in an increasingly digital world depends on our ability to lawfully examine all materials entering the U.S.," he wrote.

The problem, says Wessler, is that smartphones and laptops now contain huge amounts of private data, from financial and medical records to private conversations and photos—much more than old-fashioned luggage items typically would. It's the ACLU's view, however, that the idea of stopping contraband means examing physical packages, not digital data.

"Privacy interests are very high and, at the same time, we're not now talking about the potential for drugs or agricultural products slipping through," Wessler says. "These are digital bits, and there's no reason the government can't go and get a warrant."

In 2014, the U.S. Supreme Court ruled unanimously that without a warrant, police generally can't search a suspect's cellphone when they're taken into custody, despite established rules allowing some physical warrantless searches during an arrest.
There's no similar Supreme Court case ruling one way or another about searching travelers entering the country. A few federal courts have allowed unrestricted manual searches while requiring at least some "reasonable suspicion" before doing more intensive forensic exams of a phone's memory, but many haven't addressed the question at all, says Wessler.

"Most federal courts have yet to weigh in on this, so in most parts of the country there's no binding legal authority on the question," he says.

In practice, this means people traveling into the country should assume they have limited legal recourse to stop device searches, says Cope. How people handle that will depend on a number of factors. Some travelers without much personal information on their phones may be unconcerned, while others might encrypt their devices and refuse to disclose their passwords. That, though, can be more of a risk for noncitizens, who can be denied entry with a fair degree of latitude, Cope cautions.

"It's all about the person's risk tolerances," she says.

There's no one-size-fits-all option: Using travel-only phones and laptops, or keeping files in the cloud may work well for some, but those solutions rely on either technical sophistication or disposable funds, and assume travelers don't actually need to keep private data on their equipment while they travel.

Other travelers might prefer to encrypt and lock their phones or laptops, though it's not entirely clear which encryption tools authorities can defeat—officials reportedly paid a security company for a tool to break into the San Bernardino shooting suspect's iPhone, for instance, and it's possible federal officials have paid for similar keys to other popular phones and computers.

Travelers should also consider that officials may request their pass codes. Refusing can lead to lengthy questioning sessions, seizure of the devices, and for non-citizens it can heighten the risk of being denied entry.

"Folks should not use their thumbprint to unlock their phones—that should be disabled" when traveling she warns, since it can be easier legally and practically to obtain a fingerprint than a password. "If a border agent really wanted to be a jerk, they could grab your hand and place your thumb on your phone."

Expert mechanics can detect what's wrong with a car without lifting the hood. So can deep learning robots, says startup 3DSignals.

Sometimes, machines just don't sound right.

Brakes squeal, hard drives crunch, air conditioners rattle, and their owners know it's time for a service call. But some of the most valuable machinery in the world often operates with nobody around to hear the mechanical breakdowns, from the chillers and pumps that drive big-building climate control systems to the massive turbines at hydroelectric power plants.

That's why a number of startups are working to train computers to pick up on changes in the sounds, vibrations, heat emissions, and other signals that machines give off as they're working or failing. The hope is that the computers can catch mechanical failures before they happen, saving on repair costs and reducing downtime.

"We're developing an expert mechanic's brain that identifies exactly what is happening to a machine by the way that it sounds," says Amnon Shenfeld, founder and CEO of 3DSignals, a startup based in Kfar Saba, Israel, that is using machine learning to train computers to listen to machinery and diagnose problems at facilities like hydroelectric plants and steel mills.

And while most current efforts are currently focused on large-scale machinery, Shenfeld says the same sort of technology might one day help detect failures in home appliances or in devices like self-driving cars or rental vehicles that don't spend much time in the hands of an owner who's used to their usual sounds.

"When you have car-as-a-service, the person in the car doesn't know the car," he says. "If it sounds strange, you're losing this link with the human in the car deciding it's time to take it to the mechanic."

Initially, 3DSignals' systems can detect anomalous sounds based on physical modeling of particular types of equipment, notifying an expert mechanic to diagnose the problem. And once the problem is fixed, the mechanic's diagnosis is added to 3DSignals' database, which it uses to train its algorithms to not only detect unusual sounds, but also interpret them to understand what kind of repair is needed.

"The next time we hit this signature on the same machine for the same customer or another customer using the same type of machine, it will not just be anomaly detection," says Shenfeld.

And while 3DSignals focuses entirely on using its machine learning tools to process acoustic data—an area Shenfeld says is surprisingly neglected outside of speech recognition—other companies are using a variety of other types of data to detect and diagnose mechanical problems.

Systems from Augury, a startup with offices in New York and Haifa, Israel, monitor vibrations, temperature ultrasound, and electromagnetic emissions from machinery, typically large-scale heating and cooling systems. The company offers a portable tool a technician can use to capture a reading to an iPhone or Android device, where an app offers a real-time diagnosis as well as a continuous monitoring system, says cofounder and CEO Saar Yoskovitz.

"One of our customers, located in Ohio, had a thunderstorm at 2 a.m., and the whole building went dark," he says. "The generator was supposed to pick up automatically but it didn't, and the way that they learned about it was through an alert they got from Augury."

In more complicated situations, the system uses machine learning-based algorithms and a growing library of signals of different types of errors to suggest what's wrong and what technicians can do to repair the machine, he says.

"Over time, we've collected probably the largest malfunction dictionary in the world for our types of machines," says Yoskovitz.

For another startup, called Presenso, also based in Haifa, the exact type of data being monitored is less important than how it changes over time and what that signals about the machine's operation. The company's systems record data from sensors already installed in industrial equipment as part of existing control processes, streaming readings to Presenso's cloud servers.

"We don't care, or we don't need to know, if the sensor we're analyzing now measures temperature or voltage or flow," says CEO and cofounder Eitan Vesely.

Presenso's software tools use a variety of machine learning techniques to effectively build a model of a machine's operations based on the sensor data it receives. The tools provide visualizations of unusual readings, and how different they are from normal operations.

"They don't need any human guidance or [to] know what the physical attributes are that are being measured," Vesely says. "The goal is for them to learn by themselves how the machine operates."

And while real-time breakdown detection is obviously useful for companies operating machines in their own facilities, experts say it could also be useful to equipment vendors or insurance companies providing financial coverage in the case of downtime.

"If a company has downtime or business interruption insurance, that's something where it becomes incredibly relevant and also a money saver," says Ryan Martin, a senior analyst at ABI Research.

Having more reliable predictions of when machines need maintenance could also spur equipment makers to effectively offer time as a service, charging industrial users either extended warranties or even charging by the hour for the guaranteed use of their equipment without incurring too much risk to themselves, says Augury's Yoskovitz.

"One of the problems with this business model is they hold all the risk," he says. "If anything goes wrong, they pay for it, and they're looking for ways to lower this risk."

Since the election, media outlets are increasingly using tools that allow secure communication with sources. They're built to maintain anonymity, though extra precaution is advised.

Even before President Donald Trump took office, much of the public's understanding of the inner workings of his administration has come from what's been called an unusually large wave of leaks to the press.

Many of the leaks undoubtedly come from insiders with existing relationships with reporters—the anonymous senior administration officials familiar to readers of political journalism. But media organizations have also rolled out digital tools designed to allow would-be whistleblowers within the Trump administration, other layers of government, and private industry to communicate anonymously and securely with reporters. The stakes are potentially high for would-be sources, as Trump increasingly rails against leaks and vows to track down their sources.

The tools, such as the Freedom of the Press Foundation's SecureDrop and the GlobaLeaks platform backed by Italy's Hermes Center for Transparency and Digital Human Rights, harness encryption and the anonymizing network called Tor to let sources send secret messages and files to reporters without revealing their identities, locations, or IP addresses. And since Trump's election, a wave of media organizations have deployed this new infrastructure for secure communication, a kind of modern-day answer to the clandestine meetings in parking garages and secret signals made famous by such movies as All the President's Men.

"We've seen a massive increase in news organizations that want to run a SecureDrop since the election—so much so that we are trying mightily to keep up," says Trevor Timm, executive director of the Freedom of the Press Foundation. According to a directory maintained by the foundation, SecureDrop (which traces its history back to code written by the late internet activist Aaron Swartz) is being used by long-established publications, including The New Yorker, The New York Times, and The Washington Post, and by newer outlets like Vice, ProPublica, and BuzzFeed.

Univision's Gizmodo Media Group, which includes the millennial-focused Fusion and former Gawker properties Deadspin and Jezebel, has even taken out Facebook ads that target government officials and urge them to visit a site called TellOnTrump.com, The Wall Street Journal recently reported. The site offers sources a variety of ways to leak to Gizmodo's investigative reporters, including SecureDrop, encrypted email, and encrypted messaging tools.

The recent wave of interest was spurred on by Trump's openly hostile attitude toward the press and his early restrictions on public statements by many federal officials, Timm says. And, he adds, many outlets took a serious look at SecureDrop after seeing The New York Times adopt the platform soon after Trump's election. "Generally, when The New York Times does something, a lot of news organizations follow," he says.

Even before Trump, consequences for sources of even well-intentioned leaks can be quite serious. Former National Security Agency contractor Edward Snowden was essentially forced into exile in Russia for whistleblowing, and at the time of Chelsea Manning's scheduled release in July, the former soldier will have spent seven years in custody for transmitting a massive trove of documents to WikiLeaks. President Barack Obama commuted Manning's sentence from an initial 35-year term, but his administration presided over what transparency experts have said was an unusual number of leak prosecutions under the federal Espionage Act.

That's a trend that may continue under Trump, who has already received support from Congressional Republicans for potential plans to root out who's been leaking within his administration and who, years before taking office, had called for Snowden's execution.

"There's a high risk of retaliation when someone becomes a whistleblower," says Betsy Reed, editor-in-chief of The Intercept, an early adopter of SecureDrop. "It can rise to the level of an Espionage Act prosecution, but it can also be the loss of a job in the private sector."

Using a platform like SecureDrop or GlobaLeaks doesn't eliminate the risk that leak sources can be discovered, advocates stress. And developers urge sources to take extra precautions beyond Tor, like only transmitting files over a public wireless network and using Tails, a secure operating system that comes bundled with Tor and can be launched from a USB stick or DVD, leaving no trace on a source's laptop. Some people may also wish to use a cheap throwaway laptop to make contact, says Matthew Green, assistant professor of computer science at Johns Hopkins University's Information Security Institute.

Sources should also consider the fact that particular documents or facts might only be accessible to a few people, potentially making the identity of whoever leaked them easier to determine, Green says. And sources should consider printing and scanning digital documents before leaking them, or at least generating PDFs from original files like spreadsheets and word processing documents, he says. "Never give out a direct electronic copy of the document," Green says. "Things like Microsoft Word documents can have [identifying] metadata."

But using Tor ensures that journalists, and even internet service providers who could get search warrants or subpoenas from authorities, can't detect who's transmitting data and documents.

"The reality is that when a reporter's source can be identified through digital traces, the prosecution does not even need that reporter to testify," wrote Charles Berret, a fellow at Columbia Journalism School's Tow Center for Digital Journalism, in a report on SecureDrop.

Organizations using SecureDrop don't even open documents they receive on internet-connected computers, instead moving them to a second disconnected machine before decrypting them, so there's less chance of them being stolen by cyber eavesdroppers.

SecureDrop and GlobaLeaks also gives journalists and sources the option of sending messages back and forth, which users say can be more valuable than one-off drops. And while publications don't always disclose which stories were reported through leaking tools, outlets including the Associated Press and The Intercept have credited stories to SecureDrop leaks. GlobaLeaks has been used by outlets across Europe, Africa, and Latin America to develop stories, says Fabio Pietrosanti, president of the Hermes Center.

"GlobaLeaks play an essential role for AWP in what regards technological empowerment," says Pedro Noel, editor of the Brussels-based Associated Whistleblowing Press, in an email to Fast Company. "Without them, our organization would have very different, probably weaker, potential of action."

The AWP operates sites around the world, including Ecuador Transparente, which has reported on homegrown state surveillance, and Iceland's Ljost, which published reports of police surveillance in the country and loans to insiders at Iceland's failed bank Glitnir.

"There are maybe 40 well-known installations of GlobaLeaks that are managed by different organizations," says Hermes Center fellow Marco Calamari. That number includes the installations used for internal anonymous whistleblowing at government agencies and private companies. He emphasizes, though, that the tool's developers may not know the full list of organizations using it.

There's also no way to definitively say whether journalists overall have received more tips through whistleblowing tools since Trump was elected or since he took office. But the Freedom of the Press Foundation's Timm says he's heard anecdotal reports from journalists that they're receiving more messages through SecureDrop. The foundation is planning fundraising efforts to effectively double the current staff of five working on the tool in order to roll out new features and support existing users.

And while Reed declined to reveal the number of leaks received through SecureDrop at The Intercept, she says the type of leaks has changed with Trump in the White House.

"I definitely can say that the quality of information is greater," she says. "It's consistently more interesting and in the public interest."

The (contested) refugee ban has led to the creation of several new apps designed to help refugees, who often carry smartphones.

For refugees escaping conflict in the Middle East and elsewhere, simply communicating with people in new countries can be a big challenge.

Groups helping refugees in Canada and the United Kingdom have found language barriers make it harder for newcomers to integrate into society and find work. And even in some refugee camps, aid workers who speak the same language as the people they're there to assist can be in short supply, says Abubakar Abid, one of the cofounders of Tarjimly, a new Facebook Messenger bot designed to connect refugees and people assisting them on-site with volunteer, real-time translators who might be somewhere else on the globe.

"All they do is message the [Facebook] page and they are routed to whatever translator is available in the language they are seeking," Abid says. "They can send text messages, snippets of text, voice notes, pictures, or whatever media they like."

Tarjimly's cofounders, who are Massachusetts Institute of Technology alumni, were inspired by a friend's experience volunteering at a camp in Greece. "He went there expecting to help with operations or logistics," Abid says. "He actually spent more than 90% of his time serving as a translator because he happened to be able to speak both Arabic and English."

The founders announced the project January 30 and have since received sign-ups from more than 1,500 volunteers interested in serving as translators. They plan to test the tool with small groups of translators and users over the next few weeks and aim for a wider release in early March. They're also exploring options for funding, including accelerators and incubators focused on social enterprises. Abid believes that using the Facebook platform to handle message routing should make it easy for the app to scale.

Tarjimly is just one of many apps and digital tools aiming to help refugees communicate, navigate, and find resources as they make their way in unfamiliar places. While refugees often leave their homelands in haste with limited possessions, those traveling from Syria and elsewhere in the Middle East have typically made sure to bring mobile phones. They're often packed in a waterproof bag alongside identity papers, says a journalist based in Germany who operates the online directory AppsForRefugees.com under the pseudonym "John Tucker." He says he adopted the name to avoid online harassment from anti-refugee trolls.

And since refugees began arriving in large numbers in Europe, software developers have sought to build apps to assist them. Tucker says the most visited apps from his site have included phrasebooks and dictionaries. Other popular apps include those that connect refugees with people who are offering items like furniture and household items and another that puts them in contact with locals looking to cook meals together.

"The people [arriving] here in Germany had no idea how to deal with authorities, with daily life, with the language, and so on," Tucker says. "There are lots of apps on the market and websites which try to fill these gaps to connect these people with each other and give them some information about the country they're in."

In the United States, aid to refugees took on new urgency for some activists after President Donald Trump's executive order, now blocked by a federal appellate court, banning new refugees and people from seven majority-Muslim nations.

The London-based makers of the Refugee Aid App, which catalogs assistance available to refugees in particular areas, quickly deployed the app in 19 U.S. cities after Trump's order went into effect. "Our whole team just kind of worked round the clock to call these different organizations in these cities where there were airports where people were getting stranded," says Shelley Taylor, CEO of Trellyz, the company behind the app.

The app, also known as RefAid, provides a menu of assistance, from legal advice to nearby health care. Taylor and her colleagues wanted to make sure that information on legal support would be available to refugees stranded in U.S. airports, provided they had time to access their phones before being detained. "We wanted to make sure that they could make phone calls to local organizations to get legal help before too many bad things happened to them," she says.

Trellyz works with aid organizations around the world to help spread the word about their offerings and to assist them in digitally coordinating with each other. The company also relies on those same organizations to inform potential refugee users about the app's existence. According to an October blog post published by the Office of the United Nations High Commissioner on Refugees (UNHCR), many refugees prefer to receive information from speaking directly to aid organizations, rather than from digital sources, especially with legal issues and other factors in flux.

That's especially true after some apps released in haste during refugee crises reportedly failed to deliver on their promises, which AppsForRefugee's Tucker acknowledges can be an issue. "It's not always clear where these apps come from, and the quality is very diverse, let's say that," he explains.

But developers willing to ensure that their offerings actually match the needs of refugees and the people working with them can still make a difference, according to the UNHCR post.

"There's also just an incredible desire of people around the world to help refugees," says Abid. "And I think being able to tap into that is very, very valuable."

Apps for podcasting, tracking ovulation, and finding a lawyer extend a nonprofit's mission of supporting democracy in Iran.

There are an estimated 40 million smartphone users in Iran. But ongoing government censorship, language barriers, and economic sanctions imposed on the country have made the app ecosystem that's grown up elsewhere in the world unavailable to most Iranians.

"It's almost impossible for Iranians to pay for any apps," says Reza Ghazinouri, a program director at the Bay Area human rights organization United for Iran. "Iran's economy is still isolated from the international economy."

And while isolation has at times been a boon to some Iranian and Iran-focused entrepreneurs—the country's mostly cash-based economy has even spurred a cash-focused ride-hailing app—access to things like government and health information is still limited, partially due to censorship by the country's authoritarian regime.

"There are hardly any apps that are civic-minded that are in Persian," says United for Iran's director, Firuzeh Mahmoudi.

To try to change that, the organization launched what it calls the IranCubator—an incubator that provides funding and support to developers building software to "advance civil society" in the country. It's part of a wider push by developers and nonprofits around the world to create content deliberately aimed at populations underserved by the existing app ecosystem.

This month, the incubator released its first app—a podcast tool called RadiTo, with menus in Farsi, as well as underserved Iranian languages such as Balochi and Kurdish—and plans to deploy additional smartphone software in the coming months.

Like many podcast apps, RadiTo lets listeners download listening material through their home internet connections instead of through expensive mobile data connections. But the app features Iran-focused content from around the world, including some voices and positions that aren't typically heard in the country, Mahmoudi says.

A series called Taboo, for instance, focuses on atypical-for-Iran topics from separatist movements to the female orgasm. Another podcast, called RadioGeek, discusses issues of technology and society from around the world. Still, the group does have some limits on the podcasts it will include, in line with Iranian sensibilities: Explicit language can be a no-no, and one suggestion that could promote sectarian conflict was rejected.

To limit censorship, recordings are encrypted and downloaded from public cloud servers used by other internet services, making the app difficult to block. With the Iranian internet routinely restricted, the country's web users are no strangers to tools like encryption, proxy servers, and virtual private networks to get around pervasive censorship. In fact, the IranCubator team plans to promote RadiTo through ads on Psiphon, a circumvention tool developed by a Toronto-based company and widely used in Iran, says Ghazinouri.

Other organizations around the world are also focusing on making smartphone apps more useful at solving problems for new sets of users. In Saudi Arabia, a local foundation launched an app last year designed to help volunteers assist poor residents with health issues and information, and a Hong Kong businessman active in Africa recently sponsored a set of apps designed to help people in poorer countries diagnose and treat common eye conditions. And even internet giants like Google have begun making a serious effort to adapt their tools to the tastes and internet connection speeds of underserved parts of the world.

Mahmoudi says IranCubator's efforts could also serve as a "pilot case" for building similar apps for other places in the world. And while the IranCubator developers aren't physically on the ground in the country they serve, that can actually be an advantage, given the safety issues they might face in Iran.

"There's a very specific niche that we can play that can't be played inside the country," she says. "We have capacity that doesn't exist in the country, and we have security."

IranCubator's apps themselves are carefully audited by external testers for security, and the incubator is keeping quiet about its ties to some projects for safety reasons, he says.

"Some of them we have to be quiet about and don't take credit for them, because the subject is a little bit sensitive, and we don't want to be putting users at risk by associating users with a foreign-based nonprofit," he says.

United for Iran is also reluctant to reveal too much detail about where it receives its funding, in order to protect donors and their families in Iran and to avoid suggestions of ties to groups controversial with the government in Iran. (Commentators have occasionally suggested the group may be linked to U.S. attempts to influence public opinion in the Middle East.) Mahmoudi says funding includes foundation and individual support, as well as employer donation matching from tech companies like Google. The group also receives support from the U.S.-backed Open Technology Fund, which also funds work on well-known secure digital communication tools like Signal and Tor.

Among the next public IranCubator apps to be released is a women-centric tool called Hamdam (Persian for companion), says Mahmoudi.

"It's in Persian, and it's an ovulation and menstruation calendar, which is the first app of its kind in Persian," she says. But the app will also include information about other women's issues, from sexual health and birth control to preserving women's rights in marriage. Traditionally, women getting married in Iran forfeit many of their rights, but some of those can be preserved by putting particular legal language into a marriage contract, she says.

"It's very much a contract that you sign—anyone can write in whatever they want," says Mahmoudi, and some of that sample language will be included in the app.

Another app, built in conjunction with a human rights organization called Harana, will feature a database of lawyers across Iran and their specialties, she says.

"For example, if you are arrested for drug-related accusations in a city in the southern part of Iran, these are the lawyers you should go to," she says. "If you are arrested in Tehran for political reasons, these are the lawyers you should go to."

Users will also be able to report human rights violations, and Harana will work with its in-country sources to investigate claims, Mahmoudi says. In general, making sure government officials or others looking to sabotage the apps don't spam databases with fake reports is part of the process of securing the apps.

United for Iran already maintains an online tool called the Iran Prison Atlas, cataloguing how the country treats its political prisoners, and the organization can often focus on providing information that just wouldn't be possible inside Iran itself.

Once the initial versions of the apps are rolled out, Mahmoudi says, the project faces a problem that will be familiar to all app developers: keeping everything up to date. "As soon as the app is done, you have to work on version 2.0," says Mahmoudi.

The same technology that Pindrop uses to flag phone scams could soon let you unlock your car or phone with the unique sound of your voice.

Someday soon, you may be able to unlock your car or even log in to your favorite streaming app on a hotel TV simply with the sound of your voice.

Pindrop, an Atlanta company that now primarily offers sound-based fraud detection tools for call centers, plans to release a service later this year that will let connected devices verify who they're talking to, turning the human voice into a combination of a username and password.

"Everybody has a unique voice, and everybody has a unique behavior in the way they say things," says Pindrop cofounder and CEO Vijay A. Balasubramaniyan.

Secure voice recognition login could make it safer to conduct complex transactions through digital assistants like Apple's Siri and Amazon's Alexa—and it could prevent scenarios in which loud TV advertisements or children playing with devices accidentally give commands (and buy products) via such devices. And if different devices adopted Pindrop or another common voiceprint provider, users wouldn't have to separately program each device to recognize their individual speech patterns and could log in to new devices simply by speaking, says Balasubramaniyan.

The technology should also be able to detect when people are distorting their voices or playing recordings of other people speaking—fraud techniques he says they already encounter in the call center market, where Pindrop reports it works with eight of the top ten U.S. banks and two of the top five insurers to detect phone scams. Call center customers pay based on call volume, and Pindrop will likely roll out similar cost structures for IoT device makers he says.

For phone fraud detection, Pindrop's systems don't just listen to the sound of callers' voices as they dial in to place orders or transfer funds—it also uses other audio data to determine where people are calling from and the type of phones and networks they're using. Different models of phones introduce their own acoustic signatures into conversations, Balasubramaniyan says. And phone networks in different parts of the world transmit different sound frequency ranges based on different requirements for balancing bandwidth consumption and voice quality. Even internet calling tools, like Skype and Google Voice, break conversations into different-sized audio packets, making it possible for Pindrop's algorithms to tell them apart.

"Every time you drop a packet, you're having a break of 30 milliseconds as opposed to 20 milliseconds or whatever else," Balasubramaniyan says.

That lets the company alert its clients when calls with strange characteristics come in, like if a number registered to a U.S. cellphone carrier is actually dialing through Skype or through a telephone network in Nigeria. In other cases, the company has detected calls ostensibly from different customers and numbers all being placed by the same fraud ring, he says.

"We'll see, for example, hundreds of calls trying to access hundreds of accounts, but they all come [with] the same 147 characteristics, and it's actually the same device," he says.

Pindrop generates numeric and color-coded alert levels for its clients, letting them serve trustworthy callers more quickly and handle potential imposters with more care. Some might ask additional verification questions of users dialing in from unusual connections or even ask to call them back at their numbers on file. And others use the Pindrop score information simply for offline processing, deciding whether to approve transactions like wire transfers based on the indicated risk level.

While fraudsters do try new techniques over time, Balasubramaniyan says Pindrop is often still able to pick up on new patterns of behavior—like routing calls through hacked telephone systems in less suspicious countries—based on the new acoustic signals they generate.

"It looks even more different than regular traffic," says Balasubramaniyan.

The company can also detect known criminals connecting through new techniques based on their historic behavior patterns and voice data.

"We have created the world's largest database of well-known voice fraudsters," he says.

The company even operates a digital call center of its own, buying from carriers the phone numbers that their customers have traded in due to excessive numbers of fraudulent calls. Pindrop's system receives about 90,000 fraudulent calls per day across those numbers, generally playing randomized recordings of phrases like "I can't hear you very clearly" just long enough to keep callers on the line and fingerprint their connections.

All of the human voice data the company has captured will likely help make its IoT product more robust in preventing fraud, Balasubramaniyan says.

"Right now, we already see people who distort their voices," he says. "Just knowing how voices sound over a massive dataset, that allows us to use that knowledge to also do this better."

If you're looking to serve your country with science—but not don the uniform—Uncle Sam has a job for you.

Hey millennials: Uncle Sam wants you ...No uniform required.

Looking to tap cyber talent in the private sector, the U.S. Navy has been holding a series of hackathons. The challenge: Find a way to crack the security systems of military drones and warships.

Earlier this month, the Naval Postgraduate School's Center for Cyber Warfare let hackers test their skills against a "boat in a box"—a testbed system built by contractor Booz Allen Hamilton to mirror fleet systems. The event was a follow-up to one last June, dubbed HackTheSky, where security experts tried their skills against aerial drone control software and interface designers came up with prototypes to make it easier for drone pilots to operate the devices.

The goal is to tap into a talent base of mostly young civilian cyber experts who are interested in ways to help keep the country safe without necessarily enlisting in the Navy or joining a traditional defense contractor, says Commander Zachary Staples, director of the Center for Cyber Warfare.

"They don't want to cut their hair and go to bootcamp—that was not what they were asking me," he says. "They were asking, how could we work with you on a project sort of basis."

The Navy hackathons are purposely held in tech hubs, not at Navy bases: The first was held at Galvanize's San Francisco startup center, and the most recent one at Austin's Capital Factory coworking space.

The "boat in a box" will also be the focus of a presentation next month by Booz Allen Hamilton at South by Southwest, the Austin festival that's typically more known for appearances by Silicon Valley startups than old-school defense contractors.

The hackathons aren't the first attempt by the defense establishment to crowdsource innovation: The Defense Advanced Research Projects Agency (DARPA) has long held competitions to spur development in technology, from predicting the spread of infectious disease to building autonomous cars. And the Defense Department last year held a "Hack the Pentagon" challenge, paying out more than $150,000 in bug bounties to hackers who found vulnerabilities on public-facing defense websites.

But the Navy hackathons are among the first to bring hackers to test their skills against real-world sea and air digital systems that are increasingly important to national security, Staples says.

"At any given time, a larger and larger percentage of the world's tangible wealth is loaded on a ship or being unloaded at a port," Staples says. And if the systems onboard those ships prove vulnerable, sophisticated attackers could essentially launch a blockade with the touch of a few buttons.

"If a fleet of ships gets hacked, that's a problem for the U.S. Navy, because our nation has just been embargoed," says Staples.

The boat in a box will let hackers test their skills against actual nautical communication interfaces, from the automatic identification systems used for collision avoidance to weather satellite radio systems.

June's drone hackathon helped unearth some vulnerabilities in systems that allow one operator to control massive swarms of drones. In 2015, a Naval Postgraduate School team launched a swarm of 50 drones under the control of one pilot, using customized software to keep the devices in sync. The hackathon helped inform architectural decisions for the next generation of the software, Staples says. And one team from the event also contributed some new security improvements to the Robot Operating System, an open source robotics toolkit used by researchers and robot makers around the world.

"One team contributed that back to the ROS codebase so that the result of the Navy's swarm hackathon, at least one of those results, was available to the entire robotics community," says Staples. The code related to technology for digitally verifying that software running on a robot hasn't been tampered with, he says.

And in addition to security tweaks, the hackathon generated improvements to the digital interfaces used to control the swarming drones. Finding skilled user interface designers can be a challenge for the Navy, just as it is for many in the private sector developing apps and websites. Building more intuitive controls can mean happier sailors and less time and money spent on training, Staples says.

"If I'm going to have a swarm drone operator, it's going to be a 25-year-old kid," says Staples. "They have an expectation that the interface is going to be as good as the apps they have on their iPhone."

The design challenge was one of a series run by Booz Allen Hamilton in conjunction with TopCoder, the Connecticut-based company known for running programming and design competitions. It's part of a push by the contractor to engage with outside innovators, partly inspired by the example of the U.S. Digital Service and 18F, the two startup-inspired federal initiatives to recruit technical talent of a type not always found within the traditional civil service.

"We are realizing that the way we support our clients needs to be a little more inclusive and collaborative," says Brian MacCarthy, the head of Booz Allen Hamilton's Strategic Innovative Hub in San Francisco.

At the Austin hackathon, a design challenge will focus on alternatives to GPS for areas, such as in tunnels and buildings, where the technology gets limited reception, and a data science challenge will explore ways to use detailed data on maritime traffic to detect security risks and criminal activity such as human trafficking.

And for the Navy, the hackathon will be part of an ongoing innovation challenge of its own, to find ways to connect with hackers and small businesses willing to help keep its systems state-of-the-art.

"The focus of the hackathon this weekend is building a community for maritime security," says Staples.

AI tools could make finding jobs faster and more equitable–and cut back on recruiter spam, advocates say.

It’s no secret that finding good workers isn’t easy, particularly in the competitive tech world.

Read Full Story

MIT Researchers are using neural networks and haptic feedback to make conversations easier to navigate.

Wearable electronic devices today are, for the most part, toys for narcissists.

Read Full Story

Travelers looking to keep device data private at border crossings need to rely on technology, not the law.

Anyone who has traveled internationally knows the drill: body scans, bag searches, and lots of questions. But there’s something else U.S. Customs and Border Protection officers increasingly want to scour: your cell phone.

Read Full Story

Expert mechanics can detect what’s wrong with a car without lifting the hood. So can deep learning robots, says startup 3DSignals.

Sometimes, machines just don’t sound right.

Read Full Story

Since the election, media outlets are increasingly using tools that allow secure communication with sources. They’re built to maintain anonymity, though extra precaution is advised.

Even before President Donald Trump took office, much of the public’s understanding of the inner workings of his administration has come from what’s been called an unusually large wave of leaks to the press.

Read Full Story

The (contested) refugee ban has led to the creation of several new apps designed to help refugees, who often carry smartphones.

For refugees escaping conflict in the Middle East and elsewhere, simply communicating with people in new countries can be a big challenge.

Read Full Story

Apps for podcasting, tracking ovulation, and finding a lawyer extend a nonprofit’s mission of supporting democracy in Iran.

There are an estimated 40 million smartphone users in Iran. But ongoing government censorship, language barriers, and economic sanctions imposed on the country have made the app ecosystem that’s grown up elsewhere in the world unavailable to most Iranians.

Read Full Story

The same technology that Pindrop uses to flag phone scams could soon let you unlock your car or phone with the unique sound of your voice.

Someday soon, you may be able to unlock your car or even log in to your favorite streaming app on a hotel TV simply with the sound of your voice.

Read Full Story

If you’re looking to serve your country with science—but not don the uniform—Uncle Sam has a job for you.

Hey millennials: Uncle Sam wants you …No uniform required.

Read Full Story

GE insists its San Diego project—part of a global ‘smart city’ push, with connected sensors, cameras and microphones—will protect privacy.

Thousands of new “smart” streetlights to be rolled out across the city of San Diego will be connected to a city-wide network, allowing officials to adjust their brightness and detect outages from afar. But in addition to giving off light, they will also be pulling in information about what’s going on around them.

Read Full Story

This technology can help spot patterns that uncover cases of modern-day slavery, whether in sex work or sweatshops.

Recent United Nations estimates indicate that approximately 21 million people around the world are trapped in forced labor situations, including about 1.5 million in North America.

Read Full Story

Information imbalances benefitting Uber and other services might need new solutions.

In less than a decade, so-called sharing-economy businesses have reshaped the ways many Americans live their lives.

Read Full Story