In an effort to reach the "next billion users," the internet giant is rolling out redesigns of apps and expanding access to speedy Wi-Fi.

Critics of Silicon Valley sometimes say that tech companies focus too much on the needs of the types of people they hire, but Google and its parent company Alphabet are increasingly looking to cater to the "next billion users" of the internet.

Every hour, more than 10,000 people in India alone come online for the first time, the company estimated in a September blog post. They're typically using inexpensive phones with slow connections and limited prepaid data plans, but once they get online, they're looking to do the same things as users in Google's more established markets.

"They do similar stuff to what we do," says Caesar Sengupta, a Google vice president in charge of the company's Next Billion Users initiative. "They're searching, they're on social networks, they do a lot of video."

To make that possible, and make its services more usable in emerging markets, Google is redesigning some of its apps to use less bandwidth, working to expand access to speedier Wi-Fi connections, and talking to potential users to develop products that meet their needs.

"This is the first time for many of us that we're building a product that is not targeted at us," says Jay Akkad, product lead for the Next Billion Users project at YouTube.

The company rolled out a lighter-weight version of its search results page to serve to users on slow 2G or 3G connections, and offers a Data Saver feature in Chrome for Android that can reduce the size of web pages by 60%. The optional feature routes unencrypted web traffic through Google's servers, which compress images and minimize the HTML, CSS, and JavaScript code that defines web pages, removing whitespace, comments and other unnecessary content without changing the page itself.

Google has also rolled out Google Station, a project to bring Wi-Fi access to railroad stations and other public places across India. Currently, about 3.5 million people use the service every month, and Google plans to expand to additional stations across the country, the company said in September.

Google isn't alone in this endeavor: Facebook recently launched a service called Express Wifi, which allows local entrepreneurs to sell inexpensive internet access, after a previous initiative to deliver free access to a limited set of apps and websites was banned by the Indian government under net neutrality rules.

Users in India often download webpage content to their phones for later reading when they have Wi-Fi access, Sengupta says, and the company has added support for downloading geographical data to Google Maps for offline use—something that's now proven popular around the world, he says. The Google Play store now also caches some content when users connect to Wi-Fi, so they can continue to browse for apps even with limited connectivity, then queue up apps to download when they next have Wi-Fi signals.

In some cases, Google is making more fundamental changes to how its software operates to make it more compatible with the needs of users in new markets. For YouTube, that meant talking to hundreds of consumers in India about how they used the internet and online video, then showing them paper mockups and working prototypes to make sure their ideas were on the right track.

"We have a fantastic UX team who has gone to India basically every month or every other month for the past couple of years," Akkad says. "What we ultimately gathered was that what YouTube needed to do for these users and what these users needed from YouTube was fundamentally different from what YouTube offered at the moment."

Users had trouble streaming video over slow connections, and YouTube engineers initially built a prototype delivering highly compressed versions of videos. But when they showed it to potential users, they were dissatisfied with the poor quality of the reduced-size videos.

"We built a prototype of this," Akkad says. "Instantly, everyone was sort of like, 'I don't want an inferior experience.'"

It turned out that many potential YouTube viewers had already developed a solution for watching mobile videos: They downloaded content to their phones when they did have access to speedy internet and later swapped movies with other users.

"People build social capital by having the latest and greatest videos and being able to transfer them by Bluetooth to their friends," Akkad says.

To cater to those users, the company built YouTube Go, a new app for Google's Android operating system, which makes up 97% of India's smartphone market, according to an August report from research firm Strategy Analytics.

The app lets users download videos in a variety of resolutions when they have Wi-Fi, to optimize for quality or conserve space, and even wirelessly transfer saved YouTube videos to friends. It also includes brief previews of videos, so users can sample before they download, and includes short, six-second ads that don't take up too much space or bandwidth, says Akkad.

The company is gradually rolling out YouTube Go to more users, with plans to launch broadly in India early next year and potentially expand to other countries in the future.

"Virtually every month there's one or two studies that are ongoing with our target audience in India in order to make sure that we're on the right track," Akkad says.

For home seekers and job hunters, inaccurate background information can be devastating. And getting errors corrected can be tough.

After spending several years on a waiting list for a subsidized apartment in Tennessee, "Jack" says he was turned away by a leasing agent when incorrect information turned up in a routine background check.

"She's like, 'You have a sex offender charge on your record,' and I was shocked," says Jack, who requested that we not use his real name to avoid having it further associated with the erroneous allegation. "I never have been charged with that."

The agent showed Jack a copy of the report, which referred to a man with a similar name and birthdate, he says. The report also included a photo of the alleged sex offender, who the leasing agent acknowledged clearly wasn't him. (Among other things, the man pictured had conspicuous face tattoos.)

But Jack still lost his spot on the waitlist—he's currently staying with a relative—and had a separate housing application also declined because of the mistaken information. As he tries to correct the record and clear up the confusion, every day is a new day of limbo.

Vidhi Joshi, an attorney at the Legal Aid Society of Middle Tennessee and the Cumberlands who is representing Jack, filed a formal dispute with the screening company that included the alleged offender's picture but hasn't yet received a response. She has also requested a copy of the report used for the second rental application, which was issued by a different background check firm, she says. (Joshi declined to identify the firms involved, since their responses are still pending).

Jack's story is just one of what fair credit advocates say is a growing number of cases where incorrect information on criminal and other background checks is making it difficult for innocent people to find work and housing. Such checks are routinely required for employment and rental applications, and conducted with varying levels of diligence by hundreds of companies around the country, but they only work properly when the information they contain is accurate.

And, increasingly, that's not the case.

"We have gone from back in the early part of the century seeing a couple hundred [cases] about criminal records a year up to more than a thousand a year," says Sharon Dietrich, litigation director at Community Legal Services of Philadelphia and head of the legal aid group's employment unit.

Digitizing data like court records has made background checks significantly cheaper and faster than in years gone by: In 2012, a Society for Human Resource Management survey found that more than two-thirds of organizations polled conducted criminal background checks on job candidates.

"I think employers are scared not to do it because they feel like they may be found liable for negligent hiring," says Dietrich.

But regulation has not necessarily kept pace. Even some jurisdictions that have passed so-called "ban the box" legislation, which forbids employers from requiring potential hires to indicate on job applications whether they've been convicted of a crime, still allow criminal history checks later in the application process.

Companies typically use a third-party background screening service to verify the work and education histories of potential hires, and often to check a candidate's credit history and search for any criminal convictions that might serve as a red flag, says Mike Aitken, vice president of government affairs at SHRM.

According to a May report from business intelligence firm IBISWorld, the $2 billion background check industry is mostly composed of small, local firms, which have benefited from easy internet access to more and more public records. Fear of liability for employee misconduct and post-9/11 security concerns have also contributed to rapid growth in the industry in recent years, according to the National Association of Professional Background Screeners, an industry group which counts more than 850 background screening companies as members.

And while the majority of checks are likely accurate, Dietrich says she's seen a "good number" of cases where background reports include information about people with similar names to her clients. In one case, she says, a client with a common name received a report with 65 pages of criminal history data about unrelated people with similar names.

"I have even seen cases where a female's record is attributed to a male or vice versa, simply because they don't use gender information, even though they have it," she says.

The Devil In The Details

Even job applicants who do have criminal histories can still be unfairly harmed by slipshod background checks. Often, convictions legally expunged by a court still show up on reports. Crimes can also be misclassified, such as when misdemeanors are erroneously labeled as felonies, Dietrich says.

One problem is over-reliance on commercial databases, which can contain incomplete information or return hits based on false matches, says Larry Lambeth, the president of screening firm Employment Screening Services.

Lambeth is also the founder of Concerned CRAs, an association of consumer reporting agencies, as screening companies are known under federal law, voluntarily adhering to certain ethical standards. The group requires its members to consult original documents, like court records, and compare information like birthdates and addresses to verify data from databases actually corresponds to the subject of a background check.

"If you use a database for his criminal records, if you find a hit, we require that you pull a criminal record from the courthouse and find the identifiers and ensure it's the right person," he says.

But not all background screeners conduct such rigorous research, and not all clients are willing to pay for it, he says. And even well-intentioned screeners can have difficulty verifying records in jurisdictions where privacy-minded courts redact personal information like addresses and dates of birth, says Melissa Sorenson, executive director of the National Association of Professional Background Screeners.

"It's extending the time frame of getting people to work," she says, since it takes screeners longer to complete reports for employers. The organization generally tries to work with state and local regulators to make sure background screeners have access to the data they need to do their jobs, she says.

Under the federal Fair Credit Reporting Act, applicants do have the legal right to receive copies of their background checks and to contest any inaccuracies with the agencies that did the screening. But in practice, critics say the process can be arduous even for experts, let alone for individuals looking to correct their own records. Dietrich says she's had a screening firm resist providing her with an address to send mail on behalf of a client.

"You have to know even where to look—look for a little link at the bottom [of a firm's website] that says 'consumers,' and maybe that's where you'll find the dispute procedures," says Dietrich. And those procedures often require documentation like proof of address that not everyone can easily provide, she says.

"Some people can do that, some people can't, especially my clients who tend to be low-income and on the move," she says.

Even in cases where data is obtained directly from government sources, it can still be incorrect or out of date: Former U.S. Attorney General Eric Holder, now a partner at law firm Covington & Burling, wrote letters on behalf of Uber to local legislators earlier this year, urging them not to require ride-hailing services to vet their drivers against the Federal Bureau of Investigation's fingerprint database.

Critics have long said that the FBI database, which many states use to vet childcare workers, health care professionals, and others in regulated occupations, often contains wrong, misleading, or incomplete data. The database contains arrest records from law enforcement agencies around the country but only includes final case outcome data roughly half the time, according to a 2013 report from the National Employment Law Project. That means that arrests that resulted in an acquittal, dropped charges, or even expunged records can essentially appear as unresolved.

And updating outdated or simply incorrect data in the FBI's files can be laborious and can leave jobseekers out of work while they reach out to multiple agencies to get the information fixed, says Maurice Emsellem, director of the NELP's Access and Opportunity Program.

"Basically, you have to go back to the state that created the record to get it fixed," he says. "The FBI won't fix it in the FBI system."

And while reforms to the FBI background check system have been proposed as part of bipartisan criminal justice reform legislation, they've yet to make it through Congress.

Nor, says Dietrich, have regulators addressed issues that have arisen around private background checks in the more than 45 years since the Fair Credit Reporting Act became law, such as how background screeners should handle record matching issues in computerized databases or how they should deal with potentially expunged cases.

"The world of regulation of these companies is not what it might be," she says.

Meanwhile, the promise of future regulation is cold comfort for home-seekers like Jack, whose lives are being immediately damaged by bad data in a data-obsessed world. "He has suffered from homelessness," Joshi says. "It's had a pretty big impact."

Security experts warn of new types of malware that threaten to publish instead of encrypt valuable, confidential information.

Right when internet users have learned to be wary of malware that encrypts files and holds them for ransom, security experts are warning that digital extortionists are taking more aggressive steps to get paid.

"You're seeing different techniques with the goal of improving the conversion rates of people actually paying," says Jerome Segura, lead malware intelligence analyst at the security firm Malwarebytes.

Instead of simply encoding files so that users can't access them, some blackmailers armed with a new kind of malware called doxware are threatening to leak potentially sensitive files to the public if a ransom isn't paid, says Chris Ensey, COO of Dunbar Security Solutions.

"This is a very recent change in the tactics they're using," he says, noting that they've appeared only within the past few months.

Dunbar has yet to see malware make good on threats to leak data, and Ensey says that at least some variants appear to display fake progress bars purporting to show data transfers to attackers' servers without actually uploading any files. Storing and leaking files is logistically more difficult than just encrypting them on victims' own computers, experts say.

But Ensey predicts that by next year there will be actual data leaks attributed to ransomware, if only to motivate more attack victims to pay the ransom.

"I would not guess that we're far off from public examples of that," he says.

Previously, security experts advised companies and individual users to make regular backups of important files so they'd be ready to restore them if they were encrypted or damaged by malware. But that's of less help if malware creators instead threaten to distribute information, potentially exposing companies to liability, or individual users to embarrassment or risk of identity fraud, he says.

"My thinking now is that organizations really have to focus on: How do we isolate sensitive or private information from places where ransomware tends to find itself?" he says. "You have to make it so it's incredibly hard for that ransomware to touch or gain access to any kind of sensitive data through a standard channel."

Preventing leaks by computers infected with malware is ultimately similar to protecting data against insider threats. That means that organizations shouldn't simply have an unencrypted network drive with confidential materials like sensitive business plans or medical records, Ensey says.

Earlier versions of ransomware have already struck institutions with large troves of mission-critical, confidential information, such as hospitals, which could be motivation enough for entities to pay to keep patient records from falling into the wrong hands. But individual consumers represent the bulk of ransomware victims, according to a report released in April by the security firm Symantec. People could feel forced to pay to safeguard anything from financial and medical documents to explicit pictures, particularly if ransomware attacks on smartphones become more common.

"The variants that are out today are mostly Windows-based, so it's desktop computing," Ensey says. "If they can adapt it to mobile, I think then you might have an audience for this that would in fact pay the ransom."

Ransomware creators have recently gotten more aggressive in other ways, too, according to Segura, sometimes actually permanently deleting files rather than leaving them encrypted if victims don't quickly pay up. Some malware varieties have also focused their energies on particular classes of files likely to be of interest, such as spreadsheets, and future attackers may well use more sophisticated prices to determine how much ransom to charge.

"It's a business decision. Like marketers, how do you [set] the price?" Segura says. "Finding the sweet spots where people are willing to pay is really important to the economics of the ransomware business." That might mean charging more when it comes to victims with more apparent business documents or photos, or adjusting ransom amounts for targets in certain geographical regions.

Users looking to stay safe should maintain multiple backups to minimize the risks from disk-encrypting malware and keep sensitive information encrypted or off networked machines altogether. Once files are leaked, it can be difficult or impossible to remove them from the internet.

"If the information is published in some server that's out of U.S. jurisdiction, for example, then having that information taken down is going to be very, very difficult," Segura says. That applies equally to business data and sensitive personal files like texts and photos.

"If you think you don't want your mother or grandmother to see that picture, think about putting it somewhere secure, because you don't want it leaked," he says.

Five ways the image-sharing site is harnessing AI to keep people engaged.

Thanks to recent gains in machine learning, computers are getting skilled at picking out patterns and features in text and images. That's how e-commerce giants like Amazon and eBay build sophisticated recommendation systems and how social networks like Facebook and Twitter are tweaking feeds to keep users hooked. Pinterest is no exception, with 30% of engagement tied to personalized real-time suggestions. Here's how Pinterest engineers are leveraging artificial intelligence to keep the website's 150 million–plus users pinning and sharing.

Identifying Visual Similarities

Machine learning can not only determine the subject of an image, it can also identify visual patterns and match them to other photos. Pinterest is using this technology to process 150 million image searches per month, helping users find content that looks like pictures they've already pinned. Pin a photo of a cheetah-print pillow, and Pinterest will serve up animal-print decor from other users. Future iterations of the Pinterest app may let users simply point their cameras at real-world objects to get instant recommendations.

Categorizing And Curating

If a user pins a mid-century dining-room table, the platform can now offer suggestions of other objects from the same era. The key? Metadata, such as the names of pinboards and websites where images have been posted, helps the platform understand what photos represent.

Predicting Engagement

While many platforms prioritize content from a user's friends and contacts, Pinterest pays more attention to an individual's tastes and habits—what they've pinned and when—enabling the site to surface more personalized recommendations. After all, friends who like the same recipes may not agree at all on fashion.

Prioritizing Local Taste

Pinterest is an increasingly global platform, with more than half of its users based outside the U.S. Its recommendation engine has learned to suggest popular content from users' local region in their native language. One finding: Slow-cooker recipes are more popular in the U.S. than the U.K., where the appliances aren't as common.

Going Beyond Images

Analyzing what's in a photo is a big factor in the site's recommendations, but it doesn't offer the whole story. Pinterest also looks at captions from previously pinned content and which items get pinned to the same virtual boards. That allows Pinterest to, say, link a particular dress to the pair of shoes frequently pinned alongside it, even if they look nothing alike.

Just before Donald Trump won the presidency, Canada issued new regulations making it easier for skilled workers to immigrate.

In March, when President-elect Donald Trump was still one of several candidates vying for the Republican nomination, his image appeared in a somewhat unlikely place: an ad placed by a Kitchener, Ontario, startup looking to hire software developers.

"Thinking of moving to Canada?" asked adtech firm Sortable. "Sortable is hiring."

The notice, on Facebook, got international news coverage. It was targeted more at Canadian citizens living in the United States than at American citizens looking to escape a potential Trump presidency, Sortable cofounder and CEO Christopher Reid now says.

But since the election, Canadian tech firms say far more U.S. coders are showing a serious interest in migrating north just as the Canadian government has put the welcome mat out.

In fact, days before Trump's election, Ottawa issued new regulations making it easier for foreign skilled workers to come to the country. The net effect could give a boost to the Canadian tech industry, which has long lamented a "brain drain" to Silicon Valley.

"The most significant thing is not the election," Reid says. "It's that the Canadian government is going to make it easy to recruit in the U.S."

The rules, slated to go into effect next year, are meant to reduce visa processing waits from five or six months to as little as two weeks. The new policies, hailed by Canadian tech companies, also include provisions making it easier for workers to secure short term visas for things like company training programs and filling short term staffing crunches.

"Our industry particularly needs the ability to move temporary foreign workers in and out of the country," says Information Technology Association of Canada president Robert Watson. After all, Canada is facing a technology worker shortage that ITAC estimates could leave 180,000 tech jobs unfilled by 2020.

At the same time, Trump's election does seem to be motivating some U.S. workers to take a more-than-joking look at relocating north. Reid says that in the days immediately following the election, U.S. traffic to Sortable's career site jumped from four or five hits per week to more than 200 per day, and the company even received a handful of job applications mentioning the political climate in the United States.

"I can't imagine what that looks like at a Canada-wide scale," he says. "There must have been a lot of people looking."

Other Canadian companies, including Figure 1, a social networking firm catering to medical professionals, marketing tech company Influitive, and wearables startup Thalmic Labs have all seen a spike in interest since Trump won the White House.

"We have seen a massive increase in the number of highly qualified, deeply experienced candidates applying from the U.S. specifically into our Toronto office—in fact there have been five in the past two days (20 since the election)," Influitive founder and CEO Mark Organ wrote Wednesday in an email to Fast Company. "We also have interest from super talented Canadian employees who are working in the SF Bay Area who are interested in coming back home."

Organ says he predicts a "golden age of technology" in Toronto, pushed in part by talented workers migrating to the city from around the world.

Similarly, Thalmic Labs CEO Stephen Lake says the company has seen "several people a day reaching out" since the election, including U.S. citizens and Canadians abroad, a big change from prior to November.

"I don't traditionally get a lot of Americans or Canadians in the U.S. reaching out," he says.

Not every company reported an uptick in interest from the U.S.—Jeremy Auger, chief strategy officer at Kitchener, Ontario-based education startup D2L says the company hasn't seen much of a change since the election.

"We haven't really seen a major uptick in Americans applying to Canadian jobs," he says, though he adds that some of that may be due to the fact that D2L already offers job opportunities around the world.

Jacqueline Bart, principal at Toronto immigration law firm Bartlaw, says she also saw a burst of calls from the United States immediately after Trump's Nov. 8 election. Some calls were from Americans considering a move, and others were from Canadian citizens looking into the possibilities of relocating with their families back to their home country, she says.

"They either want to move to Canada and sponsor their spouse, or they want to get Canadian citizenship for their children based on the fact that they're Canadian citizens and they're the birth parents of their children," Bart says.

It's unclear how many of those callers will ultimately follow through—Bart says many appeared to be considering options they could take if the United States takes a radical turn for the worse under Trump's leadership—but there are some signs that threats to move to Canada might be more serious than after other elections.

"People certainly aren't being discouraged to move to Canada right now," Bart says.

Students from both the U.S. and abroad are seriously considering spending their university days in Canada, rather than in the United States, The New York Times reported Thursday. Election night traffic to Canada's immigration web portal crashed the site, and travel sites reported a recent burst of interest in one-way tickets to Canada—and a decline in searches for corresponding trips to the United States.

Trump has so far done relatively little to reassure immigrants and those from families new to the United States that they'll be welcome under his administration, after repeatedly suggesting during his campaign that he intended to build a wall along the Mexican border, deport millions of undocumented immigrants, and impose "extreme vetting" or an out-and-out ban on Muslims looking to come to the United States.

Trump also at times critiqued the H-1B work visas used by many tech experts from other countries looking to work legally in the United States, ultimately leaving it unclear whether he will attempt to limit the program.

Rhetoric from his campaign and others in the Republican Party have also left many pessimistic about prospects under his administration for members of ethnic and religious minorities, as well as women and LGBT people.

'A lot of the inquiries that we're getting are from people who are trying to make up their minds and see how bad things could potentially get," Bart says, saying some of Trump's proposed policies could lead to a "little brain drain to Canada" from the United States.

But despite what might seem to be a golden opportunity to hire U.S. programmers, Reid says Sortable is unlikely to relaunch its Trump-focused recruitment ads.

"I think it would be extremely disrespectful to do that, when you have a divided country that in my opinion is probably in a lot of pain," he says. "I don't think you need to run an ad with a picture of the president as a recruiting tactic—I think that would be in poor taste."

A cybersecurity expert warns that Russian hackers may be looking to weaken the EU by bolstering Euroskeptic politicians.

Fancy Bear may have a new target.

The Russian hacking ring alleged to have leaked internal emails from the Democratic National Committee and Hillary Clinton's presidential campaign has turned its sights on upcoming European elections, a top cybersecurity expert says.

"The script that they've been able to successfully execute in the U.S. election I believe will be repeated in European elections," warns Dmitri Alperovitch, CTO of CrowdStrike, the Irvine, Calif., security company that was among the first to attribute the DNC hacks to Russia. "There's no question that they'd been able to cause a lot of chaos in the course of the election and get people to believe that the system has been rigged."

The Democratic Party hacks, attributed to a Russian hacking group dubbed "Fancy Bear", resulted in a leak of emails over the summer that seemed to imply DNC senior officials favored Clinton over primary rival Sen. Bernie Sanders. During the general election, leaks from the email account of Clinton campaign chairman John Podesta exposed further rifts in the party and evidence suggesting debate questions were leaked to the Clinton campaign.

National Security Agency chief Michael Rogers has denied that the hacks swayed the election, but did say the intrusions appear to have had nefarious intentions.

"This was a conscious effort by a nation-state to attempt to achieve a specific effect," Rogers said in a panel discussion hosted by the Wall Street Journal's CEO Council.

The Washington Post has reported that the Russians helped spread propaganda and fake news that favored Donald Trump on social media. The report drew some criticism for attributing part of the report to an anonymous group called PropOrNot.

Alperovitch declined to comment on suggestions the leaks may have helped deliver the election to President-elect Donald Trump but told Fast Company that Fancy Bear, believed to be linked to Russian military intelligence, appears to be shifting its attacks toward political and governmental systems in Europe.

"Right now we're only seeing intrusions," he says. "First they break into these networks, they steal data over a period of time—months sometimes—and then they decide whether they want to weaponize that data by leaking it in some fashion or else distorting it."

Alperovitch didn't identify specific governments or parties that have been targeted, and representatives from major European parties facing upcoming elections didn't reply to requests for comment, but Alperovitch says it's likely Russian hackers would target hacks and leaks in an effort to boost parties and candidates challenging the European Union.

European leaders have said Russia would be happy to see a weaker or more divided EU amid conflicts over the war in Syria and Russia's entry into Ukraine. The United Kingdom's surprising Brexit vote earlier this year, and the possibility of other countries pulling out of the bloc, could lead to a smaller EU less able to impose economic sanctions or other serious consequences on Russia.

In the United States, it's likely that Republican Party systems were also attacked by the Russian hackers, says Kenneth Geers, senior research scientist at security firm Comodo. But the hackers probably preferred to release information critical of the Democratic Party, given that Trump showed greater sympathy for Russia and President Vladimir Putin than Clinton did, he says.

"Trump almost stood alone, against everyone, in saying Putin is a great leader and Russian troops are not in Ukraine, for example," Geers says. "He's the only person on the planet who seems to have had that position."

At the same time, a second group of Russian hackers, nicknamed "Cozy Bear" and believed to be tied to a separate, civilian intelligence agency, have begun using tactics such as targeted phishing emails in attempting to hack influential U.S. think tanks, particularly those working on areas tied to Russia, Alperovitch says.

"They're trying to do what any nation state would be interested in doing, which is figuring out potential policy pronouncements from various organizations that could be close to the administration," he says.

Are you ready to build your own unmanned aerial vehicle? With this design tool, you may one day get your chance.

If you're not satisfied with any of the drones on the market, you may one day be able to design your own, thanks to researchers at the Massachusetts Institute of Technology.

A team from MIT's Computer Science and Artificial Intelligence Laboratory has created a software tool that allows even relative novices to drag and drop standard components—including propellers, support rods, and a range of core body frames—and craft their own drone designs. While the database is relatively small, users of the tool can compose and manipulate the components to create a widely diverse set of designs, according to a paper the team plans to present this week at the SIGGRAPH Asia Conference in Macao.

Once users finalize their designs, the program automatically adjusts it for user-selected features such as battery life and maximum carrying capacity, tweaking factors like the overall shape and the angle and spinning directions of each propeller.

"We will try to do some optimizations to replace the rotors, or change the angles, or resize the rods so we can meet the requirements of the users," Tao Du, one of the paper's authors and a graduate student in the Computational Fabrication Group, tells Fast Company.

The team developed an efficient algorithm that lets the optimizations generally finish within a few seconds, according to the paper, though the results can differ dramatically: In one case, the optimized version could carry 30% more weight than the initial concept.

The group was able to create some whimsical designs, including some eye-catching models with odd numbers of propellers and even a design with a frame based on the Stanford bunny—a 3D rabbit image widely used in computer graphics demos.

But Du says the program could have more practical applications than whimsy. The group developed the tool after observing that doing tasks like carrying unusually shaped objects, or making sure onboard camera lenses aren't obscured by a drone's rotating propellers, can be a challenge with off-the-shelf models, he says. The hope is that custom drone design could one day aid in the creation of unmanned aerial vehicles that can carry specialized payloads.

Once the custom design is optimized, users can test it in a simple flight simulator, flying the drone with an emulated controller to see how it feels. If it isn't quite what they had in mind, they can go back to the virtual drawing board for another try. The simulation is relatively simple and doesn't capture all of the aerodynamical effects that might affect a particularly fast-flying drone, but it works reasonably well for slow-moving copters, according to the paper.

In testing, the team 3D-printed some of the controllers used to attach parts like motors and propellers—which have to be installed at custom angles output by the optimization engine—and put together the drones with the parts selected from the program's database.

"Most of the parts are pretty standard off the shelf, like motors, propellers, and batteries," Du says, adding that someone with even moderate engineering ability should be able to put together the parts.

Unfortunately for aspiring DIY drone designers, the team doesn't have any immediate plans to commercialize or otherwise release the product. But here's the good news: They may release an open-source version sometime in the future.

Intel, Nvidia, and a host of startups are rolling out new chips to make machine learning faster and more powerful.

In recent years, advanced machine learning techniques have enabled computers to recognize objects in images, understand commands from spoken sentences, and translate written language.

But while consumer products like Apple's Siri and Google Translate might operate in real time, actually building the complex mathematical models these tools rely on can take traditional computers large amounts of time, energy, and processing power. As a result, chipmakers like Intel, graphics powerhouse Nvidia, mobile computing kingpin Qualcomm, and a number of startups are racing to develop specialized hardware to make modern deep learning significantly cheaper and faster.

The importance of such chips for developing and training new AI algorithms quickly cannot be understated, according to some AI researchers. "Instead of months, it could be days," Nvidia CEO Jen-Hsun Huang said in a November earnings call, discussing the time required to train a computer to do a new task. "It's essentially like having a time machine."

While Nvidia is primarily associated with video cards that help gamers play the latest first-person shooters at the highest resolution possible, the company has also been focusing on adapting its graphics processing unit chips, or GPUs, to serious scientific computation and data center number crunching.

"In the last 10 years, we've actually brought our GPU technology outside of graphics, made it more general purpose," says Ian Buck, vice president and general manager of Nvidia's accelerated computing business unit.

Speedily drawing video game graphics and other real-time images relies on GPUs that perform particular types of mathematical calculations, such as matrix multiplications, and handle large quantities of basic computations in parallel. Researchers have found those same characteristics are also useful for other applications of similar math, including running climate simulations and modeling attributes of complex biomolecular structures.

And lately, GPUs have proven adept at training deep neural networks, the mathematical structures loosely modeled on the human brain that are the workhorses of modern machine learning. As it happens, they also rely heavily on repeated parallel matrix calculations.

"Deep learning is peculiar in that way: It requires lots and lots of dense matrix multiplication," says Naveen Rao, vice president and general manager for artificial intelligence solutions at Intel and founding CEO of Nervana Systems, a machine learning startup acquired by Intel earlier this year. "This is different from a workload that supports a word processor or spreadsheet."

The similarities between graphics and AI math operations have given Nvidia a head start among competitors. The company reported that data center revenue more than doubled year-over-year in the quarter ending Oct. 31 to $240 million, partly due to deep learning-related demand. Other GPU makers are also likely excited about the new demand for product, after reports that industrywide GPU sales were declining amid decreasing desktop computer sales. Nvidia dominates the existing GPU market, with more than 70% market share, and its stock price has nearly tripled in the past year as its chips find new applications.

Graphics Cards Make The AI World Go Round

In 2012, at the annual ImageNet Large Scale Visual Recognition Challenge (a well known image classification competition), a team used GPU-powered deep learning for the first time, winning the contest and significantly outperforming previous years' winners. "They got what was stuck at sort of a 70% accuracy range up into the 85% [range]," Buck says.

GPUs have become standard equipment in data centers for companies working on machine learning. Nvidia boasts that its GPUs are used in cloud-based machine learning services offered by Amazon and Microsoft. But Nvidia and other companies are still working on the next generations of chips that they say will be able to both train deep learning systems and use them to process information more efficiently.

Ultimately, the underlying designs of existing GPUs are adapted for graphics, not artificial intelligence, says Nigel Toon, CEO of Graphcore, a machine learning hardware startup with offices in Bristol, in the U.K. GPU limitations lead programmers to structure data in particular ways to most efficiently take advantage of the chips, which Toon says can be hard to do for more complicated data, like sequences of recorded video or speech. Graphcore is developing chips it calls "intelligent processing units" that Toon says are designed from the ground up with deep learning in mind. "And hopefully, what we can do is remove some of those restrictions," he says.

Chipmakers say machine learning will benefit from specialized processors with speedy connections between parallel onboard computing cores, fast access to ample memory for storing complex models and data, and mathematical operations optimized for speed over precision. Google revealed in May that its Go computer AlphaGo, which beat the Go world champion, Lee Sedol, earlier this year, was powered by its own custom chips called tensor processing units. And Intel announced in November that it expects to roll out non-GPU chips within the next three years, partially based on technology acquired from Nervana, that could train machine learning models 100 times faster than current GPUs and enable new, more complex algorithms.

"A lot of the neural network solutions we see have artifacts of the hardware designs in them," Rao says. These artifacts can include curbs on complexity because of memory and processing speed limits.

Intel and its rivals are also preparing for a future in which machine learning models are trained and deployed on portable hardware, not in data centers. That will be essential for devices like self-driving cars, which need to react to what's going on around them and potentially learn from new input faster than they can relay data to the cloud, says Aditya Kaul, a research director at market intelligence firm Tractica.

"Over time, you're going to see that transition from the cloud to the endpoint," he says. That will mean a need for small, energy-efficient computers optimized for machine learning, especially for portable devices.

"When you're wearing a headset, you don't want to be wearing something with a brick-sized battery on your head or around your belt," says Jack Dashwood, director of marketing communications at the San Mateo, California-based machine learning startup Movidius. That company, which Intel announced plans to acquire in September, provides computer vision-focused processors for devices including drones from Chinese maker DJI.

Nvidia, too, continues to release GPUs with increasing levels of support for machine learning-friendly features like fast, low-precision math, and AI platforms geared specifically for next-generation applications like self-driving cars. Electric carmaker Tesla Motors announced in October that all of its vehicles will be equipped with computing systems for autonomous driving, using Nvidia hardware to support neural networks to process camera and radar input. Nvidia also recently announced plans to supply hardware to a National Cancer Institute and Department of Energy initiative to study cancer and potential treatments within the federal Cancer Moonshot project.

"[Nvidia] were kind of early to spot this trend around machine learning," Kaul says. "They're in a very good position to innovate going forward."

Related Video: Will AI Destroy Or Delight Us?

That may be in part due to the number of tourists and convention-goers bringing their laptops to the city.

While ransomware can infect computers anywhere, researchers from security firm Malwarebytes say certain cities have been particularly hard hit, with the Las Vegas area at the top of the list.

The Santa Clara company detects the malicious software on devices in the Las Vegas/Henderson area at a rate 500 times the average of the top 40 U.S. cities by ransomware detection, the company said this week. The city sees the most infections overall, as well as the most infections per computer and per resident, according to the company.

That may be in part due to the number of tourists and convention-goers bringing their laptops to Las Vegas—then, after days filled with work and play, forgetting all the lessons they've heard about keeping their computers safe, says Adam Kujawa, head of malware intelligence at Malwarebytes.

"They might have a few drinks, they'll relax, they'll see a show—that kind of environment is created almost specifically to lower people's guards," he says. "It makes it a really prime situation for cybercriminals to dupe users."

After ransomware gets on a computer, it typically encrypts files and effectively holds them prisoner until the owner pays a bounty. It's often installed by clicking on attachments in scam emails or through automated downloads from malware-infected websites.

While security firm Symantec said in a July report that the majority of ransomware infections are on consumer-owned, rather than corporate, machines, the attacks have also disabled computers at institutions from hospitals to police departments. Last month, the San Francisco Municipal Transportation Agency shut down fare payment systems and offered free subway rides after the agency said ransomware struck more than 900 office computers.

"The SFMTA has never considered paying the ransom," the agency said in a statement. "We have an information technology team in place that can restore our systems, and that is what they are doing."

In general, corporate networks get infected through the same types of attacks that infect personal machines, Kujawa says.

"The intentional attack of large organizations, while it does happen, is not that common, and much more often these people might get infected by a mass campaign that one of their employees might fall for, and then it spreads through the network," he says.

The company's study, which tracked 400,000 cases where ransomware was detected from July through October, including incidents in more than 200 countries, found the U.S. overall had by far the greatest share of ransomware attacks, with 26% of all studied incidents, similar to numbers reported earlier this year by Symantec.

"We wanted to see exactly what countries, what parts of the world, and even what cities were being hit the most," Kujawa says. "We found that the USA was actually hit the most, more than any country out there, with Germany being hit second."

Ransomware attackers use increasingly sophisticated and targeted phishing emails to dupe victims, often masquerading as banks or payment providers like PayPal, he says. Many attackers are customers of so-called ransomware-as-a-service providers, which develop the software and split profits with attackers who craft the actual emails luring victims.

"If they want to hire a scammer or spammer or somebody who might be running a malvertising campaign or just a drive by exploit campaign, they can do that," Kujawa says.

The most common ransomware variant found in the Malwarebytes study was Cerber, which first appeared in March and is believed to be created by a ransomware-as-a-service provider based in Russia. The malware will generally refuse to infect computers in Russia and other former Soviet nations, according to Malwarebytes.

Rounding out the list of top U.S. cities for ransomware infection were Memphis, Tennessee; Stockton, California; Detroit, Michigan; Toledo, Ohio; Cleveland, Ohio; Columbus, Ohio; Buffalo , New York; San Antonio, Texas; and Fort Wayne, Indiana, according to Malwarebytes.

Related Video: Inside The Secret World Of Code-To-Code Combat

The public divide between the intelligence agencies and the president-elect, who questions such Russian ties, is dangerous, say experts.

Just before Donald Trump takes the presidential oath of office next month, the Obama administration plans to release a full report on Russian connections to cyber attacks during the election, officials said Friday.

Trump has cast doubt on the Kremlin's involvement in politically charged hacking and fake news efforts prior to this year's election while leading members of Congress have called on the administration to share what it knows about the Russian connection and pushed for a wider investigation, leading to a potentially fiery showdown over the issue. And such a public divide between the president-elect and intelligence agencies is unusual and dangerous, say experts.

"To evaluate Congress's response appropriately, we would like all Members to have a comprehensive understanding of what the U.S. Intelligence Community knows regarding Russia's involvement in these actions and attempts to interfere in our election,"the lawmakers wrote this week in an open letter to President Barack Obama.

Officials sometimes purposely share with Congress "only what is necessary for them to perform proper oversight," partially as a safeguard against unintentional leaks, says Chris Finan, a former White House cybersecurity advisor and CEO of Bay Area security startup Manifold Technology. But in this case, Finan says it was more likely that intelligence agencies are still working to establish precisely how various efforts to influence the election link back to the top levels of the Russian government.

"They'd want to try to find deliberate command and control links—instances where people were directed by a Russian official," he says. "I think that there's probably a lot of interest and focus in, are these proxy groups operating on their own based on what they thought Moscow would want them to do or were they acting on a directive from somebody in the Russian government."

Prominent Democrats in the House of Representatives introduced a bill on Wednesday calling for a bipartisan commission to investigate Russian attempts to influence the election. Maryland Rep. Elijah Cummings cited October statements from intelligence officials blaming Russia for hacks targeting Democratic Party servers and subsequent leaks of embarrassing emails.

"The head of the National Security Agency said there is no doubt that this was a targeted and conscious effort to achieve a specific result," Cummings said in a statement. "And the Intelligence Community also said that they believe that only Russia's senior-most officials could have authorized these activities."

Prominent Republican senators including Arizona Senator and Armed Services Committee Chairman John McCain, Intelligence Committee Chairman Richard Burr of North Carolina, and Tennessee Senator and Foreign Relations Committee Chairman Bob Corker have also all expressed interest in investigating Russian hacks, despite the president-elect's skepticism. House Homeland Security Chairman Michael McCaul, R-Texas, has also recently publicly condemned Russia for the attacks despite his support for Trump.

While Obama and Vice President Joseph Biden have both hinted at the possibility of retaliation for the alleged Russian hacks, it's likely that at this point in Obama's presidency any further responses will be left to the Trump administration, Finan says.

Trump said in an interview with Time magazine, which named him 2016's "Person of the Year," that he doesn't believe Russia interfered with the election and thinks intelligence agency conclusions may have been politically motivated.

Trump's statements drew a strong rebuke from Rep. Adam Schiff, D-California, who called his comments "enormously damaging to the country" and to intelligence officials.

"Notwithstanding the abundance of evidence that Russia hacked our political institutions during the Presidential campaign and dumped documents in an effort to meddle in our political affairs, President-elect Trump's comments this morning continue to contradict our intelligence professionals and carry water for the Kremlin,"Schiff said in a statement.

Schiff declined through a spokesman to comment further on the matter, and a Trump spokeswoman didn't respond to an email seeking comment.

Such a divide over the legitimacy of intelligence reports is highly unusual, particularly when conclusions within the intelligence community seem so unanimous, Finan says.

"Not only is it unusual—it's incredibly dangerous," he says. "To question the veracity but also to question the professionalism of our intelligence community, that is potentially very damaging to morale and it's going to put the nation at risk."

A 15,000-strong group of volunteer "Techfugees" is helping migrants and asylum seekers connect to the web—and each other.

Thousands of techies the world over have banded together to help refugees flooding Europe to stay connected.

The needs of the waves of migrants from Syria, Afghanistan, and other points—more than a million in 2015—go beyond just shelter, safety, and sustenance.

"You can imagine, crossing to a border or coming to a place you don't know. Information needs are massive," says Alyoscia D'Onofrio, senior director at the governance technical unit of the International Rescue Committee, which assists refugees and displaced people around the world.

"One of the big differences between this crisis response and many that have gone before is that you're got probably a much tech-savvier population on the move and probably much better access to handsets and networks."

Helping to connect those newcomers to information—and each other—is a group of 15,000 digital volunteers who call themselves the Techfugees.

"We are here not to solve the biggest problems of hygiene, water, clean energy because these are sectors that need a lot of expertise," says Joséphine Goube, the CEO of the nonprofit that quickly came together last year.

Instead, often with the aid of smartphones many migrants and asylum seekers bring with them, the continent's tech community aids refugees and asylum seekers in getting back online to find their footing in unfamiliar places.

Soon after surviving arduous trips across the Aegean Sea to Lesbos and other Greek islands, migrants were supplied cellphone chargers alongside other essentials, D'Onofrio says. Smartphones let them connect with loved ones left behind through apps like Facebook and WhatsApp and share safety information and travel tips.

The devices also let aid organizations like the IRC provide verified information about travel options and places to seek assistance in languages refugees speak—even when conditions are changing more rapidly than paper flyers can be distributed.

"Providing a flow of real-time relevant information is so much easier through a web-based platform that people can access on their phones than it is in a much more old-fashioned way when information is changing all the time," says D'Onofrio.

The IRC has received substantial funding from tech companies to support its efforts, and individual tech workers have flocked to dozens of conferences and hackathons organized by Techfugees around the world since an initial conference in London last October.

"We were actually overwhelmed by the response to our conference," says Goube, "It just went viral."

Affiliates of the group have since helped provide infrastructure for refugees to connect to Wi-Fi—even in places with limited electricity—and energize their phones through solar-powered charging hubs. They've also developed websites and apps to teach new arrivals everything from basic coding skills that could help them earn a living to how to navigate government bureaucracies in their new countries.

"Things that seem very trivial to us can actually be very complicated," says Vincent Olislagers, a member of a team developing an interactive chatbot called HealthIntelligence, which is designed to provide refugees in Norway with information about using the country's health care system. The tool was developed after the team met with a recent arrival to the country who had difficulty arranging hospital transportation for his pregnant wife due to language barriers and financial constraints.

"He had to call, for his wife, his caretaker at the refugee center," Olislagers says. "The caretaker had to send an ambulance at the right location."

The team is working with Norwegian health officials and refugee aid groups to ultimately make the chatbot available as part of a standard package of materials provided to refugees entering the country. The project was a finalist in an October hackathon organized by Techfugees in Oslo. The hackathon's ultimate winner was a group called KomInn, which pairs families fluent in Norwegian with newcomers who come to their homes to practice the language over dinner. That group developed a digital tool to streamline finding matches, which had previously been a laborious process, says Goube.

"They had to manually match people," she says. "It was just a catastrophe."

The Techfugees group is undeterred by the mounting anti-immigrant sentiment that has made some politicians and tech companies wary of outwardly supporting efforts to help new arrivals, Goube says.

The goal isn't to push aside traditional nonprofits and volunteers, she adds: "We're here to provide tech support to the NGOs—have them enter the 21st century, basically."

A mixture of technical innovation and social incentives could make online comments readable—and even engaging.

For years, it's been a standard piece of advice to anyone reading the news online: "Don't read the comments."

It's no secret that user-submitted comments on news websites are often angry, racist, misogynistic, or simply ill-informed. That's contributed to media organizations from NPR and Reuters to Popular Science and legal news site Above the Law deciding in recent years to eliminate comment sections altogether. Other sites, such as the New York Times, have taken to heavily moderating what readers post and limiting when they can do so.

But many media companies remain committed to providing forums where their readers can pose questions, contribute their wisdom, and even civilly debate with one another. In part, many in the industry see it as important to offer a place where people can discuss current events outside "the friends and family echo chamber" of social media, says Aja Bogdanoff, cofounder and CEO of Portland comment tech startup Civil.

Denying them the opportunity to comment might drive them to other venues where "the conversation" is already taking place—social media sites like Facebook and Twitter, local networks like Nextdoor, or forums like Reddit. While some media groups have reported success with closing their comment sections and connecting with readers through social media, that trend also risks reducing reader engagement with the news outlets themselves, and disperses potentially valuable discussion about an article across the web, where it's harder to find.

More pragmatically, many in the industry have found frequent commenters are often also among sites' most dedicated readers, spending more time reading articles and looking at ads, and among those more likely to purchase paid subscriptions, says Greg Barber, director of digital news projects at the Washington Post.

"By othering this group—by saying these people who come and speak in the comment spaces, they're something else and they're undesirable—that's actually really dangerous for news organizations," says Barber, who is also head of strategy and partnerships at The Coral Project, a collaborative effort of the Post, the New York Times, and Mozilla to improve community engagement on news sites. "They are, in fact, your most loyal readers—these are the people who are paying the bills."

Civil and The Coral Project are among a group of startups, established media companies, and academic projects working to develop ways for news sites to engage active readers. Many are looking to move beyond the troll-friendly environs of a traditional comment box, which can often incentivize attention-seeking commenters to take extreme positions and harass other users.

"The comment's just a tool—it's like walking around your house and trying to take on all your household chores with just a broom," Barber says. "What we've been trying to do as an industry is, we've been tacking this one feature onto every article, and then we're shocked when it doesn't work every time."

Organizations are experimenting with a variety of tools, from streamlined human moderation to automated filtering that can improve the traditional comment experience, as well as looking at other tools like more targeted questionnaires and quizzes that can let readers interact and express themselves without providing an avenue for harassment.

"I think this is a bit of a million-dollar question right now: How can we create a space where members of the public have the opportunity to talk to each other, and it doesn't turn into a terrible cesspool?" asks Talia Stroud, associate professor of communication studies at the University of Texas at Austin and director of the Engaging News Project at the university's Annette Strauss Institute for Civic Life.

That project offers an embeddable quiz widget that it has found encourages readers to spend more time on articles without the partisan ballot box stuffing and scientifically dubious results of traditional online opinion polls. The quizzes—asking readers to do things like recall numbers presented in articles—can also help readers remember concrete figures better than simply reading a written piece, the project has found. (You can try one at the bottom of this article.)

The Coral Project also offers an embeddable polling tool called Ask that enables publishers to ask readers specific questions. It gathers information they can use for additional reporting and to produce galleries out of interesting responses that can be displayed onsite.

"Using a form where the responses aren't posted immediately creates a different tone," says Barber.

Readers tend to respond a bit more deliberately, he says, than when their posts are immediately dropped onto an existing thread. The project is also working on a sophisticated, highly configurable tool for traditional comments called Talk, slated for a beta release early next year, Barber says.

About nine months ago, Civil released a next-generation comment tool called Civil Comments that is now used by about 50 sites, including the Globe and Mail, among Canada's most widely read newspapers. Civil Comments requires readers to review others' comments before they can submit their own, specifying whether a particular comment is subjectively good and whether it's "civil."

On the back end of the system, the company's algorithms adjust for biased moderation. Civil Comments has reduced flagging of published comments as abusive by 90% to 95% on sites where it's deployed, substantially reducing the requirements for moderators employed by the sites, even as the number of comments submitted has often doubled or tripled, Bogdanoff says.

"In general, what we're seeing is, you get a wider range of voices," she says. "The old approach to commenting on news sites really prevented anybody [from participating] who didn't just want to fight with people all day."

Other organizations are taking a more automated approach to boosting civility. Toronto startup Viafoura provides what it calls an "audience development platform" that monitors user email addresses, avatars, and IP addresses to weed out aggressive spammers and trolls, says Allison Munro, the company's head of marketing and business development. Customers include the Los Angeles Times, TMZ, and the Canadian Broadcasting Corp (CBC).

After commenters are vetted, individual comments are filtered by smart algorithms that use a mix of configurable rules and machine learning to avoid publishing unwanted material.

"We'll start from the beginning where we have a basic set of rules, and we can tune those to be either very intense, as in, don't allow anything in that looks like a swear word or looks like spam, or we can adjust the levels," says Munro. "We had one customer in Texas that had 'redneck' as a positive word. In New York, we had another customer that saw that as a personal attack."

For the CBC, Viafoura appeared to lead to users commenting more and spending more time engaging with the site—and a decrease in the number of comments flagged as inappropriate, Munro said in a November blog post.

Disqus, the nine-year-old San Francisco company that says it serves more than 1.8 million comments every day across 750,000 websites, also introduced improved features to fight spam, facilitate moderation, and block trolls this year.

"We've been hearing from Disqus users that they want better blocking controls, and I'm happy that the feature is finally here," Disqus CEO Daniel Ha said in a June statement. "We hope that this makes it easier for people to dive into good discussions, plus we also know that it'll really help out publishers by reducing how much moderating they have to do."

Ultimately, quality comments will likely come as a combination of technology and human effort, says Nicholas Diakopoulos, an assistant professor of journalism at the University of Maryland, College Park.

"Commenting platforms are a classic example of a sociotechnical system: They rely not only on technology, but also people like commenters, editors, and moderators operating under a set of norms and policies that influence behavior," he wrote in an email.

Better-quality comments can also benefit media organizations when news breaks: In those moments, thoughtful reader discussion can itself go viral, encouraging yet more discussion (and more traffic), as happened for some British publications after this year's Brexit vote, says the Washington Post's Barber. And they can reduce situations where user-submitted sections of a publication's site don't meet the typical standards of the organization, he says.

"They're pixels on a page, just like all the other pixels on a page that we manage very carefully," Barber says. "By creating these spaces and then ignoring them, we're not mowing the yard and then being angry when the grass grows in a way we didn't expect."

Take a quick quiz we built using the Engaging News Project's quiz tool:

Evidence has yet to emerge to back the CIA's reported claims of Russian hacking of Republican National Committee computer networks

Ever since last Friday's bombshell reports that the CIA believes Russian hackers penetrated both Democratic and Republican computer networks before the election, GOP officials have vehemently denied that systems linked to the Republican National Committee were hacked. By contrast, Democratic officials acknowledged last summer that Russian hackers had gained access to the Democratic National Committee, the Democratic Congressional Campaign Committee, and the email account of John Podesta, Hillary Clinton's campaign chairman.

The reports, which describe a CIA assessment that the Kremlin interfered with the U.S. election in order to help Donald Trump win, have sparked a fierce divide between President-elect Trump and some Republican leaders in Congress. Trump has dismissed the reports, while Senators John McCain and Lindsey Graham have joined Democrats in calling for investigations into the hacking. Among the most hotly debated details is that Russian agents hacked the RNC's computer systems but "did not release whatever information they gleaned from the Republican networks," per the New York Times.

"Number one, the RNC was not hacked," said Reince Priebus, chairman of the RNC and President-elect Donald Trump's pick for chief of staff, on NBC's Meet the Press Sunday. "I don't know of any employee, on any of their own Gmail accounts, that was hacked."

Priebus said in the televised interview that the RNC worked with the FBI after well-publicized digital attacks on the Democratic National Committee this year, and that the FBI found no evidence the RNC's systems were compromised. Through a spokeswoman, the RNC declined to provide Fast Company with additional information about the steps it took to determine its systems were hacker-free. According to a Friday report in The Wall Street Journal citing sources close to the investigation, experts found that the RNC's filters blocked phishing emails targeting a former staffer and believed that the hackers made a less determined effort to breach the RNC's systems.

Still, cybersecurity experts say it's generally difficult for an organization to definitively determine it hasn't been hacked and even to fully prevent attacks by sophisticated adversaries.

"Especially when a state-level adversary is going after an organization, there's very little you could do to prevent that," says Danny Rogers, CEO of security firm Terbium Labs.

Still, neither federal officials nor outside researchers have yet to publicly produce any evidence that RNC systems were penetrated, though Graham said in a Wednesday CNN interview that his own campaign attack was hacked by Russians.

A few hundred relatively innocuous emails appearing to be linked to state-level Republican organizations and campaign workers for Graham and fellow senator John McCain appeared online in data dumps linked to Russian state-sponsored hackers earlier this year, but there's been no evidence that the emails came from a system linked to national Republican organizations. Other Republican-linked information, like personal data about Republican primary delegates and Trump Organization staff, also appeared online earlier this year, says Rogers, though he says there's also no indication that it came from a breach of the RNC's network rather than from other sources.

Publications including The New York Timesreported that CIA officials believed Russian hackers had deliberately withheld information about Republican campaigns while leaking Democratic Party data in order to boost President-elect Donald Trump's campaign.

Multiple news outlets reported Wednesday that intelligence officials believe Russian President Vladimir Putin personally directed how some Democratic Party information was leaked. While the White House stopped short of directly accusing Putin of directing the attack, deputy national security advisor Ben Rhodes told MSNBC it's unlikely something "of this consequence" would take place without Putin's knowledge. Russian Foreign Minister Sergei Lavrov has dismissed the accusations against Putin, according to Reuters.

The RNC hasn't identified any security firms involved in securing its email systems or certifying them as free from breaches. But senior staff at a multiple companies involved in securing digital systems used for this summer's Republican National Convention told Fast Company they saw no sign of targeted, sophisticated attacks on convention networks, though the companies weren't directly involved with other RNC-related systems outside the convention.

"We didn't see anything that we felt was a nation-state adversary working in a very sophisticated manner," says Katherine Gronberg, vice president of government affairs at security firm ForeScout. "We didn't see evidence of something that we thought, just from our experience, could be attributed to one particular country or actor or group, like these groups in Russia."

Nor, says CEO Vince Crisler of Dark Cubed, did the companies securing convention systems hear at the time of any such attacks on other Republican Party computers.

"I know from all our interactions with them they have been engaged and proactive on this issue," Crisler says of the RNC.

Trump and some of his political allies have repeatedly questioned the reported assertions by federal intelligence agencies and outside analysts that Russian hackers were responsible for election-related hacks. He has also questioned the ability of analysts to determine who was responsible for a cyberattack after it's taken place.

"Unless you catch 'hackers' in the act, it is very hard to determine who was doing the hacking,"Trump wrote in a Monday tweet. And John Bolton, considered a likely Trump nominee for deputy secretary of state, even suggested election-related hacks could be a "false flag" attack, deliberately crafted by another country's intelligence agents to implicate Russia.

Yet experts say it's often quite possible to determine who's responsible for a cyberattack based on digital clues left behind. "You could say hey, we have seen this code somewhere else or similar code and it was known to be used by this criminal group, so we can make the link," says Chenxi Wang, chief strategy officer at security firm Twistlock. Or, she says, compromised computers may connect to servers or other internet resources known to be connected with particular hacker groups.

Those sorts of clues were how security company CrowdStrike, which investigated the DNC breach, was able to identify Russian groups dubbed Fancy Bear and Cozy Bear as having accessed the network.

Online accounts at URL shorteners that were evidently used in targeted phishing attacks on some Democratic Party officials were also used for months prior to those attacks to attack political figures in Russia and Ukraine, consistent with Russian involvement, according to cybersecurity firm SecureWorks. That makes a false flag operation less likely, since another spy organization would be less likely to spend months of work attacking targets in the former Soviet countries only to frame Russia, says SecureWorks senior security researcher Phil Burdette.

"We certainly considered that possibility," he says. "If you're an intelligence service, at the end of the day you have timelines and you have budgets in which you have to execute."

Experts from security company ThreatConnect have similarly said DC Leaks, a site used to leak data tied to the Hillary Clinton campaign and the emails taken from Republican operatives, appears to be tied to Fancy Bear. That implies that even if the national party did indeed evade the pre-election wave of cyberattacks, Russian hackers still managed to grab some internal messages from Republican campaign operations.

"The DCLeaks website almost certainly was set up to be used as a public mouthpiece specifically to leak documents gleaned from Fancy Bear operations, as we assessed back in August," ThreatConnect senior intelligence researcher Kyle Ehmke wrote in an email response to questions. "We have seen no indication that DCLeaks has hosted information garnered through non-Fancy Bear operations."

Robots are working alongside people in more warehouses, though creators say they're not taking human jobs . . . yet.

Some of the gifts you order this holiday season just might get shipped with the help of a robot.

Partly in response to an improving labor market that's made seasonal warehouse worker shortages a staple of year-end news coverage for the past several years, more logistics operations are turning to robots to make operations more efficient.

"What we find a lot of our customers are experiencing challenges in, especially as you think about what happens to an operation as you enter into the holiday season, is quite simply there's not enough people to do the work that's required," says Matt Rendall, cofounder and CEO of Clearpath Robotics. "We see customers that have literally half of their facility shut down because they can't bus people from far enough way."

Clearpath Robotics is one of several companies offering what are essentially self-driving cars designed to operate in warehouses rather than on the road (their version is called Otto Motors). Warehouse workers, especially in e-commerce businesses, spend much of their time picking objects out of one type of box, walking them to packing stations, and loading them into other types of boxes for shipping. And while humans are still better at machines at finding the right item, making sure it's in shippable condition and packing it up, there's no reason not to outsource the middle transportation step to machines, Rendall says.

"The ability to pull it out of the bin and put it into a box for packing is the most valuable thing you could have a fulfillment laborer do," he says. "Once they're done picking their order, they need to walk 500 feet to the packing station, and while they're walking 500 feet to the packing station, they're not picking the next order."

Unlike some other forms of automation, Clearpath and its rivals don't require that warehouses be redesigned to accommodate robots or that the machines be kept away from humans in order to ensure their safety. Instead, the machines are working alongside people, essentially doing the easiest part of the job, says Clint Reiser, director of supply chain research at analyst firm ARC Advisory Group.

"It's not a 100% automated strategy—it's more a strategy of using robots along with human labor," he says.

For Clearpath, it generally takes only a day or two to get the machines up and running in a new environment, Rendall says.

"We take Otto on a tour of the plant—we will manually drive Otto around the facility teaching it its environment, its surroundings, and during that process it's creating a memory of the map that the factory has," he says. "We go through a basic configuration in software—it's a very point-and-click process, and then within a day we're doing pickups and deliveries."

Other startups in the field—and the field is rich with new companies eager to fill the void left by Kiva Systems, a robotics company famously acquired by Amazon in 2012 that now only provides new bots to its parent company—offer similar ease of installation. (Amazon didn't respond to multiple requests for comment).

Robots from San Jose-based Fetch Robotics can follow workers around a warehouse as they pick items from shelves, dispatching a replacement and traveling to a packing area when an order's done, or stay put until a human orders them to a particular destination.

"As a picker picks a set of goods they walk by a robot and put the goods on the robot, so the picker can still continue picking," says Fetch CEO Melonee Wise.

And machines from Locus Robotics, a Wilmington, Mass., startup spun off from retail warehouse operator Quiet Logistics, essentially divide warehouses into geographical zones, where robots roll to particular shelves and summon nearby humans to pick products they need to fill orders.

"We decided that with our robot solution you'd make it so you didn't have to disrupt anything at all," Locus founder and chairman Bruce Welty says. "All you have to do is put this little sticker—we call it a locus point—basically a barcode you want anywhere you want the robot to be able to navigate."

A tablet mounted on the machine lets workers know what items they need to grab.

"The lights are blinking green saying, 'I need to have somebody do a pick,'" Welty says. "All his job or her job is, is to walk around and look for robots that are green."

As, well, robotic as that job seems, Welty says that when Quiet Logistics gradually rolled the robots out across one of its facilities, employees clearly preferred to work in the areas with the machines. But as robots get increasingly sophisticated at warehouse chores beyond going from place to place, it's hard not to imagine they'll begin to replace some of their human colleagues altogether.

"I think it can cut into the workforce," says Reiser. "I think it's part of the idea."

In a now-annual competition for item-picking robots, sponsored (naturally) by Amazon, progress has been rapid, with this year's winner able to pick and unload a variety of consumer goods from fairly densely packed shelves. And bots from Los Angeles-area startup InVia Robotics can already grab basic items like boxed electronics and ink cartridges from shelves, says InVia CEO Lior Elazary.

"One of the things that we believe does set our robot apart is the picking aspect," he says. "We can really handle from about a pack of gum to about a 40-pound case of soda."

The machines use a sliding arm equipped with suction cups to pull stacked items onto themselves with no need for human intervention.

The company rents the robots for monthly fees that can come out to about 10 cents per pick, or even less with 24-hour operation, compared to between 40 and 60 cents for a non-automated warehouse. For some items, like boxes of toner, the robots can even transport items directly to a label printer so they can be addressed and ready to be mailed out, though anything requiring "fine manipulation" still typically needs a human touch.

So far, Elazary says, customers have tended to use the robots to boost fulfillment throughput, not eliminate jobs. Humans are also still involved restocking the shelves for robots to pick from to fulfill orders, checking goods for quality purposes before they're shipped out, dealing with special situations and training the bots to handle changes like new item packaging. Replacing miles of walking with more interesting work—such as caring for new automated colleagues—can help boost worker satisfaction and retention, he says.

"Some of those people actually become sort of robot whisperers," he says.

Related Video: If Flexport Succeeds, Everything You Buy Will Cost Less

Deep learning and AI are turning millions of satellite images of Earth into usable data, helping insurers and aid groups alike to do their jobs.

A hurricane hits a shore town. What is the estimated property damage? A city is doing an inventory of trees. How many are there? An aid group is trying to get food to an impoverished rural population. What's the best location to make a drop?

Answering those and myriad other questions about our planet usually takes painstaking boots-on-the ground work. But satellite imagery and advanced machine learning, aka deep learning, is changing all that.

Anyone who's ever toyed around with Google Earth knows the eyes in the sky can capture amazingly detailed images of what's on the ground. The new AI is now teaching those orbiting cameras what they're seeing.

"The challenge is, how do you convert all those pixels into meaningful information?" says Shay Har-Noy, vice president and general manager of platform at satellite imagery provider DigitalGlobe. And that's exactly what the Colorado-based company is doing.

DigitalGlobe collects about 73 terabytes of satellite images every day, adding to a 16-year running archive of photos of locations around the world.

DigitalGlobe has developed a "geospatial big-data" platform it calls GBDX that trains computers to automatically detect relevant objects in those satellite images. Using deep learning, a technique that harnesses large sets of data to develop multilayered mathematical structures to classify images, audio, or text, DigitalGlobe's cloud-based systems can accurately pick out and analyze points of interest in the photos without a great degree of human effort.

That has let the company reach new customers who are interested in accessing structured geographical data they can work with in tools like Excel, but aren't interested in manually sifting through huge sets of raw images. Potential users range from insurers and investors looking to understand real estate development patterns, to aid organizations aiming to serve populations in sparsely mapped areas of the globe.

Among DigitalGlobe's new customers is PSMA Australia, a consortium of Australian federal, state, and territorial agencies working with the company to build a national database of information about man-made structures across Australia. The database, called Geoscape, saw its first data release last month and is ultimately slated to include satellite-imagery-derived information about more than 20 million structures across the continent.

The GBDX platform relies on the Amazon Web Services cloud to store DigitalGlobe's enormous collection of data—the company was among the first to use Amazon's Snowmobile service, which uses a shipping container filled with data storage equipment to transfer huge volumes of information to Amazon's servers. The platform enables developers to process that image data with DigitalGlobe's machine learning algorithms or their own, either for their own use or for resale to their own customers interested in the results.

The platform has also been used for processing data in the wake of natural disasters. After Hurricane Matthew struck Haiti and the southeastern U.S. last fall, DigitalGlobe was able to detect areas covered by water before and after the hurricane and combine those regions with open-source data from OpenStreetMap to determine where roads and buildings were submerged or at heightened risk of flooding.

DigitalGlobe also maintains crowdsourcing networks of people who can quickly hand-label images after a disaster, letting aid workers know where to look for damaged buildings and people potentially in need of assistance. That technique helped the company provide guidance to nonprofits and first responders after the 2015 earthquake in Nepal, Har-Noy says.

"We were the very first company to capture imagery of Kathmandu, which had a very large amount of damage," he says. "What we were able to do is capture imagery, run it by these crowdsourcing detections in order to detect rubble, destroyed buildings, etc., and we were able to make it accessible to various NGOs that were being deployed on the ground."

Crowdsourcing can also help with training and testing machine-learning algorithms, since human workers from around the world can be paid to quickly label features on satellite images or verify labels applied by machines, he says.

And while both crowdsourcing and automation let DigitalGlobe and its customers extract more detail from satellite images than ever before, Har-Noy says the company takes steps to protect people's privacy and safety. The resolution of commercial satellite images, regulated by the U.S. Commerce Department, means people aren't recognizable at the level of detail the company releases, Har-Noy says. DigitalGlobe also doesn't release images of active U.S. combat areas.

"We take privacy very seriously with regard to the information we release, we make accessible, and also that other people derive," Har-Noy says.

Still, machine learning's apparent success in understanding satellite images raises questions about how automated processing of higher-resolution photos, like those from aerial photography, may impact privacy in the future, says Jay Stanley, senior policy analyst at the American Civil Liberties Union.

"This is part of a larger trend, which is that machine learning and artificial intelligence techniques in general are allowing monitoring to take place on a mass scale that used to be very expensive because it required human attention," he says.

As wide-area aerial imaging, such as the controversial plane-based surveillance reportedly deployed in Baltimore last year and the balloon-mounted cameras used at last year's Summer Olympics in Rio de Janeiro, potentially becomes more common, lawmakers and the public will have to decide appropriate limits on how the technology and data are used and automatically processed, Stanley says.

"These are decisions we're going to have to make as a society if we want to go down the road of allowing us to be monitored by machine," he says. "If you start adding more advanced analytics to a tool like that, you could do some very invasive things."

Cameras mounted on fishing boats combined with advanced machine learning can help better monitor legal and illegal catches, advocates say.

The numbers are shocking.

Overexploitation of the world's fish stocks is growing at an alarming rate, says the U.N. Food and Agriculture Organisation (FAO).

Nearly 90% are either at or exceeding sustainable capacity, and in less than 10 years, production is set to grow by 17%.

One of the keys to sustainable fisheries has been the employ of human monitors to watch what is being scooped up from the sea. In the United States, fishing boats are routinely accompanied by independent observers who track compliance with fishing regulations.

In other countries' waters, it's a whole different story, so those government and independent agencies hoping to halt overfishing are turning to some of the same digital tools that let social media sites recognize faces in photos.

In the region of the Pacific Ocean from Indonesia and the Philippines to Hawaii—the source of the majority of the world's tuna harvest—a mere 2% of fishing operations are watched by observers, says Mark Zimring, director of The Nature Conservancy's Indo-Pacific tuna program.

That makes it harder for scientists to understand the effects fishing is having on imperiled species in the area, and that means large numbers of valuable fish—including the yellowfin and bigeye tuna varieties found in sushi restaurants around the world—are being harvested illegally, he says.

But by effectively automating part of the job of the observer by using cameras to record what creatures are caught and sophisticated software to classify them by species, regulators would be able to get a fuller picture of legal harvests and detect unlawful operations.

Just as automation and machine learning have given internet companies detailed records and predictions of how users behave online, they can potentially enable scientists and government agencies to build similarly detailed models of the world's fisheries.

"Today it's estimated that what's called illegal, unreported, and unregulated fishing costs the region between a half billion dollars and $1.5 billion a year," Zimring says. And even low rates of accidental capture and killing of important predators such as sharks can have significant impacts on aquatic ecosystems, he says.

"We really need to understand, from a kind of science perspective, when at-risk species like sharks and turtles are being caught, and what really happens to them," he says.

The Nature Conservancy is working with governments in the region, including in Palau, Micronesia, the Marshall Islands, and the Solomon Islands, to implement alternative monitoring programs, capturing video footage of fishing vessels instead of placing observers on each boat. The group has equipped about a dozen boats with cameras so far, with plans to place more this year.

Footage is stored on hard drives that can be removed for analysis when boats come into port. But that method still produces huge amounts of raw video that need to be manually analyzed by hand. To make analyzing the footage more feasible, The Nature Conservancy is investigating ways to use machine learning techniques to process that video material.

"You've got to be able to translate that raw video data into useful information—that's where we think machine learning can really help," says Zimring. The group is sponsoring a $150,000 challenge on machine learning competition platform Kaggle, seeking algorithms that can optimally classify types of sea creatures caught by fishing boats.

Even if automated methods can't completely replace human analysis, they can still provide a big benefit if they can reduce the amount of footage people need to review, saving time and potentially making it possible to transmit the video material more efficiently, Zimring says.

Similarly, the National Oceanic and Atmospheric Administration (NOAA) is investigating using electronic monitoring in fisheries across the U.S. from the Atlantic to the Pacific. One pilot program is placing cameras on board smaller boats in Alaska that catch fish like cod and halibut.

"It's been difficult to get observers onto these boats, and we've been working with them to develop electronic monitoring as an alternative," says Chris Rilling, director of the Fisheries Monitoring and Analysis Division at NOAA's Alaska Fisheries Science Center. Right now, fishing boats participating in a voluntary pilot program are equipped with cameras and with hard drives that are periodically shipped to NOAA for human review, but the agency is developing sets of training data for machine learning algorithms to use, with an eye toward integrating automated classification around 2019.

"That training data set has to come from the fishery that you're worried about, because you need to understand the complexity of the species," says NOAA researcher Farron Wallace. And even if the machine learning algorithms aren't perfect, they can still provide valuable information, since scientists will be able to test the inherent uncertainty in the data they produce, he says.

NOAA is also experimenting with ways to electronically monitor fish populations while they're still underwater. One device debuted in 2015 uses a low-powered computer and sonar system attached to the seafloor to automatically survey fish populations.

"These instruments wake up and send out pings, and we measure the energy that's reflected, and that tells us about how many fish are there," says NOAA biologist Alex De Robertis. When the device's work is done, NOAA researchers send it a specific audio signal that causes it to detach from the sea floor using attached flotation devices to rise to the surface.

This summer, the agency plans to place some of the devices in a remote area of the Arctic usually difficult to access without expensive icebreaker ships during the winter, and leave them in place to track fish populations and movements after the winter freeze.

"I think the real strength of this is to look over long time periods," says De Robertis. "It's ice-covered and very hostile and very expensive to work in there in the winter time."

To spread his message and rise to power, Trump relied on an open internet. Will his administration now kill it?

On the eve of Donald Trump's inauguration, one of the biggest tech-related questions surrounding his administration is the future of net neutrality. The Federal Communications Commission under President Barack Obama has taken steps to prevent internet service providers from favoring certain kinds of content over others, thereby preserving the core principles of a free and open internet. But so far, it's unclear whether Trump's future FCC appointees will pursue the same policies or adopt the more hands-off approach favored by many Republicans.

Either way, experts say, the new administration could have profound consequences for how consumers use the internet and how internet-based companies connect with their customers.

Signs of a major policy shift are already in the air. In a report issued last week—just 10 days before Trump is to be sworn in—the FCC warned that AT&T and Verizon may be violating net neutrality rules by privileging their own streaming video services over those from competitors. But a Republican member of the FCC quickly denounced the document as a "regulatory spasm" unlikely to have significance after Trump takes office.

Both companies also criticized the report, saying their zero-rating programs have been welcomed by their customers, even as a group of Democratic senators praised the FCC's ruling as a victory for consumers.

"This report evaluating whether zero rating plans violate net neutrality will help make sure the internet remains the free and open platform that it's always been," said Senator Al Franken, a Minnesota Democrat, in a statement.

In the meantime, commentators on both sides of the debate are expressing skepticism that regulators will continue on the same trajectory under the new administration. Trump and other prominent Republicans have critiqued the FCC's policies as unnecessary regulation of a still-evolving industry, and FCC Chairman Tom Wheeler, an Obama appointee, is slated to depart the commission upon Trump's inauguration.

That will leave the FCC's board of commissioners with a Republican majority and an impetus to reverse course. "This report, which I only saw after the FCC released the document, does not reflect the views of the majority of Commissioners," said FCC Commissioner Ajit Pai, a Republican, in a statement. "Fortunately, I am confident that this latest regulatory spasm will not have any impact on the Commission's policymaking or enforcement activities following next week's inauguration."

Pai reportedly met with Trump this past weekend and has been floated as a possible candidate to replace Wheeler as FCC chairman.

A Non-Neutral Issue

In 2015, the FCC's commissioners enacted its current net neutrality policy, known as the Open Internet Order, in a 3-2 vote along party lines. The order asserted the FCC's right under federal law to regulate broadband internet similarly to traditional phone service, drawing praise from Democratic legislators and civil liberties groups like the American Civil Liberties Union and Electronic Frontier Foundation. They had expressed concern that internet providers could slow or even block controversial content or content from publishers unwilling or unable to pay to have it expedited to consumers.

"The FCC did the right thing by adopting strong rules for a free & open Internet," Elizabeth Warren, the Democratic senator from Massachusetts, tweeted at the time. "This is a huge victory for the little guys."

But telecom industry groups and Republican leaders slammed the ruling as a regulatory overreach, while Trump warned in a 2014 tweet that neutrality regulations would be used to "target conservative media." Internet providers and trade groups sued unsuccessfully to overturn the order—a D.C. federal appeals court that had blocked previous net neutrality rules upheld the FCC's right to regulate broadband similarly to other telecom services—and urged Congress to limit the agency's authority.

Trump hasn't commented publicly on the issue since his election, nor has he indicated who he'd nominate to fill Wheeler's position or another open commissioner's seat at the FCC. But some of his top tech advisers have backed calls to reduce the telecom regulator's clout. Adviser Mark Jamison, an affiliate of the conservative American Enterprise Institute, seemed to call in October for sharply shrinking the agency. Internet mogul and Trump ally Peter Thiel has also questioned the need for net neutrality regulations.

"In terms of net neutrality, I think their intention is to deregulate the cable and telephone industry completely," says Ernesto Omar Falcon, legislative counsel at the Electronic Frontier Foundation. "I think that's their intention right off the get-go, and I think that's a fight that we'll have to engage in pretty quickly."

Traditionally, the FCC doesn't issue major new policies in the absence of a permanent chair, as will be the case after Wheeler's departure, but it's unclear whether the regulator will maintain that tradition under Trump, Falcon says. But any push to revoke the existing Open Internet Order would require a period of public comment and could potentially spur Congressional hearings as well, he says.

And while Republicans and telecom groups have publicly denounced the FCC rules, business groups are also averse to regulatory uncertainty and shifting legal frameworks, says Harold Feld, senior vice president at the pro-neutrality group Public Knowledge.

"Ted Cruz once said that net neutrality is Obamacare for the internet," Feld says. "That turns out to be true, in that it turns out to be something that's very complicated to repeal and replace, because the internet is critically import to everybody and businesses have invested lots of money in the understanding that this is a stable platform that's gonna have some kind of net neutrality."

A spokesman for the NCTA, an industry group formerly known as the National Cable & Telecommunications Association, declined to comment.

Trump in 2014 expressed concern about the rules stifling conservative voices, negatively comparing net neutrality to the FCC's onetime Fairness Doctrine, which required broadcasters to present balanced coverage of controversial issues before it was rescinded under President Ronald Reagan. But more recently, Trump has repeatedly slammed mainstream media organizations for perceived bias against his campaign and rising administration.

"It is a totally one-sided, biased show—nothing funny at all," he wrote in a post-election tweet about Saturday Night Live, broadcast by the Comcast-owned NBC. "Equal time for us?"

And Trump and his supporters may now worry about ISPs discriminating against conservative publishers like Breitbart News, the controversial website where Trump senior advisor Steve Bannon served as chairman, Feld suggests.

"I can't see Trump or Steve Bannon—both of whom used unfiltered access to social media as a critical tool in their rise to power—want a situation in which the media companies they distrust like Comcast or AT&T can do what they like,"Feld wrote in a December blog post. "While they will certainly be all for deregulating companies, they will want to keep the threat of regulation over their heads."

The clock is now the closest it's been to midnight since the 1950s, with scientists citing "alternative facts" and threat of nuclear arms.

In a series of statements alluding to President Donald Trump's fiery rhetoric, experts from the Bulletin of Atomic Scientists moved its Doomsday Clock forward to two-and-a-half minutes to midnight on Thursday. They also cited the threat of nuclear arms and climate change, as well as a rise in nationalism and a disregard among political leaders for scientific facts. The clock is now the closest it's been to midnight—symbolizing, essentially, human extinction—since the 1950s, during the early Cold War.

"The current political situation in the United States is of particular concern," said David Titley, a member of the Bulletin's Science and Security Board and founding director of Pennsylvania State University's Center for Solutions to Weather and Climate Risk, in a press conference announcing the change.

Titley called upon the Trump administration to acknowledge the human role in warming the planet and emphasized the need to reduce carbon emissions.

"Climate change should not be a partisan, political issue—the well-established physics of the earth are neither liberal nor conservative in character," he said. "Alternative facts will not make the challenges of climate change magically go away."

The reference to "alternative facts" was an allusion to a phrase Trump advisor Kellyanne Conway used to defend disputed claims about the size of the crowd at Trump's inauguration and has come to symbolize the new administration's tendency to embrace questionable, or outright false, claims. It wasn't the only reference at the Doomsday Clock press conference to the new U.S. president, as Rachel Bronson, executive director and publisher of the Bulletin, cited other concerns beyond the particular details of nuclear buildup and greenhouse gas emissions.

"The first has been the cavalier and reckless language used across the globe, especially in the United States, during the presidential election and after, around nuclear weapons and nuclear threats," she said. "The second has been a growing disregard of scientific expertise."

The clock, first created in 1947 to highlight the risks of the Cold War-era nuclear arms race, had previously been set to three minutes to midnight in 2015. Back in the 1950s, during a particularly tense moment in the Cold War, when the United States and Soviet Union both developed and tested the hydrogen bomb, the Bulletin set the clock to just two minutes from midnight. The clock was furthest from midnight, set to 17 minutes before the hour, in 1991, as Cold War tensions eased and the U.S. and Russia made progress on arms reduction.

"The Doomsday Clock is closer to midnight than it's ever been in the lifetime of almost everyone in this room," said Lawrence Krauss, chair of the journal's board of sponsors and director of the Origins Project at Arizona State University.

Speakers from the Bulletin also pointed to efforts by the United States and Russia to modernize their nuclear arsenals, as well as nuclear developments in other nations like North Korea, India, and Pakistan and Trump's suggestion during the presidential election that South Korea and Japan might consider developing their own atomic weapons.

"Saber rattling and loose but dangerous rhetoric have become almost commonplace," said Rachel Bronson.

The publication, which was founded in 1945 by scientists who participated in the Manhattan Project that developed the initial atomic bomb and said they "could not remain aloof to the consequences of their work," also pointed to rising threats from cyberattacks and biological warfare.

While computer hacking is not a new development, its role in undermining confidence in democratic institutions and the results of the recent election is, warned Krauss. And while new gene editing technologies like the tool known as CRISPR may help cure disease, they may also make it easier for even relatively unsophisticated attackers to unleash biological attacks, he said.

"Technological innovation is occurring at a speed that challenges society's ability to keep pace, even as many citizens lose faith in the institutions upon which they must rely to make scientific innovation work for them rather than against them," said Krauss.

With the press conference streamed through the Bulletin's website and on Facebook Live, Krauss called upon individual viewers to urge political leaders to address the problems of the day.

"We chose the Doomsday Clock 70 years ago, because we feel it allows us a rare opportunity to reach the global public directly with an enduring icon and to raise the profile of urgent, global, existential threats that the public needs to be aware of to act responsibly," Krauss said. "The future of the clock, and our future, is in your hands."

The company famous for email campaigns wants to be a one-stop shop for online marketing.

While MailChimp has made its name as an email marketing tool, the Atlanta company is branching out into other areas of digital advertising, starting with a feature launched on Thursday that lets customers buy and manage Facebook ads through the platform.

That means that MailChimp's 15 million users—many of them small e-commerce businesses using the service to communicate with their own customers—will also be able to reach audiences through Facebook without having to separately log in to the social networking site to manage and track ad campaigns.

"If you use Facebook's web interface, it's actually kind of difficult," says MailChimp CEO Ben Chestnut. "We take away a lot of the steps, a lot of the minutiae, that Facebook has in its interface."

In the future, the company plans to add integrations with other ad systems, likely including other social networks and search engines, Chestnut says. The move is part of a broader push by MailChimp to provide easy-to-use marketing tools particularly aimed at small online businesses.

Last year, MailChimp began to enable e-commerce clients to add data-driven product recommendations to their marketing emails, which Chestnut says boosts participating customer revenue by an average of 31%. The company offers other commerce-focused features like automated reminder emails when users depart shopping sites with items left in their carts.

MailChimp has even done some early testing of postal mail integration, contemplating features like physical postcards with discount codes linked to email campaigns, he says.

"One day down the road you might see people using MailChimp to build direct-mail campaigns," Chestnut says.

Customers Old And New

MailChimp worked with Facebook in order to develop the feature, which will let users target ad campaigns based on their existing MailChimp email lists or to audiences developed within the Facebook platform. It will also allow businesses to use Facebook's "lookalike" audience feature to reach Facebook users similar to their existing customer base.

Once customers do pick their target Facebook audiences, they can craft ads to bring users to their home pages or—if they're one of the e-commerce businesses that represent about 46% of MailChimp's revenue—highlight particular products or encourage them to sign up for MailChimp-powered emails highlighting future offers.

Some customers had previously approached MailChimp to ask if the company was able to sell lists of email marketing leads, which isn't possible given anti-spam laws and industry conventions. But by using social media marketing tools, companies are able to reach new potential users and have a shot at convincing them to register to receive marketing emails, Chestnut says.

"We've been in beta since mid-November, and we're seeing customers use it for everything," says Chestnut. Testers include some existing Facebook advertisers who have now mostly migrated away from Facebook's own marketing interface. The system integrates with popular online store software, letting customers closely track how the ads they place impact sales, he says.

MailChimp doesn't intend to charge extra for the Facebook integration, with any money spent on Facebook ads going straight to the social networking company. But Chestnut points out that success on Facebook can lead to businesses expanding their mailing lists and signing up for larger-capacity MailChimp plans.

And moving forward, in addition to adding integrations with new advertising platforms, the company plans to further explore how its existing service can be more closely tied in with other marketing avenues, such as sending emails to customers who responded to particular Facebook ads.

"We're adding this whole new channel, which is a big deal to us, but what's really going to be interesting is interweaving new channels with email," Chestnut says.

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.

Hiring managers can spend hours using primitive keyword search tools to sift through half-relevant resumes on job boards, and workers with in-demand skills get bombarded with emails from recruiters offering them jobs they're not particularly interested in, says Ed Donner, cofounder and CEO of New York startup Untapt.

"Hiring tech people is an incredible pain point," says Donner, who previously headed a technology team with hundreds of employees at JPMorgan Chase. "It's still impossibly hard to find talent."

Untapt is one of a number of companies looking to make it easier to digitally dig through piles of resumes, using machine learning techniques to develop algorithms that predict how well-suited a candidate is for a job. Advocates and industry experts say that adding automation to recruiting can save time and money and can potentially help hiring managers find and consider a more diverse set of applicants.

Helping to Eliminate Bias

"People are very subject to unconscious bias, and with the help of the machine, you can overcome a lot of the unconscious bias," says Tom Haak, director of the HR Trend Institute. "When the machine learns what you're looking for, it can say, if you find this type of candidate positive, here are other ones."

Untapt, which launched in early 2015, is focused on filling financial tech positions, and its software is designed to locate candidates with particular technical skills more efficiently than could be done through traditional text searches of resumes and cover letters. If a company is looking to hire developers adept at functional programming, a software development technique popular in parts of the fintech world, Untapt's tool can learn to recognize that experience in certain programming languages is likely a good sign, Donner says.

"You're dealing with this very noisy data set with a lot of information," he says. "That kind of problem is perfectly suited to artificial intelligence and to machine learning."

The company keeps track of which candidates its software surfaces actually get interviewed and for which positions, and uses that information to continually refine its algorithms, which Donner says often already outperform human recruiters in terms of the percentage of candidates called in for interviews.

Untapt also looks to surface candidates from underrepresented groups—such as women, veterans, and people of color—and more generally find candidates who might be overlooked by traditional recruiters.

"We also have features in Untapt [where] clients can choose to mask names and photos on the resumes they're presented with, so they're making blind decisions," says Donner.

The company attracts jobseekers through a mix of techniques, including online ads, blog posts, and appearances at industry events, and only shares their resume information with companies they agree to connect with, he says. That helps attract candidates who might not want to actively participate on more traditional job search boards but are interested in exploring their options, he says.

Machine learning can also help companies efficiently find potential candidates amid resumes they already have on file when new positions open up, says Steve Goodman, cofounder and CEO of San Francisco-based Restless Bandit. The company specializes in what it calls "talent rediscovery," resurfacing applicants to its clients' previous job postings that could be promising for new positions. To suggest potential candidates for a position, it looks at application records and digital hiring system data indicating how various previous applicants to similar roles have fared, learning over time which resumes to highlight for which openings. "We learn the hiring patterns of the company," Goodman says.

In the future, the company will likely incorporate more data about how successful candidates fare at their new jobs, including information pulled from human resources systems tracking how candidates progress up company ladders. Restless Bandit also can already fetch publicly available data from the internet about job candidates, so companies can keep their files up to date as they consider them for new positions.

"One of the things we do for companies is, we actually go out on the open web," says Goodman. "We try to find that person, and we try to enrich that resume with the latest data."

A Replacement For The "Gut Feeling"?

What machine learning systems can't, at present, do is eliminate the need for job interviews, he says. They can't evaluate applicants' body language and conversational skills, or estimate how they'll fit into a given team's existing culture, meaning there's still a definite need for humans in the hiring process.

"We can tell you the 15 people to interview, or the 20 people to interview," he says. "We can't tell you who to hire."

Still, some companies are even developing tools that use machine learning to evaluate job candidates' personality traits and aptitudes for a job. Ideally, artificial intelligence can help companies move beyond basic personality quizzes that critics have long said can be easily gamed by applicants able to guess the "right answers" for a given job.

Cognisess, based in Bath, England, offers about 40 different game-like assessments through its software platform, measuring factors like working memory, empathy, and ability to divide attention among multiple tasks. Based on scores of existing successful workers, the platform develops profiles of who would be well-suited to a particular position and indicates how candidates measure up, says cofounder and CEO Chris Butt.

Since the optimal results on the various tests vary from position to position, it's difficult for applicants to rig performance, he says. "You don't know what a good match looks like as the candidate," he says.

In general, applying artificial intelligence to hiring may mean a shift away from intuitive decisions and toward more data-driven processes, says Dan Ryan, principal at Ryan Search and Consulting and a member of the Society for Human Resource Management's talent acquisition panel.

"Part of the problem with the employment, and also with the sourcing process, is there is still a reliance on gut decisions versus data-driven decisions," he says. "If you could actually have data that shows using the data-driven approach could produce better results versus using the gut feeling, that's where the results will come."